This is Bugs94-I.gz in view mode; [Up]
Newsgroups: comp.sys.next.bugs,comp.sys.next.programmer From: laurie@eskimo.com (Laurie Anderson) Subject: packages on white Message-ID: <CIvGHC.8Lt@eskimo.com> Summary: can't install packages on white Organization: Eskimo North (206) For-Ever Distribution: usa Date: Thu, 30 Dec 1993 23:41:31 GMT Hi. I have developed an application under NEXTSTEP486 that I wish to distribute. The application's size is larger than the capacity of a 1.44m floppy. Therefore, I have chunked and packaged it. My problem comes when I go to install the application. Installer asks for the first floppy, partially installs it, the complains that it can't eject the floppy. This is when my troubles begin. Installer doesn't unmount the floppy. If I attempt to eject the floppy or drag it to the recycler I get a window that states I can't do that because someone is using it. I haven't found the technique to install chunked packages for NEXTSTEP486. What's the trick?
Newsgroups: comp.sys.next.bugs,comp.sys.next.programmer From: laurie@eskimo.com (Laurie Anderson) Subject: problems with packages on white Message-ID: <CIvGzs.8w3@eskimo.com> Organization: Eskimo North (206) For-Ever Distribution: usa Date: Thu, 30 Dec 1993 23:52:34 GMT Hi. I have developed an application under NEXTSTEP486 that I wish to distribute. The application's size is larger than the capacity of a 1.44m floppy. Therefore, I have chunked and packaged it. My problem comes when I go to install the application. Installer asks for the first floppy, partially installs it, the complains that it can't eject the floppy. This is when my troubles begin. Installer doesn't unmount the floppy. If I attempt to eject the floppy or drag it to the recycler I get a window that states I can't do that because someone is using it. I haven't found the technique to install chunked packages for NEXTSTEP486. What's the trick?
Newsgroups: comp.sys.next.bugs From: hsr@cs.Stanford.EDU (Scott Roy) Subject: Copy is broken in List, HashTable, and other basic classes Message-ID: <1993Dec31.093849.24519@CSD-NewsHost.Stanford.EDU> Sender: news@CSD-NewsHost.Stanford.EDU Organization: Computer Science Department, Stanford University, California, USA Date: Fri, 31 Dec 1993 09:38:49 GMT BUG: Copy is broken in List, HashTable, and other basic classes. OS: NEXTSTEP 3.1 Motorola WORKAROUND: Define copyFromZone: in all subclasses. Instance variables declared in subclasses are not copied automatically, as one might expect. Here are some test results: schmendrick% cc a.m -lNeXT_s ; a.out TEST: All values should equal 100 List_x = 100 Copy = 0 Object_x = 100 Copy = 100 Storage_x = 100 Copy = 0 View_x = 100 Copy = 100 Control_x = 100 Copy = 100 HashTable_x = 100 Copy = 0 NXBundle_x = 100 Copy = 100 Cell_x = 100 Copy = 100 NXData_x = 100 Copy = 0 NXStringTable_x = 100 Copy = 0 ButtonCell_x = 100 Copy = 100 schmendrick% The test creates a subclass with a single new instance variable, asigns it the value 100, and then calls the copy method. Ouch! It seems obvious that List, Storage, and the other broken classes are never passing copyFromZone: messages up to Object. Should be a one line fix. For now, the obvious workaround is to define copyFromZone: in all subclasses: - copyFromZone: (NXZone *) zone { id copy = [super copyFromZone: zone]; copy->x = self->x; return copy; } I've attached my test code in case you wish to duplicate the problem. Scott Roy Department of Computer Science Stanford University ------------------------------------------------------------------------- #import <appkit/appkit.h> #define SimpleSubClass(CLASS, SUBCLASS) \ \ @interface SUBCLASS:CLASS { @public int x; } @end \ @implementation SUBCLASS \ - init \ { \ self = [super init]; x = 100; return self; \ } \ @end #define PrintOut(SUBCLASS) \ { \ SUBCLASS *b, *a = [[SUBCLASS alloc] init]; \ \ b = [a copy]; \ printf("%17s = %d\tCopy = %d\n", [a name], a->x, b->x); \ } SimpleSubClass(List, List_x) SimpleSubClass(Storage, Storage_x) SimpleSubClass(Object, Object_x) SimpleSubClass(View, View_x) SimpleSubClass(Control, Control_x) SimpleSubClass(HashTable, HashTable_x) SimpleSubClass(NXBundle, NXBundle_x) SimpleSubClass(Cell, Cell_x) SimpleSubClass(NXData, NXData_x) SimpleSubClass(NXStringTable, NXStringTable_x) SimpleSubClass(ButtonCell, ButtonCell_x) main () { [Application new]; printf("\nTEST: All values should equal 100\n\n"); PrintOut(List_x); PrintOut(Object_x); PrintOut(Storage_x) PrintOut(View_x) PrintOut(Control_x) PrintOut(HashTable_x) PrintOut(NXBundle_x) PrintOut(Cell_x) PrintOut(NXData_x) PrintOut(NXStringTable_x) PrintOut(ButtonCell_x) printf("\n"); }
From: otto@tukki.jyu.fi (Otto J. Makela) Newsgroups: comp.sys.next.bugs Subject: Re: Fax problems Date: 2 Jan 1994 07:47:58 +0200 Organization: The Crimson Permanent Assurance Company Sender: otto@network.cc.jyu.fi Message-ID: <OTTO.94Jan2074756@tukki.jyu.fi> References: <9312282335.AA00218@csn.org> In-reply-to: zaphod!tpg@csn.org's message of 28 Dec 1993 19:40:27 -0600 Talking about faxes and all that... where could I find a good set of how-to instructions on setting up a ZyXEL data/fax modem for use with a NeXTstation? I tried the paper documentation (how quaint!) and the NeXTanswers, but to no avail. -- /* * * Otto J. Makela <otto@jyu.fi> * * * * * * * * * * * * * * * * * * */ /* Phone: +358 41 613 847, BBS: +358 41 211 562 (V.32bis/USR-HST,24h/d) */ /* Mail: Kauppak. 1B18/FIN-40100 Jyvaskyla/Finland, ICBM: 62.14N 25.44E */ /* * * Computers Rule 01001111 01001011 * * * * * * * * * * * * * * * * */
From: sanguish@digifix.com Newsgroups: comp.sys.next.announce,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.bugs,comp.sys.next.programmer,comp.sys.next.marketplace Subject: Quick Guide to NEXTSTEP information on the Internet Date: 3 Jan 1994 01:35:26 -0500 Organization: Next Announcements Message-ID: <2g8ebe$b0h@digifix.digifix.com> This post is made weekly, to help 'point' users to more NEXTSTEP information Topics include: comp.sys.next newsgroups related newsgroups comp.sys.next newsgroups mailing list ftp sites NeXTanswers comp.sys.next.* newsgroups -------------------------- Comp.Sys.Next.Advocacy This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. Comp.Sys.Next.Announce Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) The NEXTSTEP FAQs are posted here monthly as well. This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Comp.Sys.Next.Bugs A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT- specific groups as well. Comp.Sys.Next.Hardware Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. Comp.Sys.Next.Marketplace NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. Comp.Sys.Next.Misc For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! Comp.Sys.Next.Programmer Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. The NEXTSTEP programmer FAQs are posted here. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. Comp.Sys.Next.Software This is a place to talk about [third party] software products that run on NEXTSTEP systems. Comp.Sys.Next.Sysadmin Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. related Newsgroups ------------------ Comp.Soft-Sys.Nextstep Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. Comp.Lang.Objective-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. Comp.Object Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original Comp.Sys.Next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News ------------------------------------------- Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep-d next-advocacy-d next-announce-d next-bugs-d next-hardware-d next-marketplace-d next-misc-d next-programmer-d next-software-d next-sysadmin-d (For a full description, send mail saying LISTS to <digestif@antigone.com>). The subscription syntax is essentially the same as LISTSERV's. To subscribe, send a message to <digestif@antigone.com> saying: SUB Listname YourName Example: SUB next-hardware-d John Doe The ftp sites ------------- cs.orst.edu: The main site for North American submissions nova.cc.purdue.edu: Lots of older stuff, but very short on disk space ftp.informatik.uni-muenchen.de: In Germany. ftp.next.com: See the below ftp.next.com and NextAnswers@next.com ------------------------------------- From the document 1000_Help from ftp.next.com Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, or you can transfer them by anonymous ftp. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to NeXTanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO <your-address> If you have any problem with the system or suggestions for improvement, please send mail to NeXTanswers-request@NeXT.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to NeXTanswers-request@NeXT.com. Written by: Eric P. Scott eps@toaster.SFSU.EDU and Scott Anguish sanguish@digifix.com Additions from: Greg Anderson (Greg_Anderson@afs.com) and Michael Pizolato (Michael_Pizolato@afs.com)
Newsgroups: comp.sys.next.bugs From: g402@hrz.uni-giessen.de (Karsten Wolf) Subject: Audio-CDs donīt play on Intel-Box with Apple-CD 150. Any fixes yet? Sender: news@muster.hrz.uni-giessen.de Message-ID: <CJ1qD7.97x@muster.hrz.uni-giessen.de> Date: Mon, 3 Jan 1994 09:00:42 GMT Organization: University of Giessen, Germany Subject says all and I remember some earlier mentions of this problem. Are there any fixes yet? Cheers, Karsten (karsten.wolf@erziehung.uni-giessen.de)
Newsgroups: comp.sys.next.programmer,comp.sys.next.advocacy,comp.sys.next.bugs From: cuilla@black_silk (Chris Cuilla) Subject: "The Not So Perfect Perfectionist" or "The Imperfectionist And His Dog" Message-ID: <1994Jan3.185136.25703@il.us.swissbank.com> Sender: root@il.us.swissbank.com (Operator) Organization: Swiss Bank Corporation CM&T Division Date: Mon, 3 Jan 1994 18:51:36 GMT I have just been humbled...(not the first time, surely won't be the last)... A friend (thankfully), gently (again, thankfully) brought my "perfect" signature line to my attention: "There's no such thing a as 'part-time' perfectionist." No...in case anyone ever wondered...this was not intended as a joke, but has turned out that way. Anyhow...I have corrected things now...and my new years resolution is to practice perfectionism while I preach it. -- Chris Cuilla --- NEXTSTEP Developer chris@its.com --------------------------------------------------------------------- "There's no such thing as a 'part-time' perfectionist." -- NewsGrazer, a NeXTstep(tm) news reader, posting -- M>UQR=&8P7&%N<VE[7&9O;G1T8FQ<9C!<9FUO9&5R;B!#;W5R:65R.WT*7&UA M<F=L,3(P"EQM87)G<C$R,`I<<&%R9%QT>#$Q-3)<='@R,S`T7'1X,S0U-EQT M>#0V,#A<='@U-S8P7'1X-CDQ,EQT>#@P-C1<='@Y,C$V7'1X,3`S-CA<='@Q M,34R,%QF,%QB,%QI,%QU;&YO;F5<9G,R-%QF8S!<8V8P($D@:&%V92!J=7-T M(&)E96X@:'5M8FQE9"XN+BAN;W0@=&AE(&9I<G-T('1I;64L('-U<F5L>2!W M;VXG="!B92!T:&4@;&%S="DN+BY<"EP*02!F<FEE;F0@*'1H86YK9G5L;'DI M+"!G96YT;'D@*&%G86EN+"!T:&%N:V9U;&QY*2!B<F]U9VAT(&UY(")P97)F M96-T(B!S:6=N871U<F4@;&EN92!T;R!M>2!A='1E;G1I;VXZ7`I<"@DB5&AE M<F4G<R!N;R!S=6-H('1H:6YG(`I<8B!A(&%S"EQB,"`@)W!A<G0M=&EM92<@ M<&5R9F5C=&EO;FES="XB7`I<"DYO+BXN:6X@8V%S92!A;GEO;F4@979E<B!W M;VYD97)E9"XN+G1H:7,@=V%S(&YO="`*7&(@:6YT96YD960*7&(P("!A<R!A M(&IO:V4L(&)U="!H87,@='5R;F5D(&]U="!T:&%T('=A>2X@($%N>6AO=RXN M+DD@:&%V92!C;W)R96-T960@=&AI;F=S(&YO=RXN+F%N9"!M>2!N97<@>65A M<G,@<F5S;VQU=&EO;B!I<R!T;R!P<F%C=&EC92!P97)F96-T:6]N:7-M('=H M:6QE($D@<')E86-H(&ET+@I<"EP*+2U<"D-H<FES($-U:6QL85P*+2TM7`I. M15A44U1%4"!$979E;&]P97)<"F-H<FES0&ET<RYC;VU<"BTM+2TM+2TM+2TM M+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM M+2TM+2TM+2TM+2TM+5P*(E1H97)E)W,@;F\@<W5C:"!T:&EN9R!A<R!A("=P >87)T+71I;64G('!E<F9E8W1I;VYI<W0N(EP*"GT* `
Newsgroups: comp.sys.next.bugs From: brianw@sounds.wa.com (Brian Willoughby) Subject: Re: FAX won't render some PostScript? Message-ID: <CJ2LzJ.4zq@sounds.wa.com> Organization: Sound Consulting, Bellevue, WA, USA References: <MS-C.756719781.377401575.mrc@Ikkoku-Kan.Panda.COM> Date: Mon, 3 Jan 1994 20:23:42 GMT Mark Crispin <mrc@Ikkoku-Kan.Panda.COM> writes: >NeXT won't talk to you unless you give them a VISA number, and there is an >objection to paying to be told ``we don't know the answer''. Nor can it wait >until a new release; this application requires a plug & play solution. I remember asking 800-NeXT about their pricing, and I seem to remember that there is no charge for "we don't know the answer". Plus, their rates are (were?) much lower than standard consulting rates. I wasn't working for a company doing NeXT work at the time I paid NeXT for some troubleshooting help, and I must say that the $40 I paid is much less than I would have charged them for the same amount of time. I would suggest calling NeXT. -- Brian Willoughby Software Design Engineer, BSEE from NCSU NeXTmail welcome Sound Consulting: Software Design and Development BrianW@SoundS.WA.com Bellevue, WA
From: glenn@rightbrain.com (Glenn Reid) Newsgroups: comp.sys.next.bugs Subject: Re: FAX won't render some PostScript? Message-ID: <1447@rtbrain.rightbrain.com> Date: 4 Jan 94 09:05:22 GMT References: <CJ2LzJ.4zq@sounds.wa.com> Sender: glenn@rightbrain.com Mark Crispin <mrc@Ikkoku-Kan.Panda.COM> writes: >NeXT won't talk to you unless you give them a VISA number, and there is an >objection to paying to be told ``we don't know the answer''. Nor can it wait >until a new release; this application requires a plug & play solution. So what happened with this? I got a form letter response from you that indicated that you were flooded with responses, presumably some of them helpful, and that you had forwarded them to the agent who was actually working on the case, er, I mean the programmer. Did he/she find the problem? Enquiring minds want to know (especially those of us who took the time to offer suggestions). If the problem isn't fixed, the answers to some simple questions, or a sample file that causes the problem, will help us get to the bottom of it. The net is a resource. Use it, don't complain to it. -- Glenn Reid glenn@rightbrain.com Woodside, California Help save my book, "Thinking in PostScript", from the cut-out bins!
From: dave@truth.asg.arlut.utexas.edu (Dave Slotnick) Newsgroups: comp.sys.next.bugs,comp.sys.next.programmer Subject: NXWriteType() bug? Date: 4 Jan 1994 16:11:37 GMT Organization: The University of Texas at Austin, Austin, Texas Message-ID: <2gc4fp$e8p@geraldo.cc.utexas.edu> This program: main () { NXTypedStream *stream; char foo[] = "foo"; // NOTICE THIS LINE stream = NXOpenTypedStreamForFile ("foo.test", NX_WRITEONLY); NXWriteType (stream, "*", &foo); NXCloseTypedStream (stream); } generates a segmentation fault, while this program main () { NXTypedStream *stream; char *foo = "foo"; // NOTICE THIS LINE stream = NXOpenTypedStreamForFile ("foo.test", NX_WRITEONLY); NXWriteType (stream, "*", &foo); NXCloseTypedStream (stream); } does not. Any thoughts? -- *** David Slotnick *** Applied Research Laboratories *** dave@truth.asg.arlut.utexas.edu NeXTmail desired and appreciated
From: Mark Crispin <mrc@Tomobiki-Cho.CAC.Washington.EDU> Newsgroups: comp.sys.next.bugs Subject: Re: FAX won't render some PostScript? Date: Tue, 4 Jan 1994 08:26:47 -0800 Organization: University of Washington Message-ID: <MS-C.757700807.1103527590.mrc@Tomobiki-Cho.CAC.Washington.EDU> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII To: glenn@rightbrain.com In-Reply-To: <1447@rtbrain.rightbrain.com> Hi Glenn - As I said, I forwarded all the messages to the fellow who was doing the work. I hadn't seen him since before the holidays. I just stopped by his office and asked. It turned out that none of the net.suggestions were the problem, but he independently figured it out. The FAX software puts a save/restore sequence around the entire job; this isn't done to the screen or printer. Certain complex PostScript may end up with junk on the stack (such as a redefinition of restore), so a save/restore wrapper isn't as safe as it would sound. The workaround is to nuke the stack at the end of the job, so the wrapper's restore is something sane. I hope I described this sensibly; I'm not a PS hacker. -- Mark --
From: M.Crawford@dcs.shef.ac.uk (Malcolm Crawford) Newsgroups: comp.sys.next.bugs Subject: Re: FAX won't render some PostScript? Date: 4 Jan 1994 11:35:30 -0600 Organization: UTexas Mail-to-News Gateway Sender: daemon@cs.utexas.edu Message-ID: <9401041736.AA09938@dcs.shef.ac.uk> > It turned out that none of the net.suggestions were the problem, but > he independently figured it out. > <phew> All by himself?! Me, I'd have thought that the net.suggestions were potential *solutions* to the problem, and that the problem itself was something to do with rendering PostScript; but then I never was much good at grammar... Have fun, mmalcolm.
From: e91tl@efd.lth.se (Tomas Liedman) Newsgroups: comp.sys.next.bugs,comp.sys.next.programmer Subject: Re: NXWriteType() bug? Date: 4 Jan 1994 18:20:13 GMT Organization: Lund Institute of Technology, Sweden Message-ID: <2gcc0t$kv1@nic.lth.se> References: <2gc4fp$e8p@geraldo.cc.utexas.edu> In article <2gc4fp$e8p@geraldo.cc.utexas.edu> dave@truth.asg.arlut.utexas.edu (Dave Slotnick) writes: > >This program: > >main () >{ > NXTypedStream *stream; > char foo[] = "foo"; // NOTICE THIS LINE > stream = NXOpenTypedStreamForFile ("foo.test", NX_WRITEONLY); > NXWriteType (stream, "*", &foo); > NXCloseTypedStream (stream); >} > >generates a segmentation fault, while this program > >main () >{ > NXTypedStream *stream; > char *foo = "foo"; // NOTICE THIS LINE > stream = NXOpenTypedStreamForFile ("foo.test", NX_WRITEONLY); > NXWriteType (stream, "*", &foo); > NXCloseTypedStream (stream); >} > >does not. > >Any thoughts? Could it be thet the compiler produces { 'f', 'o', 'o' } on the first case, and { 'f', 'o', 'o', '\0' } in the second case -- ie, the string is null-terminated only when declared as char *foo ? NXWriteType, expecting a null-terminated string, then falls past the end of foo[] in the first example, resulting in the segmentation fault mentioned. I think that's it. > >-- >*** David Slotnick >*** Applied Research Laboratories >*** dave@truth.asg.arlut.utexas.edu > >NeXTmail desired and appreciated -- | Christian Brunschen, Husmansv. 26, S-227 38 Lund, Sweden | | voice/fax/data +46 (0)46 139345, email d89cb@efd.lth.se | | PGP 2.2 Public Key available on request. | ObCelebrityQuote: | | "Oh, foo." | | - Andrew R. Koenig (ark@research.att.com) after dropping a piece | | of chalk during a lecture on C++ in Lund, Sweden, April 1993 |
From: "Thomas G. Mon" <tm31+@andrew.cmu.edu> Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs Subject: Re: NXWriteType() bug? Date: Tue, 4 Jan 1994 14:20:25 -0500 Organization: Senior, Math/Computer Science, Carnegie Mellon, Pittsburgh, PA Message-ID: <Mh_Q5tK00iV1Q2sV57@andrew.cmu.edu> In-Reply-To: <2gcc0t$kv1@nic.lth.se> >Could it be thet the compiler produces { 'f', 'o', 'o' } on the first >case, and { 'f', 'o', 'o', '\0' } in the second case -- ie, the string >is null-terminated only when declared as char *foo ? NXWriteType, >expecting a null-terminated string, then falls past the end of foo[] >in the first example, resulting in the segmentation fault mentioned. > >I think that's it. No, this isn't the case, according to K&R, both foo[] and *foo have the null character, '\0' at the end one looks like this: *foo []---->["The text that you want\0"] and the other one looks like this: foo[] ["The text that you want\0"] Tom. tm31@andrew.cmu.edu [NeXTmail welcome]
Newsgroups: comp.sys.next.bugs,comp.sys.next.programmer From: michal@phys.ualberta.ca (Michal Jaegermann) Subject: Re: NXWriteType() bug? Message-ID: <1994Jan4.200150.13732@kakwa.ucs.ualberta.ca> Sender: news@kakwa.ucs.ualberta.ca Organization: Department of Physics, University of Alberta References: <2gcc0t$kv1@nic.lth.se> Date: Tue, 4 Jan 1994 20:01:50 GMT e91tl@efd.lth.se (Tomas Liedman) writes: : In article <2gc4fp$e8p@geraldo.cc.utexas.edu> dave@truth.asg.arlut.utexas.edu (Dave Slotnick) writes (paraphrased): : > : >This program: : > : >main () : >{ : > NXTypedStream *stream; : > char foo[] = "foo"; // NOTICE THIS LINE - segmentation fault : > // char *foo = "foo"; // It is ok with this line instead : > stream = NXOpenTypedStreamForFile ("foo.test", NX_WRITEONLY); : > NXWriteType (stream, "*", &foo); : > NXCloseTypedStream (stream); : >} : > : >Any thoughts? : : Could it be thet the compiler produces { 'f', 'o', 'o' } on the first : case, and { 'f', 'o', 'o', '\0' } in the second case -- ie, the string : is null-terminated only when declared as char *foo ? No, this is not the case. It is easy to check that sizeof of array foo is four in the first case and foo[3] = '\0', like it should. The problem is that declarations 'char foo[] = "foo";' and 'char *foo = "foo";' are not at all equivalent accordingly to a definition of C - UNLESS passed as arguments to a function. You may blame C designers for that often occuring confusion, especially that in many context you may indeed use one or another declaration of foo without any ill effects. A quick session with gdb shows what is going on. In the first case, i.e. 'char foo[] = "foo";' you will see: (gdb) p foo $1 = {"foo\000"} (gdb) p &foo $2 = (char (*)[4]) 0x3fffd60 "foo" and for the other one: (gdb) p foo $1 = 0x3ff0 "foo" (gdb) p &foo $2 = (char **) 0x3fffd60 From a trace it looks like that _NXEncodeString() is not willing to accept as a pointer to a string a pointer to something totally different. :-) Adding 'char *bar = foo;' and passing &bar to NXWriteType() makes this code working again. Please notice also another, often overlooked, difference between two different kinds of foo. You may write into an array of characters, but constant strings, accrodingly with ANSI Standard, are not writable. gcc allows you to hack around this limitation with -fwriteable-strings, but a code generated and its memory layout will be different. Michal Jaegermann michal@gortel.phys.ualberta.ca
From: gad@eclipse.its.rpi.edu (Garance A. Drosehn) Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs Subject: Re: NXWriteType() bug? Date: 4 Jan 1994 21:13:20 GMT Organization: Rensselaer Polytechnic Institute, Troy NY, USA Distribution: world Message-ID: <2gcm5g$5c5@usenet.rpi.edu> References: <2gc4fp$e8p@geraldo.cc.utexas.edu> dave@truth.asg.arlut.utexas.edu (Dave Slotnick) writes: > > This program: > > main () > { > NXTypedStream *stream; > char foo[] = "foo"; // NOTICE THIS LINE > stream = NXOpenTypedStreamForFile ("foo.test", NX_WRITEONLY); > NXWriteType (stream, "*", &foo); > NXCloseTypedStream (stream); > } > > generates a segmentation fault, while this program > > main () > { > NXTypedStream *stream; > char *foo = "foo"; // NOTICE THIS LINE > stream = NXOpenTypedStreamForFile ("foo.test", NX_WRITEONLY); > NXWriteType (stream, "*", &foo); > NXCloseTypedStream (stream); > } > > does not. > > Any thoughts? I believe you are busily noticing the wrong line. You need to also notice this line: NXWriteType (stream, "*", &foo); Now think about what you are asking the compiler to do in both cases. In the first case, you have an array, and you're sending the address of that array to NXWriteType. In the second case you have a pointer to an array, and you're sending the address of that *pointer* to NXWriteType. One extra level of indirection. Many C books talk about how an array of char and a pointer to a character string are the same thing. They are in some ways, but you have to pay attention when passing them as parameters, particularly when playing around with "&" to get the address of "something" (either of the characters themselves, or of a pointer to the characters, which is what NXWriteType really wants you to be sending it). This is not a bug in NXWriteType(). -- Garance Alistair Drosehn = gad@eclipse.its.rpi.edu ITS Systems Programmer (handles NeXT-type mail) Rensselaer Polytechnic Institute; Troy NY USA
Newsgroups: comp.sys.next.bugs,comp.sys.next.programmer From: samurai@amber.hasc.ca (Darcy Brockbank) Subject: Re: NXWriteType() bug? Message-ID: <2gca8l$4pg@amber.hasc.ca> Sender: news@sifon.cc.mcgill.ca Organization: The Hutchison Avenue Software Corp. References: <2gc4fp$e8p@geraldo.cc.utexas.edu> Date: Tue, 4 Jan 1994 17:50:13 GMT In article <2gc4fp$e8p@geraldo.cc.utexas.edu> dave@truth.asg.arlut.utexas.edu (Dave Slotnick) writes: > > char foo[] = "foo"; // NOTICE THIS LINE >generates a segmentation fault, while this program > char *foo = "foo"; // NOTICE THIS LINE > NXWriteType (stream, "*", &foo); > >does not. Arrays are not strings are not arrays although the C syntax in most cases allows us to use them interchangeably. You supplied the "*" type, which is a character string, when you should have used the array type "[4c]" I think... Remember that when you read these things back, you'll get your string pointer set to a new string, but the array should be allocated, and will be filled with what you wrote out. Anyway, that's my understanding of how it works, but I haven't done these things for a while, so I might be wrong... - darcy -- If it's dark when you leave the office and there's snow on the ground, you're in winter. If it's dark when you get up [...] you're in winter. This is not magical, this is not a wonderland; This is snow, ice and sunlight deprivation. Mon pays c'est shitty weather you can't see. -- Montreal Mirror (Qc. Winter)
From: scott@nic.gac.edu (Scott Hess) Newsgroups: comp.sys.next.bugs,comp.sys.next.programmer Subject: Re: NXWriteType() bug? Date: 4 Jan 94 17:35:00 Organization: Is a sign of weakness Message-ID: <SCOTT.94Jan4173500@nic.gac.edu> References: <2gc4fp$e8p@geraldo.cc.utexas.edu> In-reply-to: dave@truth.asg.arlut.utexas.edu's message of 4 Jan 1994 16:11:37 GMT In article <2gc4fp$e8p@geraldo.cc.utexas.edu>, dave@truth.asg.arlut.utexas.edu (Dave Slotnick) writes: >This program: > >main () >{ > NXTypedStream *stream; > char foo[] = "foo"; // NOTICE THIS LINE > stream = NXOpenTypedStreamForFile ("foo.test", NX_WRITEONLY); > NXWriteType (stream, "*", &foo); > NXCloseTypedStream (stream); >} > >generates a segmentation fault, while this program > >main () >{ > NXTypedStream *stream; > char *foo = "foo"; // NOTICE THIS LINE > stream = NXOpenTypedStreamForFile ("foo.test", NX_WRITEONLY); > NXWriteType (stream, "*", &foo); > NXCloseTypedStream (stream); >} > >does not. > >Any thoughts? That's a C problem. Note that the last parameter to NXWriteType() should be a pointer to the data you are writing - in this case a char ** (or, a pointer to a pointer to char). In the second case, you take a pointer to char and apply & to it - generating a pointer to a pointer to char. The important point is that there are two pieces of storage involved here, the char[] area that foo points to, and the char * area that &foo points to. But, in the first case, you have an array of char, and saying & at that generates a pointer to an array of char. There is only one area of storage, a char[]. Hmm. I've probably confused you more than helped. Here's a sample program that demonstrates the problem: howard:/tmp> cat barf.c main () { char foo[] = "foo"; char *bar = "bar"; printf( "foo==%p\n", foo); printf( "&foo==%p\n", &foo); printf( "bar==%p\n", bar); printf( "&bar==%p\n", &bar); } howard:/tmp> barf foo==0x3fffc54 &foo==0x3fffc54 bar==0x3fd4 &bar==0x3fffc50 howard:/tmp> Explaination? bar is a char * variable that points at a char[]. So, when I say bar, the compiler is going to get the value _from_ bar and use that. But, foo _is_ a char[], so when I say foo, the compiler is going to replace my reference with the value of foo directly. It may be easier to see if you consider a simpler problem that's conceptually almost identical: #define foo 10 unsigned bar=10; In this case, &bar is a pointer to the storage for bar. But, &foo is nonsense. [Why &foo works when foo is a char[] is because in that case there _is_ storage involved, whereas in the #define case there's nothing to point to.] Later, -- scott hess <scott@nic.gac.edu> <I can handle NeXTMail, but don't like it> Home: 12901 Upton Avenue South, #326 Burnsville, MN 55337 (612) 895-1208 Office: 101 W. Burnsville Pkwy, Suite 108F, Burnsville, MN 55337 890-1332 <I hate when something I've never heard of is called "The world famous ...">
From: dockd@storm.cs.orst.edu (Dion Dock) Newsgroups: comp.sys.next.bugs Subject: Re: Fax problems Message-ID: <2gd271INNkbg@flop.ENGR.ORST.EDU> Date: 5 Jan 94 00:38:57 GMT References: <9312282335.AA00218@csn.org> <OTTO.94Jan2074756@tukki.jyu.fi> Organization: Computer Science Department, Oregon State University Hmm, I thought you can either configure your modem for fax or data but not both at once. If you want to do that, you will need to buy a third party program like NXFax or Dfax. Either that or go to PrintMonitor and turn the fax on and off. (I can't get faxes to work with my Supra, but the NXFax demo makes it works just fine. Am I missing something?) -dion -- Dion Dock __ __ dockd@storm.cs.orst.edu / ) / ) / NeXT archive admin / / o ______ / / _____. /_ next-ftp@cs.orst.edu (no NeXTmail) /__/_<_(_) / <_ /__/_(_) (__/ <_
From: mra@bink.mit.edu (Michael Raskin Andrews) Newsgroups: comp.sys.next.bugs Subject: printer_manager Date: 5 Jan 1994 14:11:50 GMT Organization: Massachvsetts Institvte of Technology Message-ID: <2gehr6$20m@senator-bedfellow.MIT.EDU> Does anyone know why my printer_manager program gives me an error when I try to delete a queued item for faxing? I get an alert panel, saying "Can't change to spooling directory." Thanks.
From: max@Kolmogorov.gac.edu (Max Hailperin) Newsgroups: comp.sys.next.bugs,comp.sys.next.programmer Subject: Re: NXWriteType() bug? Date: 5 Jan 94 09:00:45 Organization: Gustavus Adolphus College, St. Peter, MN Message-ID: <MAX.94Jan5090045@Kolmogorov.gac.edu> References: <2gcc0t$kv1@nic.lth.se> <1994Jan4.200150.13732@kakwa.ucs.ualberta.ca> In-reply-to: michal@phys.ualberta.ca's message of Tue, 4 Jan 1994 20:01:50 GMT This post is only for academics and others who like details and history: In article <1994Jan4.200150.13732@kakwa.ucs.ualberta.ca> michal@phys.ualberta.ca (Michal Jaegermann) writes: ... The problem is that declarations 'char foo[] = "foo";' and 'char *foo = "foo";' are not at all equivalent accordingly to a definition of C - UNLESS passed as arguments to a function. You may blame C designers for that often occuring confusion, especially that in many context you may indeed use one or another declaration of foo without any ill effects. ... Actually, the two are 'equivalent' in contexts other than as arguments to a function. For example, if you do 'bar = foo' or 'foo+2' or 'foo[3]'. In general, any place the *rvalue* of foo is used, the two will be 'equivalent'. Arguments to functions are just one special case of this. I put 'equivalent' in quotation marks, because as Jaegermann later notes, there is a the subtle distinction of writability. Finally, rather than just blaming "C designers" (i.e. Dennis M. Ritchie) for the confusion caused by pointers and arrays being somewhat interchangable, but not entirely, it is worth asking the question: why did he do it that way? It turns out that there is some fascinating programming language history wrapped up in this question. Ritchie designed C by evolution from B and BCPL (by way of a short-lived intermediary called NB), and in those language there is no semantic distinction between foo declared to be an array vs. a pointer to an array. Even when foo is delcared to be an array, it is actually the name of a memory location holding the address at which the first element of the array is found. However, when Ritchie introduced structures (structs) into the language, he decided (for reasons made more explicit in the article cited below) to abandon this extra memory location storing the base address and instead have foo directly name the block of consecutive memory locations itself. However, and this is the crux of the matter, Ritchie didn't want to be any more incompatable with the existing B language than necessary. So, he put the now troublesome wart into the language, namely that when foo is an array, the expression foo is treated as though it were &foo. In other words, the reason why people are having trouble with arrays and pointers being almost-but-not-quite interchangable is because Ritchie worked from a historical legacy in which arrays literally were pointers. He wanted to make arrays no longer literally be pointers, but still act as much as possible as though they were. This, and much more of the fascinating history of C, is explained in his paper "The Development of the C Language," pp. 201-208, ACM SIGPLAN Notices, Volume 28, Number 3, March 1993, Preprints of The Second ACM SIGPLAN History of Programming Languages Confere (HOPL-II). [Presumably this has appeared or will appear in the final HOPL-II book.]
Newsgroups: comp.sys.next.bugs,comp.sys.next.programmer From: brad@instep.wimsey.bc.ca Subject: Bug? SybaseAdaptor and MAB applications Message-ID: <1994Jan5.193016.362@instep.wimsey.bc.ca> Sender: brad@instep.wimsey.bc.ca (Bradley Head) Organization: InStep Mobile Communications Inc. Date: Wed, 5 Jan 1994 19:30:16 GMT Problem: SybaseAdaptor fails to load when running a FAT app (compiled on Black hardware) on Intel hardware. FAT app successfully locates SybaseAdaptor on Black hardware. Details: * Compiled MAB (for Intel and M68K) * SybaseAdaptor not hard linked. * undefines in Makefile.preamble as per release notes: OTHER_LDFLAGS = -u libdbkit_s -u libNeXT_s -u libsys_s * Using 4.6 SybaseAdaptor on both black and intel hardware. * Using NEXTSTEP 3.2 on both architectures When the app is compiled (on black) skinny for either architecture, the SybaseAdaptor is successfully located at runtime. Does anyone have any clues as to why this is happening? Is this a genuine bug? thanks in advance for the response. Brad. -- Bradley Head Software Developer, InStep Mobile Communications Inc. brad@instep.wimsey.bc.ca (NeXTmail accepted) 604 872-7116 fax: 604 872-7125
From: friel@sccs.swarthmore.edu (Patrick W. Friel) Newsgroups: comp.sys.next.bugs Subject: NI Domain Problem Followup-To: comp.sys.next.bugs Date: 5 Jan 1994 20:36:04 GMT Organization: Swarthmore College Message-ID: <friel-050194153432@mac04.public.swarthmore.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit I am currently working on a machine which is not connecting with the local NetInfo domain. As a result, I cannot add any users. Could you please give me a few tips? I cannot find the relevant documentation which was supposedly on-line. Thanks. -PWF
Newsgroups: comp.sys.next.bugs From: alex@cs.umd.edu (Alex Blakemore) Subject: Backspace locks up in 3.1,3.2 until reboot Message-ID: <CJ6zqw.EL@genoa.com> Sender: alex@genoa.com (Alex Blakemore) Organization: Genoa Software Systems Date: Thu, 6 Jan 1994 05:11:20 GMT with NS 3.1, 3.2 running on a turbo slab, BackSpace frequently locks up (freezes in the middle of some module) I dont think other processes halt. The only solution I've found is to reboot. This happens about every 24-48 hours. I seem to remember this starting about the time I got NeXT's UUCP to work, but that may be a coincidence. It didnt do this in NS3.0 without UUCP. Does this happen to anyone else? Is there a workaround or fix? Sorry to provide so few details. -- Alex Blakemore alex@cs.umd.edu NeXT mail accepted
Newsgroups: comp.sys.next.bugs From: wave@media.mit.edu (Michael B. Johnson) Subject: PB's "install" command can fail... Message-ID: <1994Jan6.080236.20261@news.media.mit.edu> Sender: news@news.media.mit.edu (USENET News System) Organization: MIT Media Laboratory Date: Thu, 6 Jan 1994 08:02:36 GMT this is a bug in 3.1: As I was readying some software for distribution, I noticed that if your working directory for a project is nested deep in the file system's hierarchy, i.e. /wave/foo/bar/baz/yabba/dabba/doo/boy/this/is/a/long/path/myApp/PB.project, that the standard install provided by Project Builder will fail horribly. This seems to be due to the fact that it uses tar (as opposed to say, gtar) and that it has problems with path name length. When I'm a bit less tired I will submit this to NeXT, after I check it in 3.2. -- --> Michael B. Johnson -- wave@media.mit.edu --> MIT Media Lab -- Computer Graphics & Animation Group --> 20 Ames St. E15-023G -- (617) 547-0563 (day office) --> Cambridge, MA 02139 -- (617) 253-0663 (night office)
From: dave@truth.asg.arlut.utexas.edu (Dave Slotnick) Newsgroups: comp.sys.next.bugs Subject: Re: NXWriteType() bug? Date: 6 Jan 1994 14:04:58 GMT Organization: The University of Texas at Austin, Austin, Texas Message-ID: <2gh5qa$qd5@geraldo.cc.utexas.edu> References: <SCOTT.94Jan4173500@nic.gac.edu> Scott Hess writes > That's a C problem. Note that the last parameter to NXWriteType() > ... > scott hess <scott@nic.gac.edu> <I can handle NeXTMail, but don't like it> > Home: 12901 Upton Avenue South, #326 Burnsville, MN 55337 (612) 895-1208 > Office: 101 W. Burnsville Pkwy, Suite 108F, Burnsville, MN 55337 890-1332 > <I hate when something I've never heard of is called "The world famous .."> This is by the *far* the most elegant explanation of the lot!! Thanks for the lucid commentary, Scott. And thanks to everyone else who posted responses! -- *** David Slotnick *** Applied Research Laboratories *** dave@truth.asg.arlut.utexas.edu NeXTmail desired and appreciated
From: dave@truth.asg.arlut.utexas.edu (Dave Slotnick) Newsgroups: comp.sys.next.bugs Subject: Re: NXWriteType() bug? Date: 6 Jan 1994 14:05:17 GMT Organization: The University of Texas at Austin, Austin, Texas Message-ID: <2gh5qt$qeo@geraldo.cc.utexas.edu> References: <SCOTT.94Jan4173500@nic.gac.edu> Scott Hess writes > That's a C problem. Note that the last parameter to NXWriteType() > ... > scott hess <scott@nic.gac.edu> <I can handle NeXTMail, but don't like it> > Home: 12901 Upton Avenue South, #326 Burnsville, MN 55337 (612) 895-1208 > Office: 101 W. Burnsville Pkwy, Suite 108F, Burnsville, MN 55337 890-1332 > <I hate when something I've never heard of is called "The world famous .."> This is by the *far* the most elegant explanation of the lot!! Thanks for the lucid commentary, Scott. And thanks to everyone else who posted responses! -- *** David Slotnick *** Applied Research Laboratories *** dave@truth.asg.arlut.utexas.edu NeXTmail desired and appreciated
From: william@pinoko.berkeley.edu (William E. Grosso) Newsgroups: comp.sys.next.bugs Subject: Re: AAAAHHH! (simple request) Date: 6 Jan 1994 23:08:41 GMT Organization: University of California, Berkeley Message-ID: <2gi5lp$a5@agate.berkeley.edu> References: <1993Dec18.074900.13853@news.unt.edu> Byron Goodman writes >[entire post deleted] A simple request: Use descriptive subject titles. "AAAAHHH" is only a good subject line in alt.dentist.what.to.say The subject field was put there so that we can all save time/energy by avoiding articles we don't care about. Please use it. Bill Grosso
Newsgroups: comp.sys.next.bugs From: bobcook@slac.stanford.edu (Bob Cook) Subject: Re: Backspace locks up in 3.1,3.2 until reboot Message-ID: <CJ8nqs.Av8@unixhub.SLAC.Stanford.EDU> Sender: news@unixhub.SLAC.Stanford.EDU Organization: Stanford Linear Accelerator Center References: <CJ6zqw.EL@genoa.com> Date: Fri, 7 Jan 1994 02:47:16 GMT Alex Blakemore writes >with NS 3.1, 3.2 running on a turbo slab, >BackSpace frequently locks up (freezes in the middle >of some module) I dont think other processes halt. > >The only solution I've found is to reboot. >This happens about every 24-48 hours. > >I seem to remember this starting about the time I >got NeXT's UUCP to work, but that may be a coincidence. >It didnt do this in NS3.0 without UUCP. > >Does this happen to anyone else? I use NeXT's UUCP and BackSpace with 3.1 on a color turbo slab. I've never had this happen. -- Bob Cook bobcook@slac.stanford.edu Stanford Linear Accelerator Center NeXT mail okay
Newsgroups: comp.sys.next.bugs,comp.sys.next.hardware From: markus@isolde.stgt.sub.org (Markus Stoll) Subject: AudioCD and Apple CD150 Content-Type: text/plain; charset=ISO-8859-1 Message-ID: <1994Jan6.134912.7312@isolde.stgt.sub.org> Content-Transfer-Encoding: 8bit Organization: GbR Stoll & Stoeffler, Stuttgart, Germany Mime-Version: 1.0 Date: Thu, 6 Jan 1994 13:49:12 GMT Karsten Wolf (g402@hrz.uni-giessen.de) claimed: > Audio-CDs donīt play on Intel-Box with Apple-CD 150 Well, they dont play with CDPlayer.app on any NeXT. L. Anathea Brooks (gaia@wam.umd.edu) wrote: > Hmmm. I bought an Apple CD150 (on sale for $150) which > *seems* to be a white NeXT CD-ROM drive. But CDPlayer > does not work with it (it did w/ NeXT drive). Apparently I'm not the only one with an Apple CD150, who wants to play Audio-CDs. I wrote a very basic audio-CD-player for it. Unfortunately, I'm very busy right now. In february I will make a real CD-player out of it. If you guys cant wait and are willing to put up with this basic version (which allows start, stop, skipping tracks and shows actual track, index and time), i can place it on some ftp-servers. -- Markus Stoll, markus@isolde.stgt.sub.org (NeXTMail preferred) stoll@vaxph.mpi-stuttgart.mpg.de
From: rogata@is-next.umd.edu (Richard Scott Ogata) Newsgroups: comp.sys.next.hardware,comp.sys.next.bugs Subject: Toshiba 4400C Mouse/Keyboard dead in 3.2 Date: 8 Jan 1994 06:47:21 GMT Organization: University of Maryland, College Park Message-ID: <2glktp$l5u@umd5.umd.edu> OK, I am at my wit's end. I have a Toshiba 4400C which was running 3.1 just fine. I decided to upgrade to 3.2, and, as others upgrading this laptop to 3.2 have found, the keyboard croaked. Furthermore, my PS/2 mouse won't work either, unless I plug it into the serial port and use it as a serial mouse. I edited the kernel according to the instructions I found in this newsgroup, but there is no change in behavior. I can't find anything about this in NeXTanswers. Seeing how this is a bit of a "show-stopper", and that I believe there are other models affected, I don't understand this. Has anyone contacted NeXT about this? Or has anyone found an effective patch? Rich Ogata rogata@arpa.mil
Newsgroups: comp.sys.next.bugs From: mckelvey@fafnir.com (James W. McKelvey) Subject: Re: Backspace locks up in 3.1,3.2 until reboot Message-ID: <CJBu7M.838@fafnir.com> Sender: mckelvey@fafnir.com (James W. McKelvey) Organization: Fafnir Software, Saugus, CA, USA References: <CJ8nqs.Av8@unixhub.SLAC.Stanford.EDU> Date: Sat, 8 Jan 1994 19:59:45 GMT In article <CJ8nqs.Av8@unixhub.SLAC.Stanford.EDU> bobcook@slac.stanford.edu (Bob Cook) writes: > Alex Blakemore writes > >with NS 3.1, 3.2 running on a turbo slab, > >BackSpace frequently locks up (freezes in the middle > >of some module) I dont think other processes halt. > > > >The only solution I've found is to reboot. > >This happens about every 24-48 hours. > > > >I seem to remember this starting about the time I > >got NeXT's UUCP to work, but that may be a coincidence. > >It didnt do this in NS3.0 without UUCP. > > > >Does this happen to anyone else? > > I use NeXT's UUCP and BackSpace with 3.1 on a color turbo slab. I've never had > this happen. > > -- > Bob Cook bobcook@slac.stanford.edu > Stanford Linear Accelerator Center NeXT mail okay What module are you using? I've had freezes with teapot. Best avoid that one for general use. -- The gentle journey jars to stop. The drifting dream is done. The long gone goblins loom ahead; the deadly, that we thought were dead, stand waiting, every one. -- Walt Kelly Jim McKelvey mckelvey@fafnir.com
From: byron@ocf.nms.unt.edu (Byron Goodman) Newsgroups: comp.sys.next.bugs Subject: Re: AAAAHHH! (simple request) Date: 9 Jan 1994 08:45:38 GMT Organization: University of North Texas Message-ID: <2gog7i$34s@hermes.unt.edu> References: <1993Dec18.074900.13853@news.unt.edu> <2gi5lp$a5@agate.berkeley.edu> William E. Grosso (william@pinoko.berkeley.edu) wrote: : Byron Goodman writes : >[entire post deleted] : A simple request: Use descriptive subject titles. : "AAAAHHH" is only a good subject line in : alt.dentist.what.to.say : The subject field was put there so that we can all : save time/energy by avoiding articles we don't care : about. Please use it. : Bill Grosso Sounds like you have an attitude problem. -- +--------------------------------------+--------------------------------------+ | Byron Goodman | OS/2 and FreeBSD Programmer/Advocate | | byron@ocf.nms.unt.edu | Prefers not to use Microsoft..... | +--------------------------------------+--------------------------------------+
Newsgroups: comp.sys.next.bugs From: lloyd@world.std.com (Christopher Lloyd) Subject: Re: AAAAHHH! (simple request) Message-ID: <CJCv2E.L46@world.std.com> Organization: The World Public Access UNIX, Brookline, MA References: <1993Dec18.074900.13853@news.unt.edu> <2gi5lp$a5@agate.berkeley.edu> <2gog7i$34s@hermes.unt.edu> Date: Sun, 9 Jan 1994 09:15:49 GMT In article <2gog7i$34s@hermes.unt.edu> byron@ocf.nms.unt.edu (Byron Goodman) writes: >William E. Grosso (william@pinoko.berkeley.edu) wrote: >: Byron Goodman writes > >: >[entire post deleted] > >: A simple request: Use descriptive subject titles. >: "AAAAHHH" is only a good subject line in >: alt.dentist.what.to.say > >: The subject field was put there so that we can all >: save time/energy by avoiding articles we don't care >: about. Please use it. > >: Bill Grosso > >Sounds like you have an attitude problem. Hehehe. We ALL have an attitude problem! Or, as an old buddy used to say: "It's not a bad attitude, it's a *different* attitude" and now for something completely different, -- :: Christopher Lloyd :: Yrrid, Inc. :: lloyd@yrrid.com lloyd@world.std.com :: :: If I had known it was harmless, I would have killed it myself ::
Newsgroups: comp.sys.next.bugs,comp.sys.next.hardware From: josh@vegiwopr.calpoly.edu (Josh) Subject: Re: AudioCD and Apple CD150 Message-ID: <1994Jan09.180611.164966@zeus.aix.calpoly.edu> Sender: news@zeus.calpoly.edu Organization: Cal Poly State University, San Luis Obispo References: <1994Jan6.134912.7312@isolde.stgt.sub.org> Date: Sun, 09 Jan 1994 18:06:11 GMT In article <1994Jan6.134912.7312@isolde.stgt.sub.org>, Markus Stoll <markus@isolde.stgt.sub.org> wrote: >Karsten Wolf (g402@hrz.uni-giessen.de) claimed: >> Audio-CDs donot play on Intel-Box with Apple-CD 150 > >Well, they dont play with CDPlayer.app on any NeXT. > >L. Anathea Brooks (gaia@wam.umd.edu) wrote: >> Hmmm. I bought an Apple CD150 (on sale for $150) which >> *seems* to be a white NeXT CD-ROM drive. But CDPlayer They are both sony mechs? Sony didn't make nearly as many different mech's as they had models - a lot of them simply had a different eprom. I bought a used Sun cdrom for cheap and replaced the eprom and voila! Instant Sony 541 :) Works great. So any hardware types out there with access to a next cdrom feel like seeing what other sony mech's it's compatible with? >Apparently I'm not the only one with an Apple CD150, who wants to >play Audio-CDs. I wrote a very basic audio-CD-player for > [deleted promised program] >-- >Markus Stoll, markus@isolde.stgt.sub.org (NeXTMail preferred) > stoll@vaxph.mpi-stuttgart.mpg.de josh@vegiwopr.calpoly.edu --> nextmail ok
From: sanguish@digifix.com Newsgroups: comp.sys.next.announce,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.bugs,comp.sys.next.programmer,comp.sys.next.marketplace Subject: Quick Guide to NEXTSTEP information on the Internet Date: 9 Jan 1994 23:52:13 -0500 Organization: Next Announcements Message-ID: <2gqmtt$bap@digifix.digifix.com> This post is made weekly, to help 'point' users to more NEXTSTEP information Topics include: comp.sys.next newsgroups related newsgroups comp.sys.next newsgroups mailing list ftp sites NeXTanswers comp.sys.next.* newsgroups -------------------------- Comp.Sys.Next.Advocacy This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. Comp.Sys.Next.Announce Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) The NEXTSTEP FAQs are posted here monthly as well. This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Comp.Sys.Next.Bugs A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT- specific groups as well. Comp.Sys.Next.Hardware Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. Comp.Sys.Next.Marketplace NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. Comp.Sys.Next.Misc For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! Comp.Sys.Next.Programmer Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. The NEXTSTEP programmer FAQs are posted here. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. Comp.Sys.Next.Software This is a place to talk about [third party] software products that run on NEXTSTEP systems. Comp.Sys.Next.Sysadmin Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. related Newsgroups ------------------ Comp.Soft-Sys.Nextstep Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. Comp.Lang.Objective-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. Comp.Object Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original Comp.Sys.Next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News ------------------------------------------- Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep-d next-advocacy-d next-announce-d next-bugs-d next-hardware-d next-marketplace-d next-misc-d next-programmer-d next-software-d next-sysadmin-d (For a full description, send mail saying LISTS to <digestif@antigone.com>). The subscription syntax is essentially the same as LISTSERV's. To subscribe, send a message to <digestif@antigone.com> saying: SUB Listname YourName Example: SUB next-hardware-d John Doe The ftp sites ------------- cs.orst.edu: The main site for North American submissions nova.cc.purdue.edu: Lots of older stuff, but very short on disk space ftp.informatik.uni-muenchen.de: In Germany. ftp.next.com: See the below ftp.next.com and NextAnswers@next.com ------------------------------------- From the document 1000_Help from ftp.next.com Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, or you can transfer them by anonymous ftp. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to NeXTanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO <your-address> If you have any problem with the system or suggestions for improvement, please send mail to NeXTanswers-request@NeXT.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to NeXTanswers-request@NeXT.com. Written by: Eric P. Scott eps@toaster.SFSU.EDU and Scott Anguish sanguish@digifix.com Additions from: Greg Anderson (Greg_Anderson@afs.com) and Michael Pizolato (Michael_Pizolato@afs.com)
Newsgroups: comp.sys.next.bugs From: rfi@jeans.fokus.gmd.de (Robert Fischer) Subject: PB: Spaces in path not recommended? Message-ID: <1994Jan10.092117.18605@fokus.gmd.de> Sender: news@fokus.gmd.de (News system) Organization: GMD-Fokus Date: Mon, 10 Jan 1994 09:21:17 GMT If your application path does contain spaces (or other special characters?) the ProjectBuilder function "Finder" does not work properly: WM's console gives the hint: "cd: Too many arguments.". Robert. -- ------------------------------------------------------------------------ Robert Fischer @ GMD-Fokus -------- __o ------- _`\<,_ fischer@fokus.gmd.de ------- (*)/ (*) ## NeXT-Mail welcome ## ------------------------------------------------------------------------
Newsgroups: comp.sys.next.bugs,comp.sys.next.hardware From: g402@hrz.uni-giessen.de (Karsten Wolf) Subject: Re: AudioCD and Apple CD150 Sender: news@muster.hrz.uni-giessen.de Message-ID: <CJEuoB.9M4@muster.hrz.uni-giessen.de> Date: Mon, 10 Jan 1994 11:02:35 GMT References: <1994Jan6.134912.7312@isolde.stgt.sub.org> Organization: University of Giessen, Germany In article <1994Jan6.134912.7312@isolde.stgt.sub.org>, markus@isolde.stgt.sub.org (Markus Stoll) writes: |> Karsten Wolf (g402@hrz.uni-giessen.de) claimed: |> > Audio-CDs donīt play on Intel-Box with Apple-CD 150 |> |> Well, they dont play with CDPlayer.app on any NeXT. |> |> L. Anathea Brooks (gaia@wam.umd.edu) wrote: |> > Hmmm. I bought an Apple CD150 (on sale for $150) which |> > *seems* to be a white NeXT CD-ROM drive. But CDPlayer |> > does not work with it (it did w/ NeXT drive). |> |> Apparently I'm not the only one with an Apple CD150, who wants to |> play Audio-CDs. I wrote a very basic audio-CD-player for |> it. Unfortunately, I'm very busy right now. In february I will make a |> real CD-player out of it. If you guys cant wait and are willing to |> put up with this basic version (which allows start, stop, skipping |> tracks and shows actual track, index and time), i can place it on some |> ftp-servers. |> Yeah, put it on ftp, please. Karsten |> |> |> -- |> Markus Stoll, markus@isolde.stgt.sub.org (NeXTMail preferred) |> stoll@vaxph.mpi-stuttgart.mpg.de
From: rfschtkt%banruc60.bitnet@cc1.kuleuven.ac.be Newsgroups: comp.sys.next.bugs Subject: DPS client library errors, Mail.app Date: 10 Jan 1994 05:40:33 -0600 Organization: UTexas Mail-to-News Gateway Sender: daemon@cs.utexas.edu Message-ID: <9312301643.AA00234@flexus> LS, >>>>> KBNS.12.0.009_o3.2o When oh when is this ever going to end? On the console: ``Dec 27 13:29:19 flexus WM[202]: DPS client library error: Error while writing to connection, DPSContext b6cd0, data -110''. And the gruesome NXRunAlertPanel("Workspace","Internal error 1103 occurred.\nContinue (at your own risk)?","Continue","Log Out",NULL). How about user-friendly error messages? I think I had better log out now. KBNS.12.0.011_o3.2o Again like KBNS.12.0.009 This time on the console: ``Dec 29 17:09:43 flexus WM[271]: DPS client library error: Error while writing to connection, DPSContext b6cd0, data -102''. KBNS.12.0.012_o3.2o And now ProjectBuilder This time on the console: `` Dec 30 10:46:35 flexus ProjectBuilder[2808]: DPS client library error: Error while writing to connection, DPSContext 5ede0, data -110 Dec 30 10:46:35 flexus ProjectBuilder[2808]: Exiting due to Window Server death ''. And gone was ProjectBuilder.app (no core file was dumped). Let's try again (I was doing a search for a string in the project): `` Dec 30 10:57:00 flexus ProjectBuilder[3109]: DPS client library error: Error while writing to connection, DPSContext 5eca4, data -110 Dec 30 10:57:01 flexus ProjectBuilder[3109]: Exiting due to Window Server death Dec 30 10:57:56 flexus ProjectBuilder[3118]: DPS client library error: Error while writing to connection, DPSContext 5eca4, data -110 Dec 30 10:57:56 flexus ProjectBuilder[3118]: Exiting due to Window Server death Dec 30 10:59:00 flexus ProjectBuilder[3123]: DPS client library error: Error while writing to connection, DPSContext 5eca4, data -110 Dec 30 10:59:00 flexus ProjectBuilder[3123]: Exiting due to Window Server death ''. I searched for the same string two more times, and for another string the third time. Only ProjectBuilder.app was affected. I think I'm going to do something else than being messed with by NEXTSTEP, now. KBNS.12.0.013_c3.2o Really awful bug, but probably not (entirely) Mail.app's fault? I'll classify it here because I saw this bug in Mail.app, but it probably is caused by something else, as you will see. It is not possible anymore to transfer rich messages between Mailboxes: NXRunAlertPanel("Transfer","Unable to transfer %s",NULL,NULL,NULL,thePath), where thePath is something like ``Active.mbox/No_subject.attach''. I've tested this up to moving my Mailboxes folder somewhere else and having Mail.app create one from scratch, and restoring 3.1's Mail.app, to no avail. Maybe Mail.app has its current directory in ~/Mailboxes and some [Application workspace] protocol now assumes a full path? I don't know, but this is a damned nuisance. Should I now downgrade to 3.1's Workspace? If you have had the same experience, or if this is totally strange to you, do drop me a note. XXX Now I can't send messages to nobody anymore: they're returned instead of thrown away. Is this a bug? <<<<< If you have experienced either of these phenomena, please drop me a note. I'm running 3.2 on an 8-MB NeXTstation. For the rest, I wish you A Happy New Year! Raf Schietekat, RfSchtkt@banruc60.bitnet, Flanders, Belgium RfSchtkt%banruc60.bitnet@uunet.uu.net or so if the above does not work for you (real, i.e., with triangle in the Deliver button) NeXTmail preferred I can't reach sites with ! or % in their address, or ending in .at or .uucp
From: kay@nagasena.muc.de (Kay Schulz) Newsgroups: comp.sys.next.bugs Subject: Serial Drivers Date: 10 Jan 1994 19:46:03 +0100 Organization: Firestone Chaotic Group Message-ID: <2gs7pb$14q@nagasena.muc.de> Hi Netters I would like to know about your experiences with NSI 3.2 and the serial drivers. Are they fast? Do they work for you. Someone said that most of the guys are satisfied with them. I am not. Never. With a 14400 modem connection I get with DOS 1600 CPS via Zmodem. Stable and fast. With ISC, too! With NeXTStep 3.2 I get 300 cps if I am lucky. But most of the blocks are sent million times. I can't do file download or something similar. I get NEWS but I am afraid of my telephone bill. Mail, ok, they are small, but anythign else. in 3.1 the driver update was ok. the new drivers on 3.2 are really shit. Any UNIX is able to have good drivers, NSI 3.1, too why not 3.2 I use a 16550 UART. Don't flame me, I am a fan of NSI and will never change (except OpenStep or Gecko of HP with NS). So write to the group what your experiences are , if it's my interface which works bad. Thanks -- Kay Schulz kay@nagasena.muc.de NeXTmail accepted
From: jweiss@casbah.acns.nwu.edu (Jerry Weiss) Newsgroups: comp.sys.next.bugs Subject: Re: Serial Drivers Date: 10 Jan 1994 22:44:37 GMT Organization: Northwestern University, Evanston IL USA Message-ID: <2gslol$ga@news.acns.nwu.edu> References: <2gs7pb$14q@nagasena.muc.de> In article <2gs7pb$14q@nagasena.muc.de>, Kay Schulz <kay@nagasena.muc.de> wrote: > >Hi Netters > >I would like to know about your experiences with NSI 3.2 and the serial >drivers. Are they fast? Do they work for you. Someone said that most of the >guys are satisfied with them. I am not. Never. With a 14400 modem >connection I get with DOS 1600 CPS via Zmodem. Stable and fast. With ISC, >too! >With NeXTStep 3.2 I get 300 cps if I am lucky. But most of the >blocks are sent million times. I can't do file download or something >similar. I get NEWS but I am afraid of my telephone bill. Mail, ok, they >are small, but anythign else. >in 3.1 the driver update was ok. the newdrivers on 3.2 are really shit. >Any UNIX is able to have good drivers, NSI 3.1, too why not 3.2 >I use a 16550 UART. Well, I won't apologize for Next's Serial drivers, but you should be able to manage more than 300bps. The fact that you have to send blocks multiple times is not a good sign. I'm not familiar with Zmodem (I use mostly kermit and slip) myself. In any case NeXT serial speed will probably never equal raw Dos. Dos is just an overgrow program loader. The connections most programs make with the serial ports are direct or almost as good as direct. Ie: There is little or no overhead with DOS. NeXTstep on the other hand is much more sophisticated. With that comes more complexity, Each byte has to pass through a more layers than dos and this can slow things down, whether its more or less efficient than other Unix's is another story. With that said, I can say that I get good reliablity and good speed with the current driver. With a Slip connection on a Multitech V32 (and compression) modem at 9600 baud (serial line set to 19,200) I can get FTP transfers of over 1100 bytes/sec. I have a Gateway 486dx2-66 which doesn't have a discrete UART, its emulated in firmware. I'll be trying a faster modem soon. What speed is your serial line set to (38400,19200)? Have you tried other speeds? Are you using hardware or software handshaking? What packet sizes are you using? -- Jerry S. Weiss j-weiss@nwu.edu Dept. Medicine, Northwestern Univ. Medical School, Chicago, Illinois %SYSTEM-S-PHALOKTARG, Phasers Locked on Target, Ready to Fire
From: william@pinoko.berkeley.edu (William E. Grosso) Newsgroups: comp.sys.next.bugs Subject: Attitude Problem ? Date: 11 Jan 1994 00:45:13 GMT Organization: University of California, Berkeley Message-ID: <2gssqp$216@agate.berkeley.edu> References: <2gog7i$34s@hermes.unt.edu> Byron Goodman writes > Sounds like you have an attitude problem. I don't think so. This group is, primarily, for reporting bugs. I need to know about *some* of these bugs. Some of them are inconsequential (to me). For example, I do not currently use DBKit. This is unlikely to change in the near future. So, I ignore bug reports dealing with DBKit (by the time I need to know about DBKit, current bugs will have either been fixed or subsumed into the bug FAQ [currently on 11.3 and highly recommended]). Nor am I the only programmer who only uses some of the classes NeXT provides. I would venture to guess that most NeXT programmers don't use all the classes (and can, therefore, skip many of the bug reports). However, I have to read all the ones labelled "AAAAHHH!" because I have no way of knowing, in advance, what they are referring to. This is a waste of my time (I do not enjoy reading bug reports). Etcetera. The key point is this: it is *not* an attitude problem to rquest that other people behave politely and follow common courtesy. Bill Grosso
Newsgroups: comp.sys.next.bugs From: dennis@nebulus.ampr.ab.ca (Dennis S. Breckenridge) Subject: Re: Serial Drivers References: <2gs7pb$14q@nagasena.muc.de> Organization: "Alchemy Mindworks" Date: Tue, 11 Jan 1994 06:25:05 GMT Message-ID: <CJGCHv.26t@nebulus.ampr.ab.ca> kay@nagasena.muc.de (Kay Schulz) writes: >Hi Netters >I would like to know about your experiences with NSI 3.2 and the serial drivers. Are they fast? Do they work for you. Someone said that most of the guys are satisfied with them. I am not. Never. With a 14400 modem connection I get with DOS 1600 CPS via Zmodem. Stable and fast. With ISC, too! >With NeXTStep 3.2 I get 300 cps if I am lucky. But most of the blocks are sent >million times. I can't do file download or something similar. I get NEWS but I am afraid of my telephone bill. Mail, ok, they are small, but anythign else. >in 3.1 the driver update was ok. the new drivers on 3.2 are really shit. >Any UNIX is able to have good drivers, NSI 3.1, too why not 3.2 >I use a 16550 UART. >Don't flame me, I am a fan of NSI and will never change (except OpenStep or Gecko of HP with NS). >So write to the group what your experiences are , if it's my interface which works bad. >Thanks I have just installed Mark Salyzyn's new and improved serial port driver on my machine. Initially it instantly increased my average uucp speed from 800 chr/sec to >1200 chrs/sec. I understand that it still has a few annoying problems but he assures me he will get them fixed, after all it's only the alpha release. I would like to publically congratulate Mark on the fine job he did. He has done something that the rest of the net has been unable to fix yet. Good work Mark! -- ----------------------------------------------------------------------------- Dennis Breckenridge Microsoft Windoze. Genetic throwback from the dennis@nebulus.ampr.ab.ca natural selection of UNIX. -----------------------------------------------------------------------------
From: lemson@ux1.cso.uiuc.edu (David Lemson) Newsgroups: comp.sys.next.bugs,comp.sys.next.sysadmin Subject: hostname messed up in 3.2 during boot sequence Date: 12 Jan 1994 00:19:55 GMT Organization: University of Illinois at Urbana Message-ID: <2gvfnb$60b@vixen.cso.uiuc.edu> I am upgrading many machines to 3.2 and it appears that the lines: if [ "$SETHOSTNAME" = -YES- ]; then if [ -z "$HOSTNAME" ]; then HOSTNAME=-AUTOMATIC- fi if [ "$HOSTNAME" != -NO- ]; then echo "Setting hostname to $HOSTNAME" docmd hostname $HOSTNAME fi fi aren't doing their job in /etc/rc.net. I see "Setting hostname to -AUTOMATIC-" as I should, but for some reason, 'docmd hostname -AUTOMATIC-' isn't setting the hostname as it should at that time (it stays 'localhost' as set in rc.boot I guess) If I run 'hostname -AUTOMATIC-' from the command line as root, it sets the hostname as it should (BOOTP). Does this work during the boot process for anyone else? I am considering putting a "hostname -AUTOMATIC-" in the /etc/rc.local as a workaround. This is far from optimal, as syslogd reports 'localhost' as the hostname until it gets around to the /etc/rc.local at the end. Does anyone have any clue what needs to run before hostname responds correctly by sending a BOOTP request? The BOOTP servers are running 3.2 as well. -- David Lemson (217) 244-1205 University of Illinois Computing & Comm Services Office System Administrator Internet : lemson@uiuc.edu UUCP :...!uiucuxc!uiucux1!lemson NeXTMail & MIME accepted BITNET : LEMSON@UIUCVMD -- David Lemson (217) 244-1205 University of Illinois Computing & Comm Services Office System Administrator Internet : lemson@uiuc.edu UUCP :...!uiucuxc!uiucux1!lemson NeXTMail & MIME accepted BITNET : LEMSON@UIUCVMD
From: wwille.hanse.de!wwille (Winfried Wille) Newsgroups: comp.sys.next.bugs Subject: Re: AudioCD and Apple CD150 Date: 11 Jan 1994 09:17:20 +0100 Organization: Terraprolls Sender: wwille@wwille.hanse.de Message-ID: <2gtnag$5vl@wwille.hanse.de> References: <1994Jan09.180611.164966@zeus.aix.calpoly.edu> Keywords: CD In article <1994Jan09.180611.164966@zeus.aix.calpoly.edu> josh@vegiwopr.calpoly.edu (Josh) writes: > I bought a used Sun cdrom for cheap and replaced the eprom and > voila! Instant Sony 541 :) Works great. To avoid needless actions. I put the eproms of a NeXT-CD in a Apple CD-150: audio works but data access sucks. Quite surprising since I inspected the drives in a Apple CD-150 and a NeXT CD and couldn't find a difference. Its said that one chip differs. But you are right the Sun cdrom will work fine with the NeXT firmware and vice versa. Greetings Winfried Wille -- * email: wwille@wwille.hanse.de, NeXT mail accepted * * Voice: +49 40/456543, Data +49 40/457826 * * HMH e.V. section NeXTfriends *
From: elric@mik.uky.edu (chris dotson) Newsgroups: comp.sys.next.bugs Subject: Re: Serial Drivers Date: 12 Jan 1994 16:45:24 GMT Organization: University of Kentucky Message-ID: <2h19f4$kou@s.ms.uky.edu> References: <2gs7pb$14q@nagasena.muc.de> <2gslol$ga@news.acns.nwu.edu> jweiss@casbah.acns.nwu.edu (Jerry Weiss) writes: >Kay Schulz <kay@nagasena.muc.de> wrote: [...] >>With NeXTStep 3.2 I get 300 cps if I am lucky. But most of the [...] > >Well, I won't apologize for Next's Serial drivers, but you should be able >to manage more than 300bps. The fact that you have to send blocks multiple [...] 300 cps (chars per second) would be about 2400 baud theoretically, and actually about 3000 baud (240 cps is about the best _I_ get with 2400 baud). As far as Bits Per Second, I've always been told this is different than baud but I don't see how. Anyone know? E-mail me, please... Sorry, this is DEFINITELY off the subject... >%SYSTEM-S-PHALOKTARG, Phasers Locked on Target, Ready to Fire I wouldn't classify that as a Serious Error... How about: %SYSTEM-W-PHALOKTARG...? :) That's great... -- Elric ___________________ Chris Dotson - elric@mik.uky.edu /NeXTMail welcome. / /________-=-_______/ "If it ain't broke - mess with it until it is. Then fix it." -Me
From: kay@nagasena.muc.de (Kay Schulz) Newsgroups: comp.sys.next.bugs Subject: Re: Serial Drivers Date: 12 Jan 1994 19:39:25 +0100 Organization: Firestone Chaotic Group Message-ID: <2h1g4t$gp@nagasena.muc.de> References: <2gs7pb$14q@nagasena.muc.de> <2gslol$ga@news.acns.nwu.edu> : Well, I won't apologize for Next's Serial drivers, but you should be able : to manage more than 300bps. The fact that you have to send blocks multiple I should be. : times is not a good sign. I'm not familiar with Zmodem (I use mostly : kermit and slip) myself. In any case NeXT serial speed will probably never : equal raw Dos. Dos is just an overgrow program loader. The connections : most programs make with the serial ports are direct or almost as good as The thing is I get 1600 cps with SCO and ISC, all UNIXES and with LINUX. So where is the difference in it. : direct. Ie: There is little or no overhead with DOS. NeXTstep on the : other hand is much more sophisticated. With that comes more complexity, : Each byte has to pass through a more layers than dos and this can : slow things down, whether its more or less efficient than other Unix's is : another story. But why are others UNIXes faster? : With that said, I can say that I get good reliablity and good speed with : the current driver. With a Slip connection on a Multitech V32 (and : compression) modem at 9600 baud (serial line set to 19,200) I can get FTP : transfers of over 1100 bytes/sec. I have a Gateway 486dx2-66 which : doesn't have a discrete UART, its emulated in firmware. I'll be : trying a faster modem soon. : What speed is your serial line set to (38400,19200)? Have you tried 19200 : other speeds? Are you using hardware or software handshaking? no Hardware Handshaking : What packet sizes are you using? don't know, how can I find out -- Kay Schulz kay@nagasena.muc.de NeXTmail accepted
From: Charles William Swiger <infidel+@CMU.EDU> Newsgroups: comp.sys.next.bugs Subject: Re: Serial Drivers Date: Wed, 12 Jan 1994 14:02:19 -0500 Organization: Fifth yr. senior, Chemistry, Carnegie Mellon, Pittsburgh, PA Message-ID: <8hB4Yv600iV4E6LZku@andrew.cmu.edu> In-Reply-To: <2h19f4$kou@s.ms.uky.edu> Excerpts from netnews.comp.sys.next.bugs: 12-Jan-94 Re: Serial Drivers by chris dotson@mik.uky.edu > 300 cps (chars per second) would be about 2400 baud theoretically, and > actually about 3000 baud (240 cps is about the best _I_ get with 2400 > baud). Remember the stop bits. For a standard 8-N-1 (8 data bits, no parity, 1 stop bit), the maximum effective transfer rate from a 2400 baud modem is 267 cps. 240 cps is about as good as you can expect to get in practical terms. > As far as Bits Per Second, I've always been told this is different than > baud but I don't see how. Anyone know? E-mail me, please... Sorry, this > is DEFINITELY off the subject... Most people use the terms interchangably, but baud refers to the number of pulses per second (Hz), with the potential of having more than one bit encoded per pulse (by having move than 2 tone levels that the signal changes between). BPS refers to the actual amount of information being transferred. For example, the standard phone line has a limited bandwidth of under 8KHz, which means that you can't possibly put a signal through faster than half that frequency (from sampling theory). In practice, a 2400 baud signal is about as fast as you can throw data through a phone link. In order to get 9600KBps (or faster modems), they have to encode multiple bits per pulse, but they only actually transmit at a rate of 2400 baud. This is why you'll find things like 2400baud modems that can do 9600 Kbps FAX transfers (but have a high enough error rate that they can't be used as a 9600 Kbps modem). -Chuck Charles William Swiger -- CMU...*splat*! | 1. You can't fly. --------------------------------------------+ 2. Cars are always real, even AMS & normal mail: infidel@cmu.edu | when they're not. Failing that: cs4w+@andrew.cmu.edu | 3. Police are not your friends. NeXTmail: chuck@cswiger.slip.andrew.cmu.edu | 4. Fire burns.
Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs From: chris@its.com (Chris Cuilla) Subject: Matrix subclassing... Message-ID: <1994Jan12.213421.15585@il.us.swissbank.com> Sender: root@il.us.swissbank.com (Operator) Organization: Swiss Bank Corporation CM&T Division Date: Wed, 12 Jan 1994 21:34:21 GMT I'm playing around with a Matrix subclass...whenever I get a mouseDown: and print out the results of "getRow:andCol:forPoint:" I seem to be getting bizarre/inconsistent results. Is this broken. Anyone have any experience with this? -- Chris Cuilla --- NEXTSTEP Developer chris@its.com --------------------------------------------------------------------- "There's no such thing as a 'part-time' perfectionist."
From: Charles William Swiger <infidel+@CMU.EDU> Newsgroups: comp.sys.next.bugs Subject: Re: Serial Drivers Date: Wed, 12 Jan 1994 17:07:56 -0500 Organization: Fifth yr. senior, Chemistry, Carnegie Mellon, Pittsburgh, PA Message-ID: <chB7Gwy00iV4E_07Ew@andrew.cmu.edu> In-Reply-To: <8hB4Yv600iV4E6LZku@andrew.cmu.edu> Excerpts from netnews.comp.sys.next.bugs: 12-Jan-94 Re: Serial Drivers by Charles W. Swiger@CMU.EDU > Remember the stop bits. For a standard 8-N-1 (8 data bits, no parity, 1 > stop bit), the maximum effective transfer rate from a 2400 baud modem is > 267 cps. 240 cps is about as good as you can expect to get in practical > terms. Sigh. Remember the start bits too, Chuck, as someone gently pointed out to me. Repeat calculation with 10 bits per frame (1 start, 8 data, 1 stop), to get a maximum transfer rate of 240 cps. Subtract a few bytes for overhead, and that's your maximum transfer rate. (Replying to your own posts is a little tacky, but oh well.... ;-) -Chuck Charles William Swiger -- CMU...*splat*! | 1. You can't fly. --------------------------------------------+ 2. Cars are always real, even AMS & normal mail: infidel@cmu.edu | when they're not. Failing that: cs4w+@andrew.cmu.edu | 3. Police are not your friends. NeXTmail: chuck@cswiger.slip.andrew.cmu.edu | 4. Fire burns.
From: haker@whitewater.chem.wisc.edu (M Haker) Newsgroups: comp.sys.next.bugs Subject: DriverKit Bug: can't handle multiple DMA channels Date: 13 Jan 1994 01:04:17 GMT Organization: Division of Information Technology Message-ID: <2h26mh$2o6@news.doit.wisc.edu> Hello All, I'm having problems getting my A/D driver working with DMA. Depending on how many samples the board will do per trigger, I might need to configure my DMA buffer to be up to 200kb which seems to cause the machine to crash. So I tried breaking it up using a smaller buffer and having the board generate terminal-count interrupts where I would 1) disable the channel, 2) quickly copy the data out of the DMA buffer, 3) reenable the channel before the FIFO on the board overruns. That didn't work either. Then I found out that this board can use two DMA channels in an alternating fashion. So you can copy the data out of one while the other channel is involved in data collection, and switch back and forth until the board is done sampling. Unfortunately IOConfigTable and IOEISADeviceDescription don't want to reserve more than one DMA Channel even though they are documented as having the capability. Also Configure.app seems to be setup to deal with handling multiple DMA channels. It can't switch between one and two channels, but if the default is two channels, it allows you to pick which two to use. My Default.table has the following entries "DMA Channels" = "1 3"; "Valid DMA Channels" = "0 1 2 3 5 6 7"; and on my system channels 1 and 3 are available. But when the driver is loaded and initialized (initFromDeviceDescription) it reports that it only has one DMA channel using [deviceDescription numChannels]. And even if I go up to the IOConfigTable and query it with [[deviceDescription configTable] valueForStringKey:"DMA Channels"] it returns "1", not "1 3" I've tried reserving the channel by hand, but that fails too. [self reserveChannel:3]; Has anyone else noticed this bug, and more importantly has anyone figured out a way around it. Thanks for any Help (please respond by email). Michael Haker Dept. Chemistry UW-Madison haker@whitewater.chem.wisc.edu
From: elric@mik.uky.edu (chris dotson) Newsgroups: comp.sys.next.bugs Subject: Re: Serial Drivers Date: 13 Jan 1994 02:06:14 GMT Organization: University of Kentucky Message-ID: <2h2aam$m2c@s.ms.uky.edu> References: <2gs7pb$14q@nagasena.muc.de> <2gslol$ga@news.acns.nwu.edu> <2h19f4$kou@s.ms.uky.edu> In article <2h19f4$kou@s.ms.uky.edu> elric@mik.uky.edu (chris dotson) writes: >300 cps (chars per second) would be about 2400 baud theoretically, and actually >about 3000 baud (240 cps is about the best _I_ get with 2400 baud). If Chuck can reply to his own posts, I can too. I reversed 'theoretically' and 'actually', as someone kindly pointed out. (Thanks Mike) -- Elric ___________________ Chris Dotson - elric@mik.uky.edu /NeXTMail welcome. / /________-=-_______/ "If it ain't broke - mess with it until it is. Then fix it." -Me
From: mow@marsu.tynet.sub.org (Markus Wenzel) Newsgroups: comp.sys.next.hardware,comp.sys.next.bugs Subject: Re: Toshiba 4400C Mouse/Keyboard dead in 3.2 Date: 12 Jan 1994 10:29:25 +0100 Organization: Palumbian Research Labs Message-ID: <2h0ftl$9e@marsu.tynet.sub.org> References: <2glktp$l5u@umd5.umd.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit rogata@is-next.umd.edu (Richard Scott Ogata) writes: > OK, I am at my wit's end. I have a Toshiba 4400C which was running >3.1 just fine. I decided to upgrade to 3.2, and, as others upgrading this >laptop to 3.2 have found, the keyboard croaked. Furthermore, my PS/2 mouse >won't work either, unless I plug it into the serial port and use it as a >serial mouse. > I edited the kernel according to the instructions I found in this >newsgroup, but there is no change in behavior. > I can't find anything about this in NeXTanswers. Seeing how this is >a bit of a "show-stopper", and that I believe there are other models affected, >I don't understand this. Has anyone contacted NeXT about this? Or has >anyone found an effective patch? NeXTanswers document #1480 says something about PSKeyboard/PS2Mouse interferences and gives a workaround to avoid my posted kernel patch. I cannot say if it may cure your problem, but you should read it anyway. Regards, Markus. -- Marsu: "Es gibt tatsaechlich Leute, die ohne Computer gluecklich und zufrieden leben." -- Frankie: "Ach was, die emulieren das doch nur!" ----- Markus Wenzel System administration, Consulting, Networking mow@marsu.tynet.sub.org on... NeXTSTEP / Unix / Novell / Windows NT IRC: Marsu Intel aside.
From: jweiss@casbah.acns.nwu.edu (Jerry Weiss) Newsgroups: comp.sys.next.bugs Subject: Re: Serial Drivers Date: 13 Jan 1994 08:53:15 GMT Organization: Northwestern University, Evanston IL USA Message-ID: <2h325r$6ri@news.acns.nwu.edu> References: <2gs7pb$14q@nagasena.muc.de> <2gslol$ga@news.acns.nwu.edu> <2h1g4t$gp@nagasena.muc.de> In article <2h1g4t$gp@nagasena.muc.de>, Kay Schulz <kay@nagasena.muc.de> wrote: > > >: times is not a good sign. I'm not familiar with Zmodem (I use mostly >: kermit and slip) myself. In any case NeXT serial speed will probably never >: equal raw Dos. Dos is just an overgrow program loader. The connections >: most programs make with the serial ports are direct or almost as good as >The thing is I get 1600 cps with SCO and ISC, all UNIXES and with LINUX. >So where is the difference in it. >But why are others UNIXes faster? I don't know. Good Serial I/O is hard in any O/S. Next with its Driver kit is trying to encapsulate parts of driver interfaces to make writing new drivers much easier. Its a very worthy goal, but it may add some overhead in the process. In any case, it shouldn't be that slow on your machine especially with a real 16550 chip (unless of course it was one of the early versions which were real buggy). > > >: What speed is your serial line set to (38400,19200)? Have you tried >19200 > >: other speeds? Are you using hardware or software handshaking? >no >Hardware Handshaking > >: What packet sizes are you using? >don't know, how can I find out > One question I forgot (or can't remember from your original posting). Where you having this problem Sending or Recieving from the Next? > Well, given the other info I would lock serial ports at 9600 and have the modem to baud adjust. If you still get less than say 300 bytes/sec I'd say you have a configuration error such as incorrect cabling or some modem setting. I know you are probably going to say "but it works with XXX why should I suspect the cabling?" My reply would be that while there are (some) standards for how RS232 signaling is to performed not all interfaces and computers or their software really play by the same rules. Why else to you think there are all these darn configuration options of signal lines on these fancy modems. Now if you drop the port speed to 9600 and the tranfer rate climbs back up, my intuition would be that NeXTstep is dropping the ball at higher speeds. You might want to check the serial driver someone just posted to the net the other day. If you can find the packet size for Zmdoem and have any control over it, you might want to drop it down to something small. Ie: Give the NeXT a brief breather between packets and it might be able to keep up. This is especially true if the packets are large say >1000 bytes. ***** on Soapbox ***** The serial I/O on PC's is quite a joke to begin with. Little or no buffering (15 bytes on a 16550 isn't a hell of alot, especially at 38.4K), coupled with a limited interrupt and I/O port scheme. Anything more than 2 serial ports on a machine is risky business. Any O/S + CPU forced to do a context switch for every I/O byte is going to suffer in interactive performance. If any of these other O/S's have better serial handling, its because they've probably gone through several generations of drivers when the average modems were less demanding. But for the moment that's the world Intel and IBM made us live in. I see some of the newer modems are starting to use parallel ports for I/O to escape mire of PC serial ports. Maybe there's some hope a new standard will emerge. ((((( off Soapbox ))))) -- Jerry S. Weiss j-weiss@nwu.edu Dept. Medicine, Northwestern Univ. Medical School, Chicago, Illinois %SYSTEM-S-PHALOKTARG, Phasers Locked on Target, Ready to Fire
From: M.Crawford@dcs.shef.ac.uk (Malcolm Crawford) Newsgroups: comp.sys.next.bugs Subject: PB bug -- .c file compiled in preference to .m Date: 13 Jan 1994 07:32:03 -0600 Organization: UTexas Mail-to-News Gateway Sender: daemon@cs.utexas.edu Message-ID: <9401131328.AA12536@dcs.shef.ac.uk> A relatively inexperienced NEXTSTEP-developing colleague here recently tore out most of his hair trying to track down why his application had suddenly stopped compiling properly. Builder seemed to progress correctly, but then reported: --- cc -g -O -Wall -I./sym -arch m68k -I. -c DTController.m -o ./m68k_obj/DTController.o cc -g -O -Wall -I./sym -arch m68k -I. -c DTFilter.m -o ./m68k_obj/DTFilter.o cc -g -O -Wall -I./sym -arch m68k -I. -c DTGraph.m -o ./m68k_obj/DTGraph.o cc -g -O -Wall -I./sym -arch m68k -I. -c DTInChan.c -o ./m68k_obj/DTInChan.o cc -g -O -Wall -I./sym -arch m68k -ObjC -sectcreate __ICON __header DataTool.iconheader -segprot __ICON r r -sectcreate __ICON app /usr/lib/NextStep/Workspace.app/application.tiff -o DataTool.app/DataTool m68k_obj/Connection.o m68k_obj/DisplayableObject.o m68k_obj/DisplayList.o m68k_obj/DTController.o m68k_obj/DTFilter.o m68k_obj/DTGraph.o m68k_obj/DTInChan.o m68k_obj/DTInFile.o m68k_obj/DTOutChan.o m68k_obj/FilterWindow.o m68k_obj/GraphView.o m68k_obj/DataTool_main.o -lMedia_s -lNeXT_s ld: Undefined symbols: .objc_class_name_DTInChan *** Exit 1 Stop. *** Exit 1 Stop. --- Spot the mistake? I'll leave a couple of lines' space in case you want to test yourself -- no prizes though! For some reason Jon ("the folically-challenged", as we now know him!) had created a file called DTInChan.c, which was actually empty: Builder nevertheless picked up the file and pretended to compile it in place of DTInChan.m. Is this a bug? Have fun, mmalcolm.
Newsgroups: comp.sys.next.bugs From: kwang@data.acs.calpoly.edu (Kevin Wang (The Scarecrow)) Subject: chgrp is broken! Message-ID: <1994Jan13.200834.171773@zeus.aix.calpoly.edu> Sender: news@zeus.calpoly.edu Organization: The Outland Riders Date: Thu, 13 Jan 1994 20:08:34 GMT "chgrp wheel booga" and "chown kwang.wheel foo" is broken. It claims that "chgrp: You are not a member of the wheel group" when /usr/ucb/groups tells me that I am. nidump also confirms this. Here is my configuration: A lab of 20 NeXTs setup into a 2-tier NetInfo domain. Here are the symptoms: If I were to setup a user with primary group "student" and secondary group "project", where both groups are in the netinfo / /groups/* domain. "/usr/ucb/groups" and "nidump group /" both confirm that I am in the secondary group as shown here: >groups student staff wheel >nidump group / | egrep "wheel|staff|student" wheel:*:0:kwang,peggy,alex,kathleen,brian,kelly staff:*:200:rclover,awest,jnadel,alex,kwang,brian,peggy student:*:400:test,kelly,kathleen where "student", gid 400 is my primary group: kwang:passwd:12272:400:Kevin John Wang:/lore/home/ijkl/kwang:/bin/tcsh For example, here I create a directory, only accessable to wheel group, and I can certainly still access its files: data:/usr/local/etc>su Password: 12:00pm [ROOT] later> mkdir test 12:00pm [ROOT] later> chmod 770 test 12:00pm [ROOT] later> chown root.wheel test 12:00pm [ROOT] later> exit data:/usr/local/etc>ls -al test total 2 drwxrwx--- 2 root 1024 Jan 13 12:00 ./ drwxr-xr-x 21 root 1024 Jan 13 12:00 ../ data:/usr/local/etc>cd test data:/usr/local/etc/test>touch blah data:/usr/local/etc/test>ls -al total 2 drwxrwx--- 2 root 1024 Jan 13 12:01 ./ drwxr-xr-x 21 root 1024 Jan 13 12:00 ../ -rw------- 1 kwang 0 Jan 13 12:01 blah data:/usr/local/etc/test> I can also chgrp any file from "any_random_group" to "student" without a problem. However...anytime I try and change the group of a file (or directory for that matter) to either of my secondary gruops (staff or wheel), I get the error: chgrp: You are not a member of the wheel group This is a problem. One nasty solution: If I were to add myself to "wheel" in the /data group (the local nidb) then the chgrp's work. This implies to me that the NetInfo look-to-the-root-domain function isn't working for chgrp. Am I wrong? /usr/ucb/groups seems to search to root just fine, since it does say I'm in those groups, and I can access "group" files as well... - Kevin Wang
From: pkron@corona.com (Peter Kron) Organization: Corona Design, Inc., Seattle, WA Distribution: world Date: Thu, 13 Jan 1994 06:49:03 PST Message-ID: <1994Jan13.144903.346@corona.com> Newsgroups: comp.sys.next.bugs Subject: Re: Serial Drivers References: <chB7Gwy00iV4E_07Ew@andrew.cmu.edu> From: Charles William Swiger <infidel+@CMU.EDU> > Sigh. Remember the start bits too, Chuck, as someone > gently pointed out to me. Repeat calculation with 10 > bits per frame (1 start, 8 data, 1 stop), to get a maximum > transfer rate of 240 cps. Subtract a few bytes for > overhead, and that's your maximum transfer rate. This true for asynch communications, such as from the serial port to your external modem and over direct modem links. My understanding is that the LAPM and MNP protocols, however, actually use a synchronous connection over the phone line, which does not require start/stop bits. So uncompressed data maximum is back up to 300cps at 2400baud, 1200cps at 9600baud. The problems described by Kay Schulz sound like retransmissions due to errors, which drop the *effective* rate dramatically due to the delay before the error is recognized and subsequent retransmission. I've seen conflicting opinions as to whether modem compression and error-checking actually degrade UUCP performance, since they interfere with UUCP's own (primitive) error protocol. I run mine at 4800, compressed but not error-corrected, which gets me about 300cps for UUCP. That's the best I've ever gotten out of my gateway. Any communication port is susceptible to lost data if one side gets behind. This isn't really a bug, but an annoying inefficiency. I have a suspicion, though, that NEXTSTEP/Intel simply isn't very good at keeping up, particularly in the presence of disk activity. I notice retransmissions go up whenever I am actively hitting the disk. Perhaps Kay is transmitting in the presence of lots of other activity. It also corresponds with the erratic parallel port behavior I've seen on 3.2. If I expand a .compressed file during a print, I am almost certain to lose data. If I print on an idle system, things are ok for a while. Even idle, longer datastreams will fail, and one repeatable failure I found occurred right after an 8K boundary in the print stream. Unfortunately, there is no error correction to the printer (HP in graphics mode), so the result is garbled pages. In my case, the problem may be worsened by running both IDE and SCSI over an ISA bus. There are potentially two transfers occurring, either of which may require a subsequent copy since the ISA bus can't address all 28Mb memory. This is largely speculation on my part, but what else is there to go on? NeXT, as usual, is silent on the matter. --- NeXTMail:Peter_Kron@corona.com Corona Design, Inc. P.O. Box 51022 Seattle, WA 98115-1022
From: Mark Crispin <mrc@Tomobiki-Cho.CAC.Washington.EDU> Newsgroups: comp.sys.next.bugs Subject: lovely 3.2 Date: Thu, 13 Jan 1994 21:30:23 -0800 Organization: University of Washington Message-ID: <MS-C.758525423.1103527590.mrc@Tomobiki-Cho.CAC.Washington.EDU> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII Well well well. NeXT seems to have outdone itself with 3.2. After trashing my filesystem -- a full day of fsck'ing and restoring to put it back together -- 3.2 seems to take a special delight in panicing. Of course, this all happens inside the teensy tiny window which scrolls past a mile a minute then automatically reboots. Other than it's an ``invalid descriptor during MMU table walk'', I haven't recorded much else. Not that I've ever been able to figure when it auto-reboots and when it does not. It seems that with some panics, it syncs the filesystem and kills all processes, whereas with others, it ends up at the panic> prompt and waits for you to swat it. 3.2 seems to be crashing even more often than Windows or the Mac OS (and *that's* saying something). Ah well, should get used to it. It may be the last version of NEXTSTEP.
From: mycroft@colourbox.utexas.edu (Alex Currier) Newsgroups: comp.sys.next.bugs Subject: Re: lovely 3.2 (jumping to conclusions?) Date: 14 Jan 1994 06:16:07 GMT Organization: The University of Texas at Austin, Austin, Texas Message-ID: <2h5db7$c79@geraldo.cc.utexas.edu> References: <MS-C.758525423.1103527590.mrc@Tomobiki-Cho.CAC.Washington.EDU> In article <MS-C.758525423.1103527590.mrc@Tomobiki-Cho.CAC.Washington.EDU> Mark Crispin <mrc@Tomobiki-Cho.CAC.Washington.EDU> writes: > NeXT seems to have outdone itself with 3.2. > > After trashing my filesystem -- a full day of fsck'ing and restoring to put > it back together -- 3.2 seems to take a special delight in panicing. Of > course, this all happens inside the teensy tiny window which scrolls past a > mile a minute then automatically reboots. Other than it's an ``invalid > descriptor during MMU table walk'', I haven't recorded much else. I think you're being a bit hasty blaming your crashing and panic problems on 3.2. I've had it installed as long as you have and I haven't had any file system troubles, no crashes, no panics (and this is with ND and optical disk, probably a bit more fragile than your average NeXT)... what have you been doing to your machine to cause such excitement? Mine doesn't exactly sit idle so I would imagine that either you seriously wear out your NeXT or that it just doesn't like you. Maybe you should provide us with some situational, user specific information regarding your 3.2 problems rather than just dumping the whole load on the OS. Could it be you're using some incompatible software? Have you tried stripping your rc.local file and removing those older fax drivers (or whatever customizations you've made) or is it just easier to lay the blame on NeXT as usual? In fact, I'm quite happy with 3.2, seems a bit snappier, seems to utilize memory better, hasn't blown up on me. -- ============================================================================== Alex Currier * mycroft@ccwf.cc.utexas.edu * NeXTmail capable. Time is just one damn thing after another. ==============================================================================
Newsgroups: comp.sys.next.bugs From: clloyd@gleap (Charles C. Lloyd) Subject: Re: lovely 3.2 Message-ID: <1994Jan14.063337.386@gleap.sccsi.com> Sender: clloyd@gleap.sccsi.com Organization: GiantLeap Software References: <MS-C.758525423.1103527590.mrc@Tomobiki-Cho.CAC.Washington.EDU> Date: Fri, 14 Jan 1994 06:33:37 GMT Mark Crispin <mrc@Tomobiki-Cho.CAC.Washington.EDU> writes > >3.2 seems to be crashing even more often than Windows or the Mac OS (and >*that's* saying something). > I'd have to guess that you have some sort of flakiness in your hardware. I have been running 3.2 several days with no problems (which is no big deal), but, in the past couple of months, I have not heard of any other complaints such as yours. I know then 3.0 came out, there were a few machines at my place of employment that kept having spurious interrupts (or something like that). It was definately hardware related since we were all doing the same things and only 1 or 2 machines out of 100 machines had this problem. Charles. --- Charles Lloyd clloyd@GLeap.sccsi.com GiantLeap Software (713) 292-5853 or 363-9001 (Hou) (713) 363-9763 (fax) -- Charles Lloyd clloyd@GLeap.sccsi.com GiantLeap Software (713) 292-5853 or 363-9001 (Hou) (713) 363-9763 (fax)
From: Mark Crispin <mrc@Tomobiki-Cho.CAC.Washington.EDU> Newsgroups: comp.sys.next.bugs Subject: Re: lovely 3.2 Date: Fri, 14 Jan 1994 00:54:17 -0800 Organization: University of Washington Message-ID: <MS-C.758537657.1103527590.mrc@Tomobiki-Cho.CAC.Washington.EDU> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII In-Reply-To: <1994Jan14.063337.386@gleap.sccsi.com> The machine in question is a 16MB cube, 68040 25MHz CPU. It runs Transarc SLIP and co-Xist. The peripherals on it are a AppleCD 300 CD-ROM and a Telebit T3000 modem. It's on Ethernet. The crash which piddled on the filesystem, and the page fault crashes, both happen while the system is being started. The filesystem crash seemed to be ticked by my hitting the brightness key on the keyboard to try and dim the screen (for some reason, 3.2 seems to want to set the screen to full brightness). I'm avoiding the problem by keeping paws off the keyboard until it finishes startup. I've also seen the problem where windows and menus can't be moved; you try to drag their title bar and find its locked. True, it could be hardware. But it's curious that these problems started on the day that I installed 3.2. Before, that NeXT only paniced about once or twice a month, same as my other two machines (which are on 3.0 more or less permanently). Stomping on a few dozen low-numbered inodes is *not* a good way to start a relationship with a new operating system version!
Newsgroups: comp.sys.next.bugs From: dsdecasp@iiic.ethz.ch (Daniel Stefan Decasper) Subject: Re: lovely 3.2 (jumping to conclusions?) Message-ID: <1994Jan14.092421.6371@neptune.inf.ethz.ch> Sender: news@neptune.inf.ethz.ch (Mr News) Organization: Dept. Informatik, Swiss Federal Institute of Technology (ETH), Zurich, CH References: <MS-C.758525423.1103527590.mrc@Tomobiki-Cho.CAC.Washington.EDU> <2h5db7$c79@geraldo.cc.utexas.edu> Date: Fri, 14 Jan 1994 09:24:21 GMT In article <2h5db7$c79@geraldo.cc.utexas.edu>, Alex Currier <mycroft@ccwf.cc.utexas.edu> wrote: >In article <MS-C.758525423.1103527590.mrc@Tomobiki-Cho.CAC.Washington.EDU> >Mark Crispin <mrc@Tomobiki-Cho.CAC.Washington.EDU> writes: > >> NeXT seems to have outdone itself with 3.2. >> >> After trashing my filesystem -- a full day of fsck'ing and restoring to put [Rest of Mark's stuff deleted] > [...] >so I would imagine that either you seriously wear out your NeXT or that it >just doesn't like you. ;-) Why should his machine be different than the rest of us... ?? > Maybe you should provide us with some situational, user >specific information regarding your 3.2 problems rather than just dumping the >whole load on the OS. Could it be you're using some incompatible software? Mark, could it be you're using your own code ? >Have you tried stripping your rc.local file and removing those older fax >drivers (or whatever customizations you've made) or is it just easier to lay >the blame on NeXT as usual? Mark, prob. you better work in single user mode, it's easier to understand. And you will never see your teensy tiny window again. > >In fact, I'm quite happy with 3.2, seems a bit snappier, seems to utilize >memory better, hasn't blown up on me. I agree with Alex. It think this is Mark's own little personal problem. Dan
Newsgroups: comp.sys.next.bugs From: uunet!lkba!lkb (Larry Blische) Subject: Re: lovely 3.2 (jumping to conclusions?) Message-ID: <1994Jan14.140720.3341@lkba.uucp> Sender: lkb@lkba.uucp (Larry Blische) Organization: LKB Associates, Inc. References: <2h5db7$c79@geraldo.cc.utexas.edu> Date: Fri, 14 Jan 1994 14:07:20 GMT In article <2h5db7$c79@geraldo.cc.utexas.edu> mycroft@colourbox.utexas.edu (Alex Currier) writes: > In article > <MS-C.758525423.1103527590.mrc@Tomobiki-Cho.CAC.Washington.EDU> > Mark Crispin <mrc@Tomobiki-Cho.CAC.Washington.EDU> writes: > > > NeXT seems to have outdone itself with 3.2. > > > > After trashing my filesystem -- a full day of fsck'ing and > > restoring to put it back together -- 3.2 seems to take a special > > delight in panicing. Of course, this all happens inside the > > teensy tiny window which scrolls past a mile a minute then > > automatically reboots. Other than it's an ``invalid descriptor > > during MMU table walk'', I haven't recorded much else. Why do you think you had to wait so long for NeXT to make you those SPECIAL disks :-) > > I think you're being a bit hasty blaming your crashing and panic > problems on 3.2. I've had it installed as long as you have and > I haven't had any file system troubles, no crashes, no panics > (and this is with ND and optical disk, probably a bit more > fragile than your average NeXT)... what have you been doing to > your machine to cause such excitement? Mine doesn't exactly sit > idle so I would imagine that either you seriously wear out your > NeXT or that it just doesn't like you. Maybe you should provide > us with some situational, user specific information regarding > your 3.2 problems rather than just dumping the whole load on > the OS. Could it be you're using some incompatible software? > Have you tried stripping your rc.local file and removing those > older fax drivers (or whatever customizations you've made) or > is it just easier to lay the blame on NeXT as usual? > > In fact, I'm quite happy with 3.2, seems a bit snappier, seems > to utilize memory better, hasn't blown up on me. Seriously though, I'm not sure about him being too hasty but I do think that Mark is being a bit too all encompassing with his blame. After the 3.2 upgrade, this station also experienced system panics and the extended periods with our friend fsck. Not every day; maybe three times in two weeks. Since I had taken the upgrade path from 2.1 to 3.0 to 3.1 to 3.2 I thought that maybe I should try an install from scratch instead. I verified my last archive tape, did the install from the CD-ROM and restored my local stuff. I haven't had a panic since. And I agree with Alex that 3.2 feels a bit snappier. At this time I'd be inclined to say that there are problems with the upgrade app but not 3.2 in total. Mark, did you upgrade or install from scratch? I'd recommend that anyone having panics after a 3.2 upgrade should try the scratch install (see "Upgrading to NEXTSTEP Release 3.2", pg 20) after they have verified that their local stuff is properly backed up. --- Larry Blische lkba!lkb@uunet.UU.NET LKB Associates, Inc. NeXTmail Welcome! 3118 Dunglow Road 410 285 2262 Dundalk, Maryland 21222-5304 USA
Newsgroups: comp.sys.next.bugs From: mikes@ceco.ceco.com (Michael Stepniczka) Subject: Re: lovely 3.2 Message-ID: <CJMJvC.AAy@ceco.ceco.com> Sender: root@ceco.ceco.com (Operator) Organization: Commonwealth Edison Co. References: <MS-C.758525423.1103527590.mrc@Tomobiki-Cho.CAC.Washington.EDU> Date: Fri, 14 Jan 1994 14:49:59 GMT If all of this NeXT stuff is so bad, why use it in the first place?... In article mrc@Tomobiki-Cho.CAC.Washington.EDU, Mark Crispin <mrc@Tomobiki-Cho.CAC.Washington.EDU> writes: > Well well well. > > NeXT seems to have outdone itself with 3.2. I would have liked to see NeXTtime in it myself, but yes, the new graphical boot for the PC and FileMerger are pretty nice. > ><MUNCH> > > 3.2 seems to be crashing even more often than Windows or the Mac OS (and > *that's* saying something). > You must have REALLY stable versions of Windows and MacOS. I'm jealous. :) > Ah well, should get used to it. It may be the last version of NEXTSTEP. > Things would certainly be nicer without NEXTSTEP around. That's true. And if only they'd get rid of the damned interface while they're at it... But seriously, I have a PC and NeXTstation on 3.2, and haven't had a crash yet. The only "SORT OF" crash has been from 3DReality- and that just killed the window server. I have upgraded my station from 3.0 to 3.1 to 3.2 just doing the upgrade, not a reinstall, and haven't had problems. I upgraded the PC over ethernet from my Station, and haven't had problems there either. Maybe somewhere in installing SLIP or something else you changed a file that is now confusing the system? 3.0 was fairly unstable. 3.2 I haven't had any problems with. Maybe it's time to backup your stuff, install from scratch, and make sure nothing else is screwy. And before we get gripes about that, look at how easy other systems are to upgrade. Even the DOS upgrade can go bad. I doubt that it's hardware since the NeXT boxes were certainly well integrated. I do have complaints too though- drivers for the PC and that swapfile! :( Mike Stepniczka mikes@ceco.ceco.com
From: robin@pencom.com (Robin D. Wilson) Newsgroups: comp.sys.next.bugs Subject: Re: Serial Drivers Date: 14 Jan 1994 15:17:37 GMT Organization: Pencom Sofware Message-ID: <2h6d2h$jrb@digdug.pencom.com> References: <1994Jan13.144903.346@corona.com> In article <1994Jan13.144903.346@corona.com> pkron@corona.com (Peter Kron) writes: :From: Charles William Swiger <infidel+@CMU.EDU> :> Sigh. Remember the start bits too, Chuck, as someone :> gently pointed out to me. Repeat calculation with 10 :> bits per frame (1 start, 8 data, 1 stop), to get a maximum :> transfer rate of 240 cps. Subtract a few bytes for :> overhead, and that's your maximum transfer rate. : :This true for asynch communications, such as from the serial port to :your external modem and over direct modem links. My understanding is :that the LAPM and MNP protocols, however, actually use a synchronous :connection over the phone line, which does not require start/stop :bits. So uncompressed data maximum is back up to 300cps at 2400baud, :1200cps at 9600baud. Except that your serial port is still Async, and still transmits all 10 bits per character to the modem. I'm sure that part of the compression done by the modems allows the extra bits to be stripped, but there still has to be some overhead for that process. :The problems described by Kay Schulz sound like retransmissions due :to errors, which drop the *effective* rate dramatically due to the :delay before the error is recognized and subsequent retransmission. :I've seen conflicting opinions as to whether modem compression and :error-checking actually degrade UUCP performance, since they interfere :with UUCP's own (primitive) error protocol. I run mine at 4800, :compressed but not error-corrected, which gets me about 300cps for :UUCP. That's the best I've ever gotten out of my gateway. That's pretty sad performance. On our black HW we regularly get 1100cps for UUCP connections at 38.4Kbps. (That is using the Telebit, with 'g' protocol spoofing too, though...) I doubt the modem compression would damage the speed for UUCP. Likewise, the error control should have little effect. More than likely, the errors will come between the serial port on the computer and serial port on the modem. (Buffer overflows caused by poor implementations of RTS/CTS flow control.) :Any communication port is susceptible to lost data if one side gets :behind. This isn't really a bug, but an annoying inefficiency. I have :a suspicion, though, that NEXTSTEP/Intel simply isn't very good at :keeping up, particularly in the presence of disk activity. I notice :retransmissions go up whenever I am actively hitting the disk. Perhaps :Kay is transmitting in the presence of lots of other activity. When using 'good' flow control, this shouldn't be the case. If I have RTS/CTS pacing (correctly implemented, that is...) turned on, it should stop the flow whenever the either side can't keep up. I shouldn't have _any_ retransmits. Worst case (depending on the transmission protocol... assuming we are talking about more than UUCP here) is that I get a timeout on the side that is sending too fast. :It also corresponds with the erratic parallel port behavior I've seen :on 3.2. If I expand a .compressed file during a print, I am almost :certain to lose data. If I print on an idle system, things are ok for :a while. Even idle, longer datastreams will fail, and one repeatable :failure I found occurred right after an 8K boundary in the print :stream. Unfortunately, there is no error correction to the printer :(HP in graphics mode), so the result is garbled pages. This just sounds like bugs in the driver. How can your disk activity cause the buffer on the printer to overflow? Technically, increased disk activity should slow your transmission down, reducing the probability of buffer overflow (due to a lower transmission rate). :In my case, the problem may be worsened by running both IDE and SCSI :over an ISA bus. There are potentially two transfers occurring, either :of which may require a subsequent copy since the ISA bus can't address :all 28Mb memory. I think there are either some clear configuration parameters that are not set right on your box, or there are some serious bugs in the drivers for these devices. Talking to a printer should not be affected by disk activity. Perhaps the parallel port is not interupting properly, so flow control from the printer is handled too late? (Wild guess...) :This is largely speculation on my part, but what else is there to go :on? :NeXT, as usual, is silent on the matter. This is pretty common among hardware vendors. You don't advertise your bugs until you have a solution for them... -- ----------------------------------------------------------------------------- *** These are my opinions... Mine! All Mine! Minemineminemineminemine! *** ----------------------------------------------------------------------------- Robin D. Wilson robin@pencom.com Pencom Software 701 Canyon Bend Dr. 9050 Capital of Texas Hwy Pflugerville, TX 78660 Austin, TX 78759
From: robin@pencom.com (Robin D. Wilson) Newsgroups: comp.sys.next.bugs Subject: Re: lovely 3.2 Date: 14 Jan 1994 15:32:55 GMT Organization: Pencom Sofware Message-ID: <2h6dv7$o46@digdug.pencom.com> References: <MS-C.758537657.1103527590.mrc@Tomobiki-Cho.CAC.Washington.EDU> In article <MS-C.758537657.1103527590.mrc@Tomobiki-Cho.CAC.Washington.EDU> Mark Crispin <mrc@Tomobiki-Cho.CAC.Washington.EDU> writes: :The machine in question is a 16MB cube, 68040 25MHz CPU. It runs Transarc :SLIP and co-Xist. The peripherals on it are a AppleCD 300 CD-ROM and a :Telebit T3000 modem. It's on Ethernet. I didn't even know Transarc had a product for the NeXT.... (We do alot of work with them on Encina and DCE, but I'm pretty sure Mark meant 'Transys SLIP'.) :I've also seen the problem where windows and menus can't be moved; you try to :drag their title bar and find its locked. : :True, it could be hardware. But it's curious that these problems started :on the day that I installed 3.2. Before, that NeXT only paniced about once or :twice a month, same as my other two machines (which are on 3.0 more or less :permanently). Mark seems to have a number of problems that most of the rest of us have never had. I've been running NEXTSTEP (in it's various capitalizations) for 2 years now, and have never seen regular 'panics' on my machine (or the machines of anyone else at our facility -- which houses about 70 NeXTs). I installed NeXT hardware for clients, and they never seemed to have these problems either. Everyone I know uses their machines for software development, so they are quite busy every day. If my machine panic'd more than once every 6 months I'd be very concerned too. But I'd start by looking at the hardware to see if there were a problem. I wonder if Mark uses the same login that shows up in his net posts. You don't suppose that NeXT got tired of his constant complaining and developed the 'Mark Crispin panic'? Anytime the machine sees a 'mrc' user, with 'Mark Crispin' in the GCOS field of his password entry, it runs a 'random crash generator'. Over time, the odds of a random crash are modified until... Boom! Nah... Couldn't be. Steve would never permit something so unprofessional. %^) -- ----------------------------------------------------------------------------- *** These are my opinions... Mine! All Mine! Minemineminemineminemine! *** ----------------------------------------------------------------------------- Robin D. Wilson robin@pencom.com Pencom Software 701 Canyon Bend Dr. 9050 Capital of Texas Hwy Pflugerville, TX 78660 Austin, TX 78759
Newsgroups: comp.sys.next.bugs From: clloyd@gleap (Charles C. Lloyd) Subject: Re: lovely 3.2 Message-ID: <1994Jan14.152147.2422@gleap.sccsi.com> Sender: clloyd@gleap.sccsi.com Organization: GiantLeap Software References: <MS-C.758537657.1103527590.mrc@Tomobiki-Cho.CAC.Washington.EDU> Date: Fri, 14 Jan 1994 15:21:47 GMT Mark Crispin <mrc@Tomobiki-Cho.CAC.Washington.EDU> writes > >True, it could be hardware. But it's curious that these problems started >on the day that I installed 3.2. Before, that NeXT only paniced about once or >twice a month, same as my other two machines (which are on 3.0 more or less >permanently). My point (previously) was that the hardware problem (spurious interrupt) manifested itself only after the upgrade from 2.1 to 3.0. -- Charles Lloyd clloyd@GLeap.sccsi.com GiantLeap Software (713) 292-5853 or 363-9001 (Hou) (713) 363-9763 (fax)
From: Charles William Swiger <infidel+@CMU.EDU> Newsgroups: comp.sys.next.bugs Subject: Re: lovely 3.2 Date: Fri, 14 Jan 1994 10:43:32 -0500 Organization: Fifth yr. senior, Chemistry, Carnegie Mellon, Pittsburgh, PA Message-ID: <AhBfqYS00iV5Q2POgZ@andrew.cmu.edu> In-Reply-To: <MS-C.758537657.1103527590.mrc@Tomobiki-Cho.CAC.Washington.EDU> Excerpts from netnews.comp.sys.next.bugs: 14-Jan-94 Re: lovely 3.2 by Mark Crispin@Tomobiki-Ch > The machine in question is a 16MB cube, 68040 25MHz CPU. It runs Transarc > SLIP and co-Xist. The peripherals on it are a AppleCD 300 CD-ROM and a > Telebit T3000 modem. It's on Ethernet. I'd be a little suspicious of the SLIP package, as it partially consists of a loadable kernel module that attempts to patch the kernel directly (!). For all that DialUpIP SLIP is a wonderful package (I use it too), it obviously cannot have been tested much with NS 3.2. A buggy loadable kernel module has much greater potential to dramatically crash your system than some application not being run as root. > The crash which piddled on the filesystem, and the page fault crashes, both > happen while the system is being started. The filesystem crash seemed to > be ticked by my hitting the brightness key on the keyboard to try and dim > the screen (for some reason, 3.2 seems to want to set the screen to full > brightness). I'm avoiding the problem by keeping paws off the keyboard > until it finishes startup. Hmm. From what I understand of the changes in 3.2, there have been modifications to a fair number of low-level things (like the tty disciplines) for POSIX support that may be related to this. Although the brightness and volume keys may not ever reach that level, now that I think of it. -Chuck Charles William Swiger -- CMU...*splat*! | 1. You can't fly. --------------------------------------------+ 2. Cars are always real, even AMS & normal mail: infidel@cmu.edu | when they're not. Failing that: cs4w+@andrew.cmu.edu | 3. Police are not your friends. NeXTmail: chuck@cswiger.slip.andrew.cmu.edu | 4. Fire burns.
Newsgroups: comp.sys.next.bugs From: kelman@niehs.nih.gov Subject: Re: lovely 3.2 (jumping to conclusions?) Message-ID: <kelman.29.0@niehs.nih.gov> Sender: postman@alw.nih.gov (AMDS Postmaster) Organization: National Institutes of Health References: <2h5db7$c79@geraldo.cc.utexas.edu> <1994Jan14.140720.3341@lkba.uucp> Date: Fri, 14 Jan 1994 19:43:33 GMT In article <1994Jan14.140720.3341@lkba.uucp> uunet!lkba!lkb (Larry Blische) writes: [stuff deleted] >After the 3.2 upgrade, this station also experienced system panics and the >extended periods with our friend fsck. Not every day; maybe three times in >two weeks. Since I had taken the upgrade path from 2.1 to 3.0 to 3.1 to 3.2 >I thought that maybe I should try an install from scratch instead. I >verified my last archive tape, did the install from the CD-ROM and restored >my local stuff. >I haven't had a panic since. And I agree with Alex that 3.2 feels a bit >snappier. >At this time I'd be inclined to say that there are problems with the upgrade >app but not 3.2 in total. >Mark, did you upgrade or install from scratch? >I'd recommend that anyone having panics after a 3.2 upgrade should try the >scratch install (see "Upgrading to NEXTSTEP Release 3.2", pg 20) after they >have verified that their local stuff is properly backed up. I have upgraded 2.1 -> 3.0 -> 3.1 -> 3.2 and have not had any of these problems. I'm also running Transys SLIP. Not a single panic or crash. I rather doubt it is the upgrade app that is the problem, rather personal modifications you may have made to the machine. >--- >Larry Blische lkba!lkb@uunet.UU.NET >LKB Associates, Inc. NeXTmail Welcome! >3118 Dunglow Road 410 285 2262 >Dundalk, Maryland 21222-5304 USA
From: wmorse@erasure-sl.cc.emory.edu (William Morse) Newsgroups: comp.sys.next.bugs Subject: Re: lovely 3.2 Date: 14 Jan 1994 20:31:03 GMT Organization: Emory University, Dept of Math and CS Distribution: world Message-ID: <2h6ve7INNl0k@emory.mathcs.emory.edu> References: <MS-C.758537657.1103527590.mrc@Tomobiki-Cho.CAC.Washington.EDU> In article <MS-C.758537657.1103527590.mrc@Tomobiki-Cho.CAC.Washington.EDU> Mark Crispin <mrc@Tomobiki-Cho.CAC.Washington.EDU> writes: > The machine in question is a 16MB cube, 68040 25MHz CPU. It runs Transarc > SLIP and co-Xist. The peripherals on it are a AppleCD 300 CD-ROM and a > Telebit T3000 modem. It's on Ethernet. > > The crash which piddled on the filesystem, and the page fault crashes, both > happen while the system is being started. The filesystem crash seemed to be > ticked by my hitting the brightness key on the keyboard to try and dim the > screen (for some reason, 3.2 seems to want to set the screen to full > brightness). I'm avoiding the problem by keeping paws off the keyboard until > it finishes startup. > > I've also seen the problem where windows and menus can't be moved; you try to > drag their title bar and find its locked. > > True, it could be hardware. But it's curious that these problems started > on the day that I installed 3.2. Before, that NeXT only paniced about once or > twice a month, same as my other two machines (which are on 3.0 more or less > permanently). > > Stomping on a few dozen low-numbered inodes is *not* a good way to start a > relationship with a new operating system version! ------- Marc, I have an almost identical set up: NeXT Turbo Color w/ Transys SLIP and ethernet. I have gone from 3.0 --> 3.2 without problems. In fact, 3.2 seems more stable and less buggy than any of the others in the 3.x series. Needless to say, I am exceedingly happy with it. I think you have got a bad upgrade. The reasons for this vary: sometimes it is the sysadmin's fault (ie: not wanting to rebuild his config. files... so he uses the old ones and ends up rebuilding the system) and sometimes it is an act of God. Quite honestly, I have never personally had a bad upgrade with NS. Now, I can tell you some stories about Windows, OS/2 and Mac upgrades! In any case, with NeXTStep, like most UNIXes I know, once a system begins the long and slow decline... of system instability, it is best to just do a dump... rebuild and restore your files. I bet that that will cure the disease. William Consult. / ITD
From: rfschtkt%banruc60.bitnet@cc1.kuleuven.ac.be Newsgroups: comp.sys.next.bugs Subject: Re: DPS client library errors, Mail.app Date: 14 Jan 1994 19:48:58 -0600 Organization: UTexas Mail-to-News Gateway Sender: daemon@cs.utexas.edu Message-ID: <9401121242.AA00796@flexus> LS, I'm very sorry to have bothered you with a false report about a particular failure of Mail.app. The following provisional entry into KBNS has been removed, because the problem has suddenly disappeared. 1. The transfer of rich messages just works again. Was it logging out and back in again, or rebooting, that solved a possible transient error? I hope I'll never find out. 2. Sending to nobody is writing or appending the message to /dev/null (it's an alias somewhere in Netinfo). For whatever reason (I don't remember changing anything and I'm the only one with access to my machine), all permissions were removed from this device. I've since restored them. If somebody has experienced anything (upgrading from 3.1 to 3.2 on NeXT hardware maybe?) that does this to /dev/null, please tell me. Otherwise let's forget this. Again, I'm sorry. I did work on the problem for a while before reporting it, though, and now I'm giving the all-clear. So it might be all right to believe me next time... :-) Raf Schietekat, RfSchtkt@banruc60.bitnet, Flanders, Belgium RfSchtkt%banruc60.bitnet@uunet.uu.net or so if the above does not work for you (real, i.e., with triangle in the Deliver button) NeXTmail preferred I can't reach sites with ! or % in their address, or ending in .at or .uucp >>>>> KBNS.32.0.013_c3.2o Really awful bug, but probably not (entirely) Mail.app's fault? I'll classify it here because I saw this bug in Mail.app, but it probably is caused by something else, as you will see. It is not possible anymore to transfer rich messages between Mailboxes: NXRunAlertPanel("Transfer","Unable to transfer %s",NULL,NULL,NULL,thePath), where thePath is something like ``Active.mbox/No_subject.attach''. I've tested this up to moving my Mailboxes folder somewhere else and having Mail.app create one from scratch, and restoring 3.1's Mail.app, to no avail. Maybe Mail.app has its current directory in ~/Mailboxes and some [Application workspace] protocol now assumes a full path? I don't know, but this is a damned nuisance. Should I now downgrade to 3.1's Workspace? If you have had the same experience, or if this is totally strange to you, do drop me a note. XXX Now I can't send messages to nobody anymore: they're returned instead of thrown away. Is this a bug? /dev/null has no access privileges anymore: is this Upgrader.app's doing? <<<<< PS: I haven't processed any replies to this message yet (if there are any).
Newsgroups: comp.sys.next.bugs From: nathan@laplace.csb.yale.edu (Nathan F. Janette) Subject: Re: lovely 3.2 (jumping to conclusions?) Message-ID: <1994Jan15.015153.2575@cs.yale.edu> Sender: news@cs.yale.edu (Usenet News) Organization: Yale University, Department of Computer Science, New Haven, CT References: <kelman.29.0@niehs.nih.gov> Date: Sat, 15 Jan 1994 01:51:53 GMT In article <kelman.29.0@niehs.nih.gov> kelman@niehs.nih.gov writes: > I have upgraded 2.1 -> 3.0 -> 3.1 -> 3.2 and have not had any of these > problems. I'm also running Transys SLIP. Not a single panic or crash. I > rather doubt it is the upgrade app that is the problem, rather personal > modifications you may have made to the machine. Many folks have reported various problems using the upgrade apps over as many NEXTSTEP revisions as I can recall. I've always avoided the problems by rebuilding from scratch every time. Then again, I also sometimes get random panics on different machines, but not often enough to worry about. I return you to the gleeful Crispin pile-on here in the wild kingdom. -- Nathan "USENET" Janette PPP link from hilbert.csb.yale.edu Please reply to: nathan@laplace.csb.yale.edu (NeXT)
Newsgroups: comp.sys.next.bugs From: trapac@netcom.com (Transpacific Container) Subject: 3.2 performance problems? Message-ID: <trapacCJn7E1.FnE@netcom.com> Organization: NETCOM On-line Communication Services (408 241-9760 guest) Date: Fri, 14 Jan 1994 23:17:59 GMT Over the last couple of months, our app has gotten increasingly slower. While we dismissed much of the problem to the increased functionality that had been added, it has become a concern. Recently the consultant who wrote the app discovered when an older 3.1 version was recompiled, it ran a lot slower. It appears that something in the 3.2 compiler has changed. We are using NSFIP on Intel GX/Pro machines with 32MB memory. Because the app as a whole is slower, I wonder if something fundamental like memory management has changed. To summarize the application's performance: Compiled Executed Relative under under performance 3.1 3.1 1 3.1 3.2 1 3.2 3.2 approx. 4X slower Has anyone encountered simular problems? Does anyone know of changes in the 3.2 compiler that may cause performance problems? Any suggestions besides going back and reloading 3.1? Keith Carpenter trapac@netcom.com
Newsgroups: comp.sys.next.bugs From: Robert_La_Ferla@hot.com Subject: Re: lovely 3.2 Message-ID: <1994Jan14.222425.6749@hot.com> Sender: robertl@hot.com Organization: Hot Technologies References: <MS-C.758537657.1103527590.mrc@Tomobiki-Cho.CAC.Washington.EDU> Date: Fri, 14 Jan 1994 22:24:25 GMT Hmmm. I wonder if your CD-ROM is properly terminated. That would definitely fry your filesystem. But it's just a wild guess on my part. Robert La Ferla Hot Technologies NEXTSTEP ISV In article <MS-C.758537657.1103527590.mrc@Tomobiki-Cho.CAC.Washington.EDU> Mark Crispin <mrc@Tomobiki-Cho.CAC.Washington.EDU> writes: > The machine in question is a 16MB cube, 68040 25MHz CPU. It runs Transarc > SLIP and co-Xist. The peripherals on it are a AppleCD 300 CD-ROM and a > Telebit T3000 modem. It's on Ethernet. > > The crash which piddled on the filesystem, and the page fault crashes, both > happen while the system is being started. The filesystem crash seemed to be > ticked by my hitting the brightness key on the keyboard to try and dim the > screen (for some reason, 3.2 seems to want to set the screen to full > brightness). I'm avoiding the problem by keeping paws off the keyboard until > it finishes startup. > > I've also seen the problem where windows and menus can't be moved; you try to > drag their title bar and find its locked. > > True, it could be hardware. But it's curious that these problems started > on the day that I installed 3.2. Before, that NeXT only paniced about once or > twice a month, same as my other two machines (which are on 3.0 more or less > permanently). > > Stomping on a few dozen low-numbered inodes is *not* a good way to start a > relationship with a new operating system version! >
From: toon@moene.indiv.nluug.nl (Toon Moene) Newsgroups: comp.sys.next.bugs Subject: Re: lovely 3.2 Message-ID: <777@moene.indiv.nluug.nl> Date: 14 Jan 94 18:49:19 GMT References: <AhBfqYS00iV5Q2POgZ@andrew.cmu.edu> Sender: toon@moene.indiv.nluug.nl Organization: Moene Computational Physics, Maartensdijk, The Netherlands In article <AhBfqYS00iV5Q2POgZ@andrew.cmu.edu> infidel+@CMU.EDU (Charles William Swiger) writes: Marc Crispin: > > ... It runs Transarc SLIP and co-Xist. The peripherals on it are a > > AppleCD 300 CD-ROM and a Telebit T3000 modem. It's on Ethernet. > > I'd be a little suspicious of the SLIP package, as it partially consists > of a loadable kernel module that attempts to patch the kernel directly > (!). For all that DialUpIP SLIP is a wonderful package (I use it too), > it obviously cannot have been tested much with NS 3.2. A buggy loadable > kernel module has much greater potential to dramatically crash your > system than some application not being run as root. Interesting ... I upgraded my NS Color (non-turbo, 12 Mbyte RAM) from 2.1 -> 3.0 -> 3.1 -> 3.2 with Louis Mamakos' DialUpIP SLIP package (dated December '91 !), which does try to patch the kernel to cure a pre-3.1 bug (at least I haven't been able to trace it in 3.1 & 3.2). It backs off, though, when it realizes this isn't a 'known' kernel (See the date !). I haven't had a single problem since I upgraded to 3.2, the 22nd of December. Other equipment hanging off: NeXT CD-ROM drive, Neuron FAX 1414 modem (!), NeXT Laser Printer; no Ethernet. I didn't really push it yet, though - 3.1, however, has seen some unpredicted power outages when people turned off the power here while refurbishing the shower :-) No real damages - an fsck is only 2 minutes away on a 400 Mbyte drive ... I might also try the demo co-Xist that came with the 'Education Software Sampler' in '92 over SLIP - it did work with 3.0 (off the Cray Y/MP8 of the European Centre for Medium Range Weather Forecasts :-), although it was molasses slow ... -- Toon Moene (toon@moene.indiv.nluug.nl) Saturnushof 14, 3738 XG Maartensdijk, The Netherlands Phone: + 31 3461 4290; Fax: + 31 3461 4286 No Disclaimers; a NeXT@home protects against this occupational hazard. NB: Strictly black hardware here - No Intel in sight :-)
Newsgroups: comp.sys.next.bugs From: flash!jon@dsi.com Subject: Re: lovely 3.2 Message-ID: <CJnC3s.48o@dsinc!flash> Sender: jon@dsinc!flash (Jonathan Hendry) Organization: Who Needs It? References: <MS-C.758525423.1103527590.mrc@Tomobiki-Cho.CAC.Washington.EDU> Date: Sat, 15 Jan 1994 00:59:52 GMT In article <MS-C.758525423.1103527590.mrc@Tomobiki-Cho.CAC.Washington.EDU> Mark Crispin <mrc@Tomobiki-Cho.CAC.Washington.EDU> writes: > 3.2 seems to be crashing even more often than Windows or the Mac OS (and > *that's* saying something). Must be a karma thing. -- Jonathan W. Hendry Inexpensive NeXTSTEP Consulting tjhendry@mcs.drexel.edu For Your "Not-So-Mission-Critical" Apps
Newsgroups: comp.sys.next.bugs From: Neil Greene <Neil@bMD.com> Subject: Re: lovely 3.2 Message-ID: <1994Jan15.013305.27950@bMD.com> Sender: neil@bMD.com (Neil Greene) Organization: benchMark Developments, Inc. (Lex., KY) References: <MS-C.758525423.1103527590.mrc@Tomobiki-Cho.CAC.Washington.EDU> Date: Sat, 15 Jan 1994 01:33:05 GMT In article <MS-C.758525423.1103527590.mrc@Tomobiki-Cho.CAC.Washington.EDU> Mark Crispin <mrc@Tomobiki-Cho.CAC.Washington.EDU> writes: > Well well well. > > NeXT seems to have outdone itself with 3.2. > > After trashing my filesystem -- a full day of fsck'ing and restoring to put it > back together -- 3.2 seems to take a special delight in panicing. Of course, > this all happens inside the teensy tiny window which scrolls past a mile a > minute then automatically reboots. Other than it's an ``invalid descriptor > during MMU table walk'', I haven't recorded much else. > > Not that I've ever been able to figure when it auto-reboots and when it does > not. It seems that with some panics, it syncs the filesystem and kills all > processes, whereas with others, it ends up at the panic> prompt and waits for > you to swat it. > > 3.2 seems to be crashing even more often than Windows or the Mac OS (and > *that's* saying something). > > Ah well, should get used to it. It may be the last version of NEXTSTEP. > I think it would just be best if you picked a new OS. -- Neil Greene benchMark Developments, Inc. [NeXT VAR] 2040 Regency Road, Suite C Lexington, KY 40503 Phone: 606-231-6599 / Fax: 606-254-4864 Email: Neil@bMD.com [NeXTmail] -- Neil Greene benchMark Developments, Inc. [NeXT VAR] 2040 Regency Road, Suite C Lexington, KY 40503 Phone: 606-231-6599 / Fax: 606-254-4864
From: hketola@agsm.ucla.edu (Heikki Ketola) Newsgroups: comp.sys.next.bugs Subject: Re: lovely 3.2 (jumping to conclusions?) Date: 15 Jan 1994 16:52:57 GMT Organization: UCLA Microcomputer Support Office Message-ID: <2h9719$5ac@news.mic.ucla.edu> References: <MS-C.758525423.1103527590.mrc@Tomobiki-Cho.CAC.Washington.EDU> <2h5db7$c79@geraldo.cc.utexas.edu> I've had NS 3.2 running on my NeXTdimension for about a week -NS 3.2 seems to be working fine. No crashes, no panics, no major excitement. One commercial application is unhappy after the arrival NS 3.2. Is it me or are the swapfiles growing faster than under NS 3.1? It is not a big deal, but it just seems so.... Heikki
From: johannes@scribes.english.uiuc.edu (John B. Friedman) Newsgroups: comp.sys.next.bugs Subject: Help with dead? mail app 3.0 Date: 15 Jan 1994 20:38:16 GMT Organization: University of Illinois at Urbana Message-ID: <2h9k7o$36g@vixen.cso.uiuc.edu> Netters, could I get some advice? I had a freeze the other day when my modem clicked off in midst of transmitting mail and cursor kept whirling. 3.0 Mail was not active then though had been working fine just 10 minutes before. I used Halt, rebooted, got a message there had been a disk segment error now fixed. Clicked on Mail app in dock. It would highlight for a second or two and then quit and take me back to Workspace menu. I tried to pull app out of dock and replace with Mail app from Next apps. Same thing. I logged in as root and replaced the app. I got the mail to open there but there was no mail in it, just "welcome new user" message. The modem is definitely passing mail and seems OK. It is just that the mail menu won't come up and the window open when mail icon clicked . Any thoughts on where to begin, what kind of error message to look for if any would be most appreciated. Thanks very much. John Friedman
From: Mark Crispin <mrc@Ikkoku-Kan.Panda.COM> Newsgroups: comp.sys.next.bugs Subject: 3.2 crashes Date: Sat, 15 Jan 1994 13:27:05 -0800 Organization: University of Washington Message-ID: <MS-C.758669225.1103527590.mrc@Ikkoku-Kan.Panda.COM> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII I am on the trail of what seems to be an honest-to-goodness bug that caused the filesystem clobberage. The problem is in drawing up a reliable scenario to reproduce the problem. Things to watch out for: 1) doing an upgrade instead of a full install from scratch 2) usage of the login window's restart button instead of double-command-` and halt 3) usage of the brightness keys while the system is starting (and before the window manager has run) Items posted on this list that have been confirmed to be *not* a factor are: 1) memory failure 2) SCSI bus termination 3) ``local mods'' Transsys SLIP is still under investigation as a possible factor, but there has been no evidence to date blame Mamakos. As for the weenie comments (e.g. ``karma'', or ``get a new OS''), such as excellent examples of why many vendors don't want to touch the NeXT community with a 10-foot pole. Such comments make the entire NeXT community look bad and only serve to validate the image of NEXTSTEP as being a fringe system.
From: Mark Crispin <mrc@Ikkoku-Kan.Panda.COM> Newsgroups: comp.sys.next.bugs Subject: Re: 3.2 crashes Date: Sat, 15 Jan 1994 14:50:06 -0800 Organization: University of Washington Message-ID: <MS-C.758674206.662824084.mrc@Ikkoku-Kan.Panda.COM> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII To: mycroft@ccwf.cc.utexas.edu In-Reply-To: <2h9pui$d0k@geraldo.cc.utexas.edu> Alex Currier - Have you thought about the consequences of what you state? Suppose I did give up on using the NEXTSTEP operating system. Then my comment would be that I had to give up on NEXTSTEP because I was unusable. A vocal unhappy EX-customer is much more damaging than an unhappy current customer. Especially an ex-customer who is stuck with several thousands of dollars worth of useless equipment. It may be emotionally satisfying to you to tell me to buzz off, but it precisely that sort of comment that hurts your cause. You might consider what those comments look like to any corporation considering the purchase of NEXTSTEP.
From: izumi@mindseye.berkeley.edu (Izumi Ohzawa) Newsgroups: comp.sys.next.bugs Subject: Re: lovely 3.2 Date: 16 Jan 1994 01:56:46 GMT Organization: /etc/organization Message-ID: <2ha6su$kl9@agate.berkeley.edu> References: <MS-C.758525423.1103527590.mrc@Tomobiki-Cho.CAC.Washington.EDU> <CJnC3s.48o@dsinc!flash> In article <CJnC3s.48o@dsinc!flash> flash!jon@dsi.com writes: >In article <MS-C.758525423.1103527590.mrc@Tomobiki-Cho.CAC.Washington.EDU> Mark >Crispin <mrc@Tomobiki-Cho.CAC.Washington.EDU> writes: > >> 3.2 seems to be crashing even more often than Windows or the Mac OS (and >> *that's* saying something). > >Must be a karma thing. Yup. I think it's the node name that is inviting a bad karma. Try changing "Tomobiki-cho" to something more appropriate. :-) -- Izumi Ohzawa [ $BBg_78^=;(B ] USMail: University of California, 360 Minor Hall, Berkeley, CA 94720 Telephone: (510) 642-6440 Fax: (510) 642-3323 Internet: izumi@pinoko.berkeley.edu (NeXTmail OK)
From: robert@steffi.demon.co.uk (Robert Nicholson) Newsgroups: comp.sys.next.bugs Subject: Re: 3.2 crashes Date: 16 Jan 1994 00:25:11 -0000 Organization: me organized? That's a joke! Message-ID: <2ha1h7$2pu@steffi.demon.co.uk> References: <MS-C.758669225.1103527590.mrc@Ikkoku-Kan.Panda.COM> <2h9pui$d0k@geraldo.cc.utexas.edu> mycroft@ccwf.cc.utexas.edu wrote >In article <MS-C.758669225.1103527590.mrc@Ikkoku-Kan.Panda.COM> Mark Crispin ><mrc@Ikkoku-Kan.Panda.COM> writes: > >> As for the weenie comments (e.g. ``karma'', or ``get a new OS''), such as >> excellent examples of why many vendors don't want to touch the NeXT >> community with a 10-foot pole. Such comments make the entire NeXT community >> look bad and only serve to validate the image of NEXTSTEP as being a fringe >> system. > >(With all due respect) I'm sorry but I think this has a lot more to do with >your incessant complaining and bad mouthing of NeXTStep than any tendency for >the NeXT community to be rude to people with problems who need help. I think >the point was that *You, Mark Crispin* need to just give up on NS and move >along rather than any Joe User/Vendor who's having a problem with NS should >give up on it. > >This is not an evidence of the low quality of the NeXT community or the >fringeness of NeXTStep, it is an evidence that the NeXT community (such as it >is here) is getting bored of hearing you go on about NeXTStep and some people >feel you're just beating your own personal dead horse. I for one am not bored with Mark's "incessant complaining and bad mouthing". I consider that statement to be rather harsh actually. Mark has always conducted himself with the utmost courtesy in this group and it's only those folks who continue to stiffle his opinions and act in an offensive manner that are a bit tiresome. I for one enjoy seeing a multidimensional view of the world rather than the tunnel vision that most people are seeing. "If you want tunnel vision go read some white papers" I think the reason why Mark hasn't moved on to something else is obvious and it's the reason all of us haven't moved onto something else. There just aint anything else worth moving onto! We all love NeXTSTEP. I probably risk alienating myself from some of the diehard NeXT fans in this group but I don't think that people should be told to shut up when you don't like what they are saying. This is advocacy after all. It's no holds bared in here, get use to it. Now I return you to your previously scheduled flamefest :-) -- "You know what's wrong with you?" (Audrey Hepburn, Cary Grant) "Nothing" Charade, 1963 robert@steffi.demon.co.uk (ASCII for text only messages)
From: jweiss@casbah.acns.nwu.edu (Jerry Weiss) Newsgroups: comp.sys.next.bugs Subject: Re: 3.2 crashes Date: 16 Jan 1994 03:25:21 GMT Organization: Northwestern University, Evanston IL USA Message-ID: <2hac31$lem@news.acns.nwu.edu> References: <MS-C.758674206.662824084.mrc@Ikkoku-Kan.Panda.COM> In article <MS-C.758674206.662824084.mrc@Ikkoku-Kan.Panda.COM>, Mark Crispin <mrc@Ikkoku-Kan.Panda.COM> wrote: >Alex Currier - > > Have you thought about the consequences of what you state? > > Suppose I did give up on using the NEXTSTEP operating system. Then my >comment would be that I had to give up on NEXTSTEP because I was unusable. > > A vocal unhappy EX-customer is much more damaging than an unhappy current >customer. Especially an ex-customer who is stuck with several thousands of >dollars worth of useless equipment. > > It may be emotionally satisfying to you to tell me to buzz off, but it >precisely that sort of comment that hurts your cause. You might consider what >those comments look like to any corporation considering the purchase of >NEXTSTEP. > Mark, most of us no longer care what you think. We don't care if your predictions are true. We don't care if your technical comments are relevant. We are tired of "the sky is falling" attitude in your notes. Welcome to my KILL file, you've earned it. -- Jerry S. Weiss j-weiss@nwu.edu Dept. Medicine, Northwestern Univ. Medical School, Chicago, Illinois %SYSTEM-S-PHALOKTARG, Phasers Locked on Target, Ready to Fire
From: william@pinoko.berkeley.edu (William E. Grosso) Newsgroups: comp.sys.next.bugs Subject: Re: 3.2 crashes Date: 16 Jan 1994 03:46:29 GMT Organization: University of California, Berkeley Message-ID: <2hadal$lo2@agate.berkeley.edu> References: <2hac31$lem@news.acns.nwu.edu> Jerry Weiss writes > In article <MS-C.758674206.662824084.mrc@Ikkoku-Kan.Panda.COM>, > Mark Crispin <mrc@Ikkoku-Kan.Panda.COM> wrote: > > > > > > It may be emotionally satisfying to you to tell me to buzz off, but it > >precisely that sort of comment that hurts your cause. You might consider what > >those comments look like to any corporation considering the purchase of > >NEXTSTEP. > > > > Mark, most of us no longer care what you think. We don't care if your > predictions are true. We don't care if your technical comments are > relevant. We are tired of "the sky is falling" attitude in your notes. > Actually, I can (sortof) verify what Mr. Crispin says : People do occasionally wonder why Mr. Crispin gets abused. During one of the interminable flame wars that Mr. Crispin (somehow :-) ) gets involved in, I posted something like "Why do people even respond to Mark". I received e-mail from 2 distinct people asking me "Who is this Mark Crispin / Why does everyone treat him so badly ?" Moral: There are newbies out there and they pay attention to us. I have no idea if flaming Mark loses NeXT sales (I tend to think that corporations would take enough of a time-slice to realise Mark is quite the anomaly), but he has a point : the deluge of flame that inevitably follows any of his postings (especially when he is being reasonable) makes *us* (the quasi-mythical NeXT community) look bad (though, in our favor, not nearly as bad as those Amiga people :-) ). Bill Grosso
From: mycroft@colourbox.utexas.edu (Alex Currier) Newsgroups: comp.sys.next.bugs Subject: Re: 3.2 crashes Date: 16 Jan 1994 04:02:55 GMT Organization: The University of Texas at Austin, Austin, Texas Message-ID: <2hae9f$m1l@geraldo.cc.utexas.edu> References: <MS-C.758674206.662824084.mrc@Ikkoku-Kan.Panda.COM> I wonder how many corporations base their purchasing decisions on what they read on usenet? For every single thing ever made or sold there are some customers for whom it just doesn't work and who end up being disgruntled ex-customers. I don't think that your one machine's problems and a few comments made in half jest about your rather unique experiences with and opinions about NeXTStep are going to be detremental to NeXT as a company or are going to lose them any sales (they do that well enough for themselves). Anyone who has followed the c.s.n.* groups for any length of time knows about you (even my next door neighbor who doesn't own a computer recognizes your name from reading your posts over my shoulder) and knows how you feel about NeXTStep and would probably take your enthusiastically negative posts with a grain of salt. Really, this is nothing personal. I just think your making mountains out of a molehill here by claiming these replies to you are damaging the character of NeXTStep. NeXT doesn't live or die by what is said on usenet and even if it did I think the comments made to you would be taken in the spirit in which they were meant. In other words, lighten up. By the way, I am sorry your experiences with NeXTStep have been less than gratifying but (and I would tell this to my own mother) if it isn't working for you maybe you *should* just use something else. Pretty good advice, I think. Highest regards, Alex -- ============================================================================== Alex Currier * mycroft@ccwf.cc.utexas.edu * NeXTmail capable. Time is just one damn thing after another. ==============================================================================
Newsgroups: comp.sys.next.bugs From: lloyd@world.std.com (Christopher Lloyd) Subject: Re: 3.2 crashes Message-ID: <CJpIrw.32E@world.std.com> Organization: The World Public Access UNIX, Brookline, MA References: <2hac31$lem@news.acns.nwu.edu> <2hadal$lo2@agate.berkeley.edu> Date: Sun, 16 Jan 1994 05:19:06 GMT In article <2hadal$lo2@agate.berkeley.edu> william@pinoko.berkeley.edu (William E. Grosso) writes: >I have no idea if flaming XXXX loses NeXT sales (I tend to think >that corporations would take enough of a time-slice to realise >XXXX is quite the anomaly), but he has a point : the deluge of flame >that inevitably follows any of his postings (especially when he is >being reasonable) makes *us* (the quasi-mythical NeXT community) >look bad (though, in our favor, not nearly as bad as those >Amiga people :-) ). Yep, he does have a point, but it's two-sided. If you want to portray a group as nuts, you can drive them crazy enough to make them act like nuts. This is exacerbated on USENET where it is almost impossible to exercise group self-control, there is always a fresh supply of victims to make out as crazies. The whole thing is just unpleasant to read. I'm sure he'll pipe in and say we're all kooks to begin with and he does nothing to agitate things, but that's to be expected. After all, when you kick a dog until it bites you it's the dogs fault, right? blah, -- :: Christopher Lloyd :: Yrrid, Inc. :: lloyd@yrrid.com lloyd@world.std.com :: :: If I had known it was harmless, I would have killed it myself ::
Control: cancel <CJpIrw.32E@world.std.com> Newsgroups: comp.sys.next.bugs From: lloyd@world.std.com (Christopher Lloyd) Subject: cmsg cancel <CJpIrw.32E@world.std.com> Message-ID: <CJpJMJ.81q@world.std.com> Sender: lloyd@world.std.com (Christopher Lloyd) Organization: The World Public Access UNIX, Brookline, MA References: <2hac31$lem@news.acns.nwu.edu> <2hadal$lo2@agate.berkeley.edu> <CJpIrw.32E@world.std.com> Date: Sun, 16 Jan 1994 05:37:29 GMT <CJpIrw.32E@world.std.com> was cancelled from within rn. -- :: Christopher Lloyd :: Yrrid, Inc. :: lloyd@yrrid.com lloyd@world.std.com :: :: If I had known it was harmless, I would have killed it myself ::
From: pkron@corona.com (Peter Kron) Organization: Corona Design, Inc., Seattle, WA Distribution: world Date: Sat, 15 Jan 1994 20:10:05 PST Message-ID: <1994Jan16.041005.406@corona.com> References: <1993Dec24.052014.261@corona.com> Newsgroups: comp.sys.next.bugs Subject: Parallel port broken in 3.2 NS/I? I wrote: > Since 3.1 came out in May, I have been reliably printing > to my HP DeskJet 550C using the configuration available > on the archives and Ghostscript 2.6. Since upgrading to > 3.2, however, printing is barely usable. If I'm lucky I > can print a full page without some character getting > dropped, breaking the data protocol and causing the > printer to spew out pages of smileys and hearts in the > finest DOS fashion. > > <etc> NextAnswers-1330 mentions that the parallel port driver does not support bidirectional data transfer in 3.1. I previously assumed that did not apply to flow control to the printer, either as a special case or due to hardware flow control. If not, why would printing have worked in 3.1, which supposedly had the same problem? In any case, it now appears that something has changed or at least no longer works, such that flow control is indeed the problem. I tested this theory by adding a throttle process in the pipe between Ghostscript and the device. This process limits data to the printer to about 1K/sec. That seems to help a lot for B&W output. The parameter for color may vary...I have to experiment with it some more. This isn't very satisfactory, since page ejects are quite slow and can back up the buffer enough to cause problems. Running out of paper will also cause failures. But at least for now it looks fairly reliable, and I ought to be able to get most anything out 1 page at a time at worst. --- NeXTMail:Peter_Kron@corona.com Corona Design, Inc. P.O. Box 51022 Seattle, WA 98115-1022
From: jweiss@casbah.acns.nwu.edu (Jerry Weiss) Newsgroups: comp.sys.next.bugs Subject: Re: Parallel port broken in 3.2 NS/I? Date: 16 Jan 1994 06:15:45 GMT Organization: Northwestern University, Evanston IL USA Message-ID: <2ham2h$oi2@news.acns.nwu.edu> References: <1993Dec24.052014.261@corona.com> <1994Jan16.041005.406@corona.com> In article <1994Jan16.041005.406@corona.com>, Peter Kron <pkron@corona.com> wrote: >I wrote: >> Since 3.1 came out in May, I have been reliably printing >> to my HP DeskJet 550C using the configuration available >> on the archives and Ghostscript 2.6. Since upgrading to >> 3.2, however, printing is barely usable. If I'm lucky I >> can print a full page without some character getting >> dropped, breaking the data protocol and causing the >> printer to spew out pages of smileys and hearts in the >> finest DOS fashion. >> >> <etc> > >NextAnswers-1330 mentions that the parallel port driver does not >support bidirectional data transfer in 3.1. I previously assumed that >did not apply to flow control to the printer, either as a special case >or due to hardware flow control. If not, why would printing have >worked in 3.1, which supposedly had the same problem? > >In any case, it now appears that something has changed or at least no >longer works, such that flow control is indeed the problem. I tested >this theory by adding a throttle process in the pipe between >Ghostscript and the device. This process limits data to the printer >to about 1K/sec. That seems to help a lot for B&W output. The >parameter for color may vary...I have to experiment with it some more. > >This isn't very satisfactory, since page ejects are quite slow and >can back up the buffer enough to cause problems. Running out of paper >will also cause failures. But at least for now it looks fairly >reliable, and I ought to be able to get most anything out 1 page at >a time at worst. >--- I use the parallel port for my HP 4ML. Flow control is not a problem and if the printer runs out of paper, the Next notifies me. When they refer to bidirectional transfer, I would think they mean that replies or data from the printer are not carried back to the Next. This does not apply to flow control or whatever signal is convention for paper out signal. If I get a paper misfeed, the Next waits patiently for the problem to clear. No data loss under 3.1 or 3.2. I would check /private/adm/lpd? for messages in the printer log and double check your cabling again (perhaps inside the PC if necessary). Jerry -- Jerry S. Weiss j-weiss@nwu.edu Dept. Medicine, Northwestern Univ. Medical School, Chicago, Illinois %SYSTEM-S-PHALOKTARG, Phasers Locked on Target, Ready to Fire
Newsgroups: comp.sys.next.bugs From: c_avery@rsgreet.andi.org (Christopher Avery) Subject: 3.2 crashes Message-ID: <CJp6o2.ArD@nextsrv1.andi.org> Keywords: Operating System Sender: usenet@nextsrv1.andi.org (usenet) Organization: Association of NeXTSTEP Developers International Date: Sun, 16 Jan 1994 00:57:37 GMT When I upgraded to 3.2, the system rebooted fine. The next day when I brought the machine up, It went to the monitor window with a panic crash. I booted the machine without turning it off and went through a long fsck. The machine has been operating fine since then. As far as I know, I did not touch any keys during bootup. I suspect there may be an upgrade problem since it is hard to perfect this type of app. Also, my more experienced friends recommend a full install. I also have not been able to read a macintosh formatted floppy under 3.2. Since I did not try to read such a floppy under 3.0 or 3.1, I can't say whether this is a recent problem or not. But the disc read fine on a Macintosh. -- Christopher Avery reply to:cavery@andi.org (NeXTMail accepted)
Newsgroups: comp.sys.next.bugs From: rick@TotSysSoft.com (Richard Jacoby) Subject: Re: 3.2 crashes Message-ID: <1994Jan16.083146.2380@ToTSySSoft.com> Sender: news@ToTSySSoft.com Organization: Total System Software References: <2ha1h7$2pu@steffi.demon.co.uk> Date: Sun, 16 Jan 1994 08:31:46 GMT In article <2ha1h7$2pu@steffi.demon.co.uk> robert@steffi.demon.co.uk (Robert Nicholson) writes: [stuff deleted] > I for one am not bored with Mark's "incessant complaining and bad mouthing". > I consider that statement to be rather harsh actually. Mark has > always conducted himself with the utmost courtesy in this group Im staring at the monitor with my mouth wide open!! [stuff deleted] Rick
From: robert@steffi.demon.co.uk (Robert Nicholson) Newsgroups: comp.sys.next.bugs Subject: Re: 3.2 crashes Date: 16 Jan 1994 11:16:26 -0000 Organization: me organized? That's a joke! Message-ID: <2hb7ma$mk@steffi.demon.co.uk> References: <MS-C.758674206.662824084.mrc@Ikkoku-Kan.Panda.COM> <2hae9f$m1l@geraldo.cc.utexas.edu> mycroft@ccwf.cc.utexas.edu wrote >By the way, I am sorry your experiences with NeXTStep have been less than >gratifying but (and I would tell this to my own mother) if it isn't working >for you maybe you *should* just use something else. Pretty good advice, I >think. Why should he? Because he'll then be silenced? Or you won't have to listen to him anymore? Regardless of wether Mark's comments effect potential sales or not _nobody_ deserves to be treated that way. It's his investment he can chose what he wants to do with it. Like I said earlier, there isn't anything else worth using and when there is I'm sure everybody will "just use something else" In mean time either read his postings or don't, it's that simple. -- "You know what's wrong with you?" (Audrey Hepburn, Cary Grant) "Nothing" Charade, 1963 robert@steffi.demon.co.uk (ASCII for text only messages)
Newsgroups: comp.sys.next.bugs From: patrick@wbar.uta.edu (Patrick Phillips) Subject: Re: lovely 3.2 Message-ID: <1994Jan14.155357.11090@news.uta.edu> Sender: news@news.uta.edu (USENET News System) Organization: University of Texas at Arlington References: <MS-C.758537657.1103527590.mrc@Tomobiki-Cho.CAC.Washington.EDU> Date: Fri, 14 Jan 1994 15:53:57 GMT In article <MS-C.758537657.1103527590.mrc@Tomobiki-Cho.CAC.Washington.EDU> Mark Crispin <mrc@Tomobiki-Cho.CAC.Washington.EDU> writes: > The machine in question is a 16MB cube, 68040 25MHz CPU. It runs > Transarc SLIP and co-Xist. I don't know if this is the same thing, but in the "Upgrading to NEXTSTEP Release 3.2" book it says (p. 7): "If you're running Transarc's AFS, you should remove it before upgrading. If you don't, you won't be able to restart the computer after you upgrade." I don't know anything about Transarc products, so I don't know if their SLIP and AFS products share this problem in common. Patrick Phillips University of Texas at Arlington patrick@wbar.uta.edu (NeXTmail)
Newsgroups: comp.sys.next.bugs From: joe@FreemanSoft.com (Joe Freeman) Subject: Re: 3.2 crashes (kill file) Message-ID: <1994Jan16.130348.1906@FreemanSoft.com> Sender: jfreeman@FreemanSoft.com Organization: FreemanSoft Inc. References: <2hb7ma$mk@steffi.demon.co.uk> Distribution: usa Date: Sun, 16 Jan 1994 13:03:48 GMT In article <2hb7ma$mk@steffi.demon.co.uk> robert@steffi.demon.co.uk (Robert Nicholson) writes: > mycroft@ccwf.cc.utexas.edu wrote > In mean time either read his postings or don't, it's that simple. I would be a log easier if NewsGrazer supported kill files. -- Joe Freeman FreemanSoft Inc. A NEXTSTEP software and consulting services company. Electronic Mail: Joe@FreemanSoft.com (NeXT Mail) Voice: 919.783.7033
Newsgroups: comp.sys.next.bugs From: bruce@TotSysSoft.com (Bruce Gingery) Subject: Re: Amiga/OpenStep? Message-ID: <1994Jan16.133649.8890@ToTSySSoft.com> Sender: news@ToTSySSoft.com Organization: Total System Software References: <2hadal$lo2@agate.berkeley.edu> Date: Sun, 16 Jan 1994 13:36:49 GMT In article <2hadal$lo2@agate.berkeley.edu> william@pinoko.berkeley.edu (William E. Grosso) writes: [most of "Bash Mark Crispin" thread omitted -> I received e-mail from 2 distinct people asking me "Who is -> this Mark Crispin / Why does everyone treat him so badly ?" -> -> Moral: There are newbies out there and they pay attention -> to us. -> -> I have no idea if flaming Mark loses NeXT sales (I tend -> to think that corporations would take enough of a -> time-slice to realise Mark is quite the anomaly), but -> he has a point : the deluge of flame that inevitably -> follows any of his postings (especially when he is -> being reasonable) makes *us* (the quasi-mythical NeXT -> community) look bad (though, in our favor, not nearly -> as bad as those Amiga people :-) ). Bill, And yet, if you consider the hardware, and the video production acceptance that Toaster equipped Amigas are gaining, you might, as I have, decide that the Ami-4000 might just be the best direct port -- far exceeding the FIP port in performance AND solidity. Consider the following: 1. m68k hardware 2. Blitted video COULD make for faster composite to screen than on anything else running NEXTSTEP if properly supported in the PostScript server. 3. Low-end pricing, unless you DO add the toaster. 4. No need to retain AmigaDOS compatibility, as it could be bootstrapped from a floppy if dual booting were desired. 5. Lots of peripherals available. 6. OTHER Amiga Unix flavors are not gaining accept- ance very fast. 7. C=/Amiga still makes hardware and it would be to their advantage to cooperate, and although they are short on true task segregation, they've had shared libraries for a long time (albeit not pre-link FIXED ones). And perhaps the most important, no MicroSoft sabotage. Check out the multi-boot compatibility with DOS 6.2-- not even compatible with other versions of DOS. From what I've heard so far, only OS/2 has really found a solid way around it, and the existance of that is complete hearsay. Well, there's my monthly 2 for this group. Back to occasional lurking here. Bruce Gingery bruce@TotSysSoft.com
From: bkr@drdhh.hanse.de (Bjoern Kriews) Newsgroups: comp.sys.next.bugs Subject: Re: lovely 3.2 Date: 15 Jan 1994 05:21:03 +0100 Organization: Digital Island Message-ID: <bkr.758607476@drdhh> References: <MS-C.758525423.1103527590.mrc@Tomobiki-Cho.CAC.Washington.EDU> Mark Crispin <mrc@Tomobiki-Cho.CAC.Washington.EDU> writes: >3.2 seems to be crashing even more often than Windows or the Mac OS (and >*that's* saying something). I run 3.2 for over a month now and I have not had a single panic. Then, I didi not use PPP dialin, only dialout. But the machine has heavy 38400-modem load (up to 21600 physical, ~10hours a day) and it is networked with a dos-pc for 3 days now and Windows definitely crashes more often than this machine. Ineresting however, since I had these page faults with PPP dialin (Cuperman's public 0.3). Greetings, Bjoern -- bkr@drdhh.hanse.de - Bjoern Kriews - Stormsweg 6 - D-22085 Hamburg [76] - FRG "Insisting on perfect safety is for people who don't have the balls to live in the real world." -- Mary Shafer, NASA Ames Dryden
From: bkr@drdhh.hanse.de (Bjoern Kriews) Newsgroups: comp.sys.next.bugs Subject: Re: lovely 3.2 Date: 15 Jan 1994 05:31:45 +0100 Organization: Digital Island Message-ID: <bkr.758607877@drdhh> References: <MS-C.758537657.1103527590.mrc@Tomobiki-Cho.CAC.Washington.EDU> <2h6dv7$o46@digdug.pencom.com> robin@pencom.com (Robin D. Wilson) writes: >If my machine panic'd more than once every 6 months I'd be very concerned >too. But I'd start by looking at the hardware to see if there were a >problem. >I wonder if Mark uses the same login that shows up in his net posts. You >don't suppose that NeXT got tired of his constant complaining and developed >the 'Mark Crispin panic'? I had my machine panic under 3.1 twice a week - I tried compiling some stuff with -posix, then removed all that stuff (YES, I've CHECKED). The panics continued for around a week afterwards, then disappeared. It always happened with serial load - the machine ran for days with heavy modem load but it never crashed without modem load :-). Mark was the only one who experienced the same problem [new_posix_proc() - posic struct already exists] - and I was not running SLIP. The Unix part is not stable. I don't get NeXTSTEP power from a Sparc, but I definitely don't get a patched SUNOS 4.1.3's reliability from a NeXT (now). - Bjoern -- bkr@drdhh.hanse.de - Bjoern Kriews - Stormsweg 6 - D-22085 Hamburg [76] - FRG "Insisting on perfect safety is for people who don't have the balls to live in the real world." -- Mary Shafer, NASA Ames Dryden
From: kay@nagasena.muc.de (Kay Schulz) Newsgroups: comp.sys.next.bugs,comp.soft-sys.nextstep,comp.sys.next.advocacy,comp.sys.next.misc,de.comp.sys.next,comp.sys.next.sysadmin Subject: driverLoader scandir error Date: 16 Jan 1994 19:38:31 +0100 Organization: Firestone Chaotic Group Message-ID: <2hc1j7$8v@nagasena.muc.de> Hi, I tried to install the Mux8 driver but I got problems with it. The problem is: I have a serial mouse on com1 and I wanted the MUX8 on COM2. But all we tried failed. No we installed the /usr/Devices from CDROM so that I can use the old drivers. at 9600 the serials work, but not with 19200 or above. But also I have the normal configuration as before (with Configure.APP) I get an error message during boottime. it looks like that: driverLoader:scandir (/usr/Devices/SerialPorts.config) error or something like this. My mouse and my modem work (if you can read this). What does the scnadir mean and where is the problem? What do I have to do to get the MUX8 working on COM2 and the normal serials for my mouse on COM1 ? Do I have to change the Instance0.table or the Default.table? I am not familiar with it. Can someone help? Thanks -- Kay Schulz kay@nagasena.muc.de NeXTmail accepted
From: spagiola@frinext.stanford.edu (Stefano Pagiola) Newsgroups: comp.sys.next.bugs Subject: Re: 3.2 crashes (kill file) Date: 16 Jan 1994 18:39:19 GMT Organization: Stanford University Distribution: usa Message-ID: <2hc1kn$ero@nntp2.Stanford.EDU> References: <1994Jan16.130348.1906@FreemanSoft.com> Joe Freeman writes > > In mean time either read his postings or don't, it's that simple. > > I would be a log easier if NewsGrazer supported kill files. It does. But it calls it "filtering." (perhaps in response to all those complaints about "too much violence on the screen" :-). cmd-k puts the subject of the currently-selected article in the "filter" file. cmd-j does the same for the author's name, cmd-l for the author's account number. Look in the group menu. -- - Stefano Pagiola Food Research Institute, Stanford University spagiola@leland.stanford.edu (NeXTMail encouraged) spagiola@FRI-nxt-Pagiola.stanford.edu (NeXTMail encouraged)
From: mycroft@colourbox.utexas.edu (Alex Currier) Newsgroups: comp.sys.next.bugs Subject: Re: 3.2 crashes Date: 16 Jan 1994 19:51:54 GMT Organization: The University of Texas at Austin, Austin, Texas Message-ID: <2hc5sq$f1f@geraldo.cc.utexas.edu> References: <2hb7ma$mk@steffi.demon.co.uk> In article <2hb7ma$mk@steffi.demon.co.uk> robert@steffi.demon.co.uk (Robert Nicholson) writes: > mycroft@ccwf.cc.utexas.edu wrote > >By the way, I am sorry your experiences with NeXTStep have been less than > >gratifying but (and I would tell this to my own mother) if it isn't working > >for you maybe you *should* just use something else. Pretty good advice, I > >think. > > Why should he? Because he'll then be silenced? Or you won't have to > listen to him anymore? Regardless of wether Mark's comments effect > potential sales or not _nobody_ deserves to be treated that way. It's > his investment he can chose what he wants to do with it. Like I said > earlier, there isn't anything else worth using and when there is I'm > sure everybody will "just use something else" From a "strictly business" point of view, if you invest in something and it just doesn't pan out you have two options really, fight with it and waste (more) time and energy with a non-workable situation or cut your losses and walk away from it. Mark could easily sell off his NeXT hardware and buy something that will work better for him (he could take the trade up to Sun deal or some such) which would make a lot more sense than struggling with a non-functional computer (as he seems to feel the NeXT is). I honestly don't care which he does and I don't mind reading his (often entertaining and sometimes informative) posts. Recall I was not one of the people who made "it's Karma" comments... I just felt he was jumping to conclusions by blaming the OS before testing for the problem. Mark can stay or go as he pleases but I think he's wasting *his own* time by hanging on to NeXTStep if he truly feels it is useless (an opinion he's been stating here for quite some time now). In any case, I'm dropping the subject. I've had more than my say about it. -- ============================================================================== Alex Currier * mycroft@ccwf.cc.utexas.edu * NeXTmail capable. Time is just one damn thing after another. ==============================================================================
From: Mark Crispin <mrc@Ikkoku-Kan.Panda.COM> Newsgroups: comp.sys.next.bugs Subject: Re: lovely 3.2 Date: Sun, 16 Jan 1994 13:13:01 -0800 Organization: University of Washington Message-ID: <MS-C.758754781.1103527590.mrc@Ikkoku-Kan.Panda.COM> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII To: Bjoern Kriews <bkr@drdhh.hanse.de> In-Reply-To: <bkr.758607877@drdhh> Bjoern - NeXT claims that the new_posix_proc() - posix struct already exists bug has been fixed in 3.2. This bug was quite difficult to reproduce. Basically, what happened was that a static cell was used to store an intermediate variable in re-entrant code. Moving it to the stack fixed the problem. I believe (this is from memory) that the best way to trigger it was to leave the machine idle, then suddenly create two processes in rapid succession. This is precisely the sort of thing that might happen on a server machine, although the 2AM crontab entry on most NeXTs might also trigger it. -- Mark --
Newsgroups: comp.sys.next.bugs From: nathan@laplace.csb.yale.edu (Nathan F. Janette) Subject: Re: 3.2 crashes Message-ID: <1994Jan16.214947.7694@cs.yale.edu> Sender: news@cs.yale.edu (Usenet News) Organization: Yale University, Department of Computer Science, New Haven, CT References: <2hac31$lem@news.acns.nwu.edu> Date: Sun, 16 Jan 1994 21:49:47 GMT In article <2hac31$lem@news.acns.nwu.edu> jweiss@casbah.acns.nwu.edu (Jerry Weiss) writes: > Mark, most of us no longer care what you think. We don't care if your > predictions are true. We don't care if your technical comments are > relevant. We are tired of "the sky is falling" attitude in your notes. Reality check, Jerry: you have no idea what most of "us" think about Mark Crispin, so how about confining your comments to your own opinions? IMHO, Mark has added much more to the interesting content of these newsgroups than most of his flamers. I often disagree with his opinions, but I'd miss them if he stopped posting. So put him in your kill file, and be happy again. I'll gut it out ;-) -- Nathan "USENET" Janette PPP link from hilbert.csb.yale.edu Please reply to: nathan@laplace.csb.yale.edu (NeXT)
From: jeske@ux4.cso.uiuc.edu (David Jeske) Newsgroups: comp.sys.next.bugs Subject: Re: Amiga/OpenStep? Date: 17 Jan 1994 01:19:26 GMT Organization: University of Illinois at Urbana Message-ID: <2hcp2u$3d4@vixen.cso.uiuc.edu> References: <2hadal$lo2@agate.berkeley.edu> <1994Jan16.133649.8890@ToTSySSoft.com> bruce@TotSysSoft.com (Bruce Gingery) writes: > 1. m68k hardware > 2. Blitted video COULD make for faster composite to > screen than on anything else running NEXTSTEP > if properly supported in the PostScript server. I have talked with several people at Adobe and all those that I talked to were not fond of blitters at all. It seems that DPS and blitters just don't have the same things in mind. NeXTStep could ONLY benefit from blitters in the department of moving windows. The people at Adobe AND next both seemed to hint at the NeXT window server NEVER being taylored to doing things such as window moves on blitters. Sorry, it's high bandwidth video only. I have not idea what the amiga video is like, but if it's not high bandwidth, then it would surely be a dog. (FYI: the preference of the Adobe people was to have 2 main processors running Symmetric, and slave one to the video. So that when it was not processing video it could do something usefull. Sounds like the realm of PowerPC and other up and coming PC-class processors designed for multiprocessing. -- David Jeske(N9LCA)/CompEng Student at Univ of Ill at Cham-Urbana/NeXT Programmer CoCreator of the GTalk Chat Software System - online at (708)998-0008 Mail: jeske@ux4.cso.uiuc.edu NeXTMail: jeske@sumter.cso.uiuc.edu
Newsgroups: comp.sys.next.bugs From: flash!jon@dsi.com Subject: Re: 3.2 crashes Message-ID: <CJr04p.1uJ@dsinc!flash> Sender: jon@dsinc!flash (Jonathan Hendry) Organization: Who Needs It? References: <CJp6o2.ArD@nextsrv1.andi.org> Date: Mon, 17 Jan 1994 00:31:36 GMT In article <CJp6o2.ArD@nextsrv1.andi.org> c_avery@rsgreet.andi.org (Christopher Avery) writes: > I also have not been able to read a macintosh formatted floppy under 3.2. > Since I did not try to read such a floppy under 3.0 or 3.1, I can't say > whether this is a recent problem or not. But the disc read fine on a > Macintosh. Hm. Personally, I use Mac disks exclusively with my 3.2 Cube, and haven't had any trouble. Heck, my cube formats Mac disks better than most Macs. It's probably the disk. - Jon -- Jonathan W. Hendry Inexpensive NeXTSTEP Consulting tjhendry@mcs.drexel.edu For Your "Not-So-Mission-Critical" Apps
Newsgroups: comp.sys.next.bugs From: flash!jon@dsi.com Subject: Re: 3.2 crashes Message-ID: <CJr0x9.3A8@dsinc!flash> Sender: jon@dsinc!flash (Jonathan Hendry) Organization: Who Needs It? References: <2h9pui$d0k@geraldo.cc.utexas.edu> Date: Mon, 17 Jan 1994 00:48:45 GMT In article <MS-C.758669225.1103527590.mrc@Ikkoku-Kan.Panda.COM> Mark Crispin <mrc@Ikkoku-Kan.Panda.COM> writes: > > As for the weenie comments (e.g. ``karma'', or ``get a new OS''), such as > excellent examples of why many vendors don't want to touch the NeXT > community with a 10-foot pole. Such comments make the entire NeXT community > look bad and only serve to validate the image of NEXTSTEP as being a fringe > system. I'm sure weenie comments are a big factor when vendors are considering what platforms to use or develop for. "We don't plan on supporting NeXTSTEP. Not so much because of the small installed base and reported instability of NeXT, but primarily because some of the users are weenies. We've studied the demographics and concluded that the vocal weenie minority heavily outweighs the financial, military, medical, legal, and intelligence markets which use NeXTSTEP." Personally, I just find it amusing that the person who rails against NeXTSTEP most has been afflicted with a (rather severe) bug which no one else has reported. As if that weren't enough, you characteristically report it in a post which basically has an attitude of "I've had a problem. Obviously this means NeXTSTEP 3.2 is an unusable, buggy piece of software. Never mind the overwhelming lack of evidence to the contrary, and the fact that this problem hasn't been reported before on the net. It happened to me, therefore it is a universal bug, thus NeXTSTEP must be worthless. It could not possibly be a problem with my machine or my software." -- Jonathan W. Hendry Inexpensive NeXTSTEP Consulting tjhendry@mcs.drexel.edu For Your "Not-So-Mission-Critical" Apps Major: CIS
From: Mark Crispin <mrc@Ikkoku-Kan.Panda.COM> Newsgroups: comp.sys.next.bugs Subject: 3.2 crashes Date: Sun, 16 Jan 1994 19:02:40 -0800 Organization: University of Washington Message-ID: <MS-C.758775760.1103527590.mrc@Ikkoku-Kan.Panda.COM> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII Jonathan Hendry writes: >I've had a problem. Obviously this means >NeXTSTEP 3.2 is an unusable, buggy piece of software. Never mind the >overwhelming lack of evidence to the contrary, and the fact that this problem >hasn't been reported before on the net. It happened to me, therefore it is a >universal bug, thus NeXTSTEP must be worthless. It could not possibly be a >problem with my machine or my software. Johnny, take a Valium. You are hallucinating. I never said anything of the sort. To be so defensive, it sounds like you yourself are worried about the quality of your NEXTSTEP software. There *have* been corroborations to my my report, and in particular the belief that doing an upgrade (instead of a complete reinstall) and/or the use of the brightness keys before the windowserver starts may lead to crashes. The most probable explanation for the inode clobberage is that the bug which cause the crashes damaged some filesystem update code without immediately crashing the system. It is probably going to be extremely unlikely to reproduce this particular problem. If you are not concerned about the possibility that such a serious bug may exist, that is your concern.
From: gad@eclipse.its.rpi.edu (Garance A. Drosehn) Newsgroups: comp.sys.next.bugs Subject: Re: 3.2 crashes Date: 17 Jan 1994 03:35:28 GMT Organization: Rensselaer Polytechnic Institute, Troy NY, USA Distribution: world Message-ID: <2hd121$rka@usenet.rpi.edu> References: <2ha1h7$2pu@steffi.demon.co.uk> robert@steffi.demon.co.uk (Robert Nicholson) writes: > I probably risk alienating myself from some of the diehard NeXT fans > in this group but I don't think that people should be told to shut up > when you don't like what they are saying. This is advocacy after > all. It's no holds bared in here, get use to it. Actually, *this* whole thread is in comp.sys.next.bugs. It is not "no hold bared" in here. The thread does not belong here (not the flamefest, at least -- Mark's original article was an attempt to pin down the circumstances of some bug, so that certainly does belong). -- Garance Alistair Drosehn = gad@eclipse.its.rpi.edu ITS Systems Programmer (handles NeXT-type mail) Rensselaer Polytechnic Institute; Troy NY USA
Newsgroups: comp.sys.next.bugs From: dhowland@gpu.srv.ualberta.ca (Deborah Howland) Subject: Re: 3.2 crashes Message-ID: <dhowland.758783682@kakwa.ucs.ualberta.ca> Sender: news@kakwa.ucs.ualberta.ca Organization: University Of Alberta, Edmonton Canada References: <MS-C.758669225.1103527590.mrc@Ikkoku-Kan.Panda.COM> Date: Mon, 17 Jan 1994 05:14:42 GMT Mark Crispin <mrc@Ikkoku-Kan.Panda.COM> writes: >I am on the trail of what seems to be an honest-to-goodness bug that caused >the filesystem clobberage. The problem is in drawing up a reliable scenario >to reproduce the problem. And the pain of reproducing the scenario, if you manage it. :) >Things to watch out for: > 1) doing an upgrade instead of a full install from scratch > 2) usage of the login window's restart button instead of double-command-` > and halt > 3) usage of the brightness keys while the system is starting (and before the > window manager has run) Just to inject my own 2 cents regarding the substance of this thread (and trying to side-step the flammage): I've done 1) and lived to tell about it, can't say I've done 2) or 3). However: I was peacefully running a fairly stock black slab with 3.0 (no kernel servers, in particular) when my external 1.2GB drive flamed out sort of like Mark described, only more so: scores of media errors and bad sectors, crashes, kernel panics, etc. This built up gradually over about 5 days until my system was totally unbootable, unfsckable, and unusable. The drive appeared to be physically toasted, as I was unable to get it back even by trying a low-level format. I concluded the drive itself had failed, rather than any software on it, so I sent it back to the vendor for replacement under warranty. Several weeks later :(, I now have a new drive and am just wrapping up a scratch install of NS 3.2. I have so far encountered no difficulties except for a few blown apps recovered from the semi-fragged filesystems I backed up on a DAT drive I bought following hard disk death (call me stupid/lazy, but I'm lucky :). The moral of my story: confirmation of NS 3.2 stability after a scratch install (touch wood), and a recommendation to make sure your disk is physically okay. Sidebar: while my main drive was off being replaced, I ran a much- reduced system off my internal 105MB drive, which had 3.2 on it, upgraded from 3.1, upgraded from 3.0, upgraded from 2.1. No problems to report after several weeks of use. Royce Howland dhowland@gpu.srv.ualberta.ca (wife's account)
Newsgroups: comp.sys.next.bugs From: wave@media.mit.edu (Michael B. Johnson) Subject: Re: 3.2 crashes Message-ID: <1994Jan17.074301.19012@news.media.mit.edu> Sender: news@news.media.mit.edu (USENET News System) Organization: MIT Media Laboratory References: <CJp6o2.ArD@nextsrv1.andi.org> <CJr04p.1uJ@dsinc!flash> Date: Mon, 17 Jan 1994 07:43:01 GMT In article <CJr04p.1uJ@dsinc!flash> flash!jon@dsi.com writes: >>In article <CJp6o2.ArD@nextsrv1.andi.org> c_avery@rsgreet.andi.org (Christopher >>Avery) writes: >>> I also have not been able to read a macintosh formatted floppy under 3.2. >>> Since I did not try to read such a floppy under 3.0 or 3.1, I can't say >>> whether this is a recent problem or not. But the disc read fine on a >>> Macintosh. >> >>Hm. Personally, I use Mac disks exclusively with my 3.2 Cube, and haven't had >>any trouble. Heck, my cube formats Mac disks better than most Macs. >> >>It's probably the disk. >> Well, if he tried more than one disk, it's probably not the disk. If you ever ran an old copy of Executor and it crashed badly, it might have moved your /usr/filesystems/mac.fs somewhere so that it could catch disk insertions. You might sniff around there... I believe current versions of Executor don't ever mess up on this anymore. -- --> Michael B. Johnson -- wave@media.mit.edu --> MIT Media Lab -- Computer Graphics & Animation Group --> 20 Ames St. E15-023G -- (617) 547-0563 (day office) --> Cambridge, MA 02139 -- (617) 253-0663 (night office)
From: sanguish@digifix.com Newsgroups: comp.sys.next.announce,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.bugs,comp.sys.next.programmer,comp.sys.next.marketplace Subject: Quick Guide to NEXTSTEP information on the Internet Date: 17 Jan 1994 02:04:20 -0500 Organization: Next Announcements Message-ID: <2hdd9k$kid@digifix.digifix.com> This post is made weekly, to help 'point' users to more NEXTSTEP information Topics include: comp.sys.next newsgroups related newsgroups comp.sys.next newsgroups mailing list ftp sites NeXTanswers comp.sys.next.* newsgroups -------------------------- Comp.Sys.Next.Advocacy This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. Comp.Sys.Next.Announce Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) The NEXTSTEP FAQs are posted here monthly as well. This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Comp.Sys.Next.Bugs A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT- specific groups as well. Comp.Sys.Next.Hardware Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. Comp.Sys.Next.Marketplace NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. Comp.Sys.Next.Misc For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! Comp.Sys.Next.Programmer Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. The NEXTSTEP programmer FAQs are posted here. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. Comp.Sys.Next.Software This is a place to talk about [third party] software products that run on NEXTSTEP systems. Comp.Sys.Next.Sysadmin Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. related Newsgroups ------------------ Comp.Soft-Sys.Nextstep Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. Comp.Lang.Objective-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. Comp.Object Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original Comp.Sys.Next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News ------------------------------------------- Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep-d next-advocacy-d next-announce-d next-bugs-d next-hardware-d next-marketplace-d next-misc-d next-programmer-d next-software-d next-sysadmin-d (For a full description, send mail saying LISTS to <digestif@antigone.com>). The subscription syntax is essentially the same as LISTSERV's. To subscribe, send a message to <digestif@antigone.com> saying: SUB Listname YourName Example: SUB next-hardware-d John Doe The ftp sites ------------- cs.orst.edu: The main site for North American submissions nova.cc.purdue.edu: Lots of older stuff, but very short on disk space ftp.informatik.uni-muenchen.de: In Germany. ftp.next.com: See the below ftp.next.com and NextAnswers@next.com ------------------------------------- From the document 1000_Help from ftp.next.com Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, or you can transfer them by anonymous ftp. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to NeXTanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO <your-address> If you have any problem with the system or suggestions for improvement, please send mail to NeXTanswers-request@NeXT.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to NeXTanswers-request@NeXT.com. Written by: Eric P. Scott eps@toaster.SFSU.EDU and Scott Anguish sanguish@digifix.com Additions from: Greg Anderson (Greg_Anderson@afs.com) and Michael Pizolato (Michael_Pizolato@afs.com)
From: dillon@apollo.west.oic.com (Matthew Dillon) Newsgroups: comp.sys.next.bugs Subject: Re: Parallel port broken in 3.2 NS/I? Date: 16 Jan 1994 14:57:02 -0800 Organization: Obvious Implementations Corp Distribution: world Message-ID: <2hcgnu$14h@apollo.west.oic.com> References: <1993Dec24.052014.261@corona.com> <1994Jan16.041005.406@corona.com> <2ham2h$oi2@news.acns.nwu.edu> In article <2ham2h$oi2@news.acns.nwu.edu> jweiss@casbah.acns.nwu.edu (Jerry Weiss) writes: :In article <1994Jan16.041005.406@corona.com>, :Peter Kron <pkron@corona.com> wrote: :>I wrote: :>> Since 3.1 came out in May, I have been reliably printing :>> to my HP DeskJet 550C using the configuration available :>> on the archives and Ghostscript 2.6. Since upgrading to :>> 3.2, however, printing is barely usable. If I'm lucky I :>> can print a full page without some character getting :>> dropped, breaking the data protocol and causing the :>> printer to spew out pages of smileys and hearts in the :>> finest DOS fashion. :>> :>> <etc> :> :>NextAnswers-1330 mentions that the parallel port driver does not :>support bidirectional data transfer in 3.1. I previously assumed that :... PC parallel ports have a data-direction register, you can actually *read* data from the data lines if you like. The bidirectionality feature of the parallel port is usually used for file transfer between two PC's or as a network link (under Linux, for example). The nominal handshaking of a printer port has NOTHING to do with its bidirectionality. There are no timing considerations other then writing to the IO and waiting for a flag that indicates acknowledgement. Parallel ports do not work like serial ports, EVERY byte is handshaked. It is not possible to overrun a printer by sending data two quickly. If NeXT is loosing data to a printer over a parallel port, it is a bug with NeXTStep. -Matt Matthew Dillon dillon@apollo.west.oic.com 1005 Apollo Way Incline Village, NV. 89451 ham: KC6LVW (no mail drop) USA Sandel-Avery Engineering (702)831-8000 [always include a portion of the original email in any response!] -- Matthew Dillon dillon@apollo.west.oic.com 1005 Apollo Way Incline Village, NV. 89451 ham: KC6LVW (no mail drop) USA Sandel-Avery Engineering (702)831-8000 [always include a portion of the original email in any response!]
From: rfi@jeans.fokus.gmd.de (Robert Fischer) Newsgroups: comp.sys.next.bugs Subject: Librarian crashes WM on double-head Message-ID: <1994Jan17.142412.2373@fokus.gmd.de> Date: 17 Jan 94 14:24:12 GMT Sender: news@fokus.gmd.de (News system) Organization: GMD-Fokus Yes, it's true! I can reproduce this fatal behaviour. Try this: - ( Quit all apps except of WM. Does not solve the problem! ) - Launch Librarian - Drag the librarian window to the second (b&w) screen - Do some searching - Open some references - The Librarian opens up the windows on the second screen. Almost the second window reaches over to the first (login) screen. - While opening the third or fourth or ... reference, no more interaction is possible. The cursor is spinning, but the clock is out of order. WHAT'S THAT? Robert. -- ------------------------------------------------------------------------ Robert Fischer @ GMD-Fokus -------- __o ------- _`\<,_ fischer@fokus.gmd.de ------- (*)/ (*) ## NeXT-Mail welcome ## ------------------------------------------------------------------------
Newsgroups: comp.sys.next.bugs From: joe@FreemanSoft.com (Joe Freeman) Subject: Re: hostname messed up in 3.2 during boot sequence Message-ID: <1994Jan17.142956.3713@FreemanSoft.com> Sender: jfreeman@FreemanSoft.com Organization: FreemanSoft Inc. References: <2gvfnb$60b@vixen.cso.uiuc.edu> Distribution: usa Date: Mon, 17 Jan 1994 14:29:56 GMT I have the same problem. Has anyone else seen this of figured out why. Our build server has this problem as do all the machines built from it. In article <2gvfnb$60b@vixen.cso.uiuc.edu> lemson@ux1.cso.uiuc.edu (David Lemson) writes: > > I am upgrading many machines to 3.2 and it appears that the lines: > if [ "$SETHOSTNAME" = -YES- ]; then > if [ -z "$HOSTNAME" ]; then > HOSTNAME=-AUTOMATIC- > fi > if [ "$HOSTNAME" != -NO- ]; then > echo "Setting hostname to $HOSTNAME" > docmd hostname $HOSTNAME > fi > fi > aren't doing their job in /etc/rc.net. I see "Setting hostname to > -AUTOMATIC-" as I should, but for some reason, 'docmd hostname > -AUTOMATIC-' isn't setting the hostname as it should at that time > (it stays 'localhost' as set in rc.boot I guess) > If I run 'hostname -AUTOMATIC-' from the command line as root, it > sets the hostname as it should (BOOTP). Does this work during the boot > process for anyone else? I am considering putting a "hostname > -AUTOMATIC-" in the /etc/rc.local as a workaround. This is far from > optimal, as syslogd reports 'localhost' as the hostname until it > gets around to the /etc/rc.local at the end. > > Does anyone have any clue what needs to run before hostname responds > correctly by sending a BOOTP request? > > The BOOTP servers are running 3.2 as well. > > -- > David Lemson (217) 244-1205 > University of Illinois Computing & Comm Services Office System Administrator > Internet : lemson@uiuc.edu UUCP :...!uiucuxc!uiucux1!lemson > NeXTMail & MIME accepted BITNET : LEMSON@UIUCVMD > > -- > David Lemson (217) 244-1205 > University of Illinois Computing & Comm Services Office System Administrator > Internet : lemson@uiuc.edu UUCP :...!uiucuxc!uiucux1!lemson > NeXTMail & MIME accepted BITNET : LEMSON@UIUCVMD -- Joe Freeman FreemanSoft Inc. A NEXTSTEP software and consulting services company. Electronic Mail: Joe@FreemanSoft.com (NeXT Mail) Voice: 919.783.7033
Newsgroups: comp.sys.next.bugs From: thf@zelator.in-berlin.de (Thomas Funke) Subject: Re: lovely 3.2 (jumping to conclusions?) Message-ID: <1994Jan16.130122.789@gamelan> Sender: thomas@gamelan (thomas) Organization: Disorganization References: <1994Jan15.015153.2575@cs.yale.edu> Date: Sun, 16 Jan 1994 13:01:22 GMT NS3.2 on a 040-Cube seems to be very stable and also faster than previous releases. Still the old bugs remain: No shrinking swapfiles, crashing OS when running Sybase in the standard installation (which has an easy work around). Annoying: Bug in Distributed Objects not freeing string space. Besides of that: I'm very content. Good and fast OS. -- ------------------------------------------------------------------ Thomas Funke ** Unix-Consultant ** thf@zelator.in-berlin.de Brooks's Law: Adding manpower to a late software project makes it later ------------------------------------------------------------------
From: sfk@dannug.dk (Steen Kroyer) Newsgroups: comp.sys.next.bugs Subject: A Freudian Slip? Date: 17 Jan 1994 19:03:41 GMT Organization: Danish NeXT User Group Distribution: world Message-ID: <2hened$k03@aho.dk> References: <MS-C.758674206.662824084.mrc@Ikkoku-Kan.Panda.COM> Keywords: Ouupss! In article <MS-C.758674206.662824084.mrc@Ikkoku-Kan.Panda.COM> Mark Crispin <mrc@Ikkoku-Kan.Panda.COM> writes: > Alex Currier - > > Have you thought about the consequences of what you state? > > Suppose I did give up on using the NEXTSTEP operating system. Then my > comment would be that I had to give up on NEXTSTEP because I was unusable. ^ ??? Come now Mark, surely that is being too modest ..... (;-) -- @-------------------------------@------------------------------@ | Steen Kroyer | | | Email/NeXTMail: sfk@dannug.dk | * This Space For Rent * | | Fax (home) : +45 98574459 | | @-------------------------------@------------------------------@
From: clarence@orion.cc.andrews.edu (Clarence L. Thomas IV) Newsgroups: comp.sys.next.bugs Subject: Global Alert For All: Jesus is Coming Soon Date: 17 Jan 1994 19:43:05 -0500 Organization: Andrews University, Berrien Springs, MI, 49104 Distribution: world Message-ID: <2hfbap$9fm@orion.cc.andrews.edu> The earthquake in Los Angeles, California, the flood in Europe, the seemingly unstoppable war in the former Yugoslavia, the devastating fires in Australia, the flood in the Midwest of the United States of America, the devastating fires near Los Angeles, California, the rapid and appalling increase in violence in cities, towns, villages all over the world, the famines, the diseases, the rapid decline of the family unit, and the destructive earthquake in India (in 1993) are signs that this world's history is coming to a climax. The human race has trampled on God's Constitution, as given in Exodus 20:1-17 (King James Version Bible), and Jesus is coming to set things right. These rapidly accelerating signs are an indication that Jesus is coming soon (Matthew 24). God's Holy Spirit is gradually withdrawing its protection from the earth and the devastating events you see are demonstrations of Satan's power. All those who are not guarded by God are in danger of forever losing eternal life. If you want to know what's about to happen, please study the books of Daniel and Revelation which are located in God's Word, the Bible. They are not sealed or closed books. They can and must be understood by all. Every word in the Bible from Genesis to Revelation is true. The Bible and the Bible only must be your guide. When God's Law (the Constitution for the Universe) is consistently ignored, disregarded, changed, and questioned, He permits certain events to occur to wake us up. I would urge all, wherever you are and regardless of the circumstances, to directly call on Jesus and ask Him to intervene in your life. Jesus who created this planet and every living creature in it and on it, died on the cross, was raised from the dead by God the Father, and is now in Heaven interceding for you. Jesus is the only One who can rescue us from the slavery, misery, and death Satan is causing us. For reference I'm including God's Constitution as given in the King James Version Bible. Please note that when God says the seventh day, he means Sabbath (the 7th day of the week) not Sunday (1st day of the week). Commandment #1: Exodus 20:1-3 (KJV) And God spake all these words, saying, I am the LORD thy God, which have brought thee out of the land of Egypt, out of the house of bondage. Thou shalt have no other gods before me. Commandment #2: Exodus 20:4-6 (KJV) Thou shalt not make unto thee any graven image, or any likeness of any thing that is in heaven above, or that is in the earth beneath, or that is in the water under the earth. And shewing mercy unto thousands of them that love me, and keep my commandments. Commandment #3: Exodus 20:7 (KJV) Thou shalt not take the name of the LORD thy God in vain; for the LORD will not hold him guiltless that taketh his name in vain. Commandment #4: Exodus 20:8-11 (KJV) Remember the sabbath day, to keep it holy. Six days shalt thou labour, and do all thy work: But the seventh day is the sabbath of the LORD thy God: in it thou shalt not do any work, thou, nor thy son, nor thy daughter, thy manservant, nor thy maidservant, nor thy cattle, nor thy stranger that is within thy gates: For in six days the LORD made heaven and earth, the sea, and all that in them is, and rested the seventh day: wherefore the LORD blessed the sabbath day, and hallowed it. Commandment #5: Exodus 20:12 (KJV) Honour thy father and thy mother: that thy days may be long upon the land which the LORD thy God giveth thee. Commandment #6: Exodus 20:13 (KJV) Thou shalt not kill. Commandment #7: Exodus 20:14 (KJV) Thou shalt not commit adultery. Commandment #8: Exodus 20:15 (KJV) Thou shalt not steal. Commandment #9: Exodus 20:16 (KJV) Thou shalt not bear false witness against thy neighbour. Commandment #10: Exodus 20:17 (KJV) Thou shalt not covet thy neighbour's house, thou shalt not covet thy neighbour's wife, nor his manservant, nor his maidservant, nor his ox, nor his ass, nor any thing that is thy neighbour's. I also recommend that the following books be obtained and closely studied: The Great Controversy By Ellen G. White Review and Herald Publishing Association Hagerstown, MD 21740 The Desire of the Ages By Ellen G. White Review and Herald Publishing Association Hagerstown, MD 21740 Patriarchs and Prophets By Ellen G. White Review and Hearld Publishing Association Hagerstown, MD 21740 Daniel and the Revelation By Uriah Smith Review and Herald Publishing Association Hagerstown, MD 21740 ------- Clarence L. Thomas IV Phone: 616-471-6116 E-mail: thomas@redwood.cc.andrews.edu
From: gad@eclipse.its.rpi.edu (Garance A. Drosehn) Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs Subject: Re: Librarian bug? Man page topics don't show up in *NS-3.2* Date: 18 Jan 1994 04:23:56 GMT Organization: Rensselaer Polytechnic Institute, Troy NY, USA Distribution: world Message-ID: <2hfo8s$s9@usenet.rpi.edu> References: <2h0ka7$mte@zip.eecs.umich.edu> Roland Telfeyan <roland@gomidas.mi.org> writes: > I've noticed that in 3.1 the man page topics don't show up in > Librarian. All you get are things like "man3/ctime.3" rather > than "ctime - convert date and time to ASCII" or whatever. > > Anybody know what causes this? I've tried reindexing using ixbuild > but it doesn't work. I noticed that in 3.2 some of the man page > topics DO show up but many still DON'T. What gives? I have no insight on NS-3.1, but the following is significant for people with NS-3.2. Or at least, it was significant for me, running NS-3.2 on NeXT (not Intel) hardware: In the Supplemental Release Notes for NeXTSTEP Developer, which were included with release 3.2 developer, the second page has a section on Unix man pages. This is tagged as "Reference 38816". What is listed there as the workaround for the problem it describes is wrong. Anyone who followed that advice will (among other things) end up without descriptions for the man pages when they list titles in Digital Librarian. What it recommends: cd /usr/man mv man/* man1 rmdir man rm whatis .index.store ixbuild -fsv -LEnglish . makewhatis . If you have not already done the workaround that NeXT recommended, then what you really want to do is: cd /usr/man mv man/*/* man1 rmdir man/man1 man rm man2/open.2.deleteme mv man8/rdate.8C man8/rdate.8c rm whatis .index.store ixbuild -fsg -LEnglish . /usr/lib/makewhatis . chmod go+r .index.store whatis if you have done the recommended workaround, then you want to: cd /usr/man mv man1/man1/* man1 rmdir man1/man1 rm man2/open.2.deleteme mv man8/rdate.8C man8/rdate.8c rm whatis .index.store ixbuild -fsg -LEnglish . /usr/lib/makewhatis . chmod go+r .index.store whatis (in all cases, you have to log into root to do these unix commands. You'd have to start up Terminal.app to get a window where you can type the unix commands into...). Comments on the differences: the release notes describe the error as some man pages being in /usr/man/man, but they're really in /usr/man/man/man1 The release notes didn't catch the "deleteme" file that Annard Brouwer (annard@theborg.stack.urc.tue.nl) noticed. Annard also noticed that the ".8C" suffix on rdate had to be ".8c". In the ixbuild, you want the "-g" parameter (which gets you the one-line man page descriptions when Digital Librarian lists the titles of man pages). You have no need for the -v option, which just prints hundreds of lines that you don't care about (in this case) to your terminal window. Thanks to Malcolm Crawford (M.Crawford@dcs.shef.ac.uk) for spotting this, I missed it the first times I rebuilt the index. makewhatis is in /usr/lib, and /usr/lib is not in the $PATH for most people (it's not in the default PATH for root). once the files are created, you want to make sure all users can read them! The way my own umask is set, they were not permitted to others by default. That might be due to a change I made, I don't remember now... --- Garance Alistair Drosehn = gad@eclipse.its.rpi.edu ITS Systems Programmer (handles NeXT-type mail) Rensselaer Polytechnic Institute; Troy NY USA
Control: cancel <2hfbap$9fm@orion.cc.andrews.edu> Newsgroups: comp.sys.next.bugs From: clarence@orion.cc.andrews.edu Subject: cmsg cancel <2hfbap$9fm@orion.cc.andrews.edu> Message-ID: <cancel$758899248@aber.ac.uk> Sender: news@aber.ac.uk (USENET news service) Organization: UseNet Cabal (West Wales Cadre) Active Service Unit Date: Tue, 18 Jan 1994 13:20:48 GMT Autocancel of: "Global Alert For All: Jesus is Coming Soon"
Newsgroups: comp.sys.next.bugs From: jimbo@oingo.umn.edu Subject: Re: 3.2 crashes Message-ID: <CJuy2t.10L@news2.cis.umn.edu> Sender: news@news2.cis.umn.edu (Usenet News Administration) Organization: University of Minnesota, Twin Cities References: <2hae9f$m1l@geraldo.cc.utexas.edu> Date: Wed, 19 Jan 1994 03:30:34 GMT > if it isn't working > for you maybe you *should* just use something else. Pretty good advice, I > think. > What he said!!
From: bkr@drdhh.hanse.de (Bjoern Kriews) Newsgroups: comp.sys.next.bugs Subject: Re: 3.2 crashes Date: 18 Jan 1994 09:07:36 +0100 Organization: Digital Island Message-ID: <bkr.758880147@drdhh> References: <MS-C.758669225.1103527590.mrc@Ikkoku-Kan.Panda.COM> Mark Crispin <mrc@Ikkoku-Kan.Panda.COM> writes: >Things to watch out for: > 1) doing an upgrade instead of a full install from scratch I did. > 2) usage of the login window's restart button instead of double-command-` > and halt I often do. > 3) usage of the brightness keys while the system is starting (and before the > window manager has run) My system once crashed when I hit Brightness-Keys at this time. Tip - and panic. >Transsys SLIP is still under investigation as a possible factor, but there has >been no evidence to date blame Mamakos. No transsys slip with me - just free ppp 0.3, non active at this time. >As for the weenie comments (e.g. ``karma'', or ``get a new OS''), such as >excellent examples of why many vendors don't want to touch the NeXT community >with a 10-foot pole. Such comments make the entire NeXT community look bad >and only serve to validate the image of NEXTSTEP as being a fringe system. I support that. Sure, it's not nice to talk down ones baby, but I hate a system that panics every now and then - what I want is a NeXTSTEP that runs with the reliablity of a well patched SUNOS 4.1.3. I want to be able to look people straight in the eye, saying: Use NS, it's superior and reliable. I also understand Marks comments - I know my own annoyed face when I come home and just see some small rectangle on the screen - wishing it said DON'T PANIC in large friendly letters - but it does not. Let's kick NeXT to make it an even better product. Regards, Bjoern -- bkr@drdhh.hanse.de - Bjoern Kriews - Stormsweg 6 - D-22085 Hamburg [76] - FRG "Insisting on perfect safety is for people who don't have the balls to live in the real world." -- Mary Shafer, NASA Ames Dryden
From: bkr@drdhh.hanse.de (Bjoern Kriews) Newsgroups: comp.sys.next.bugs Subject: Re: 3.2 crashes Date: 18 Jan 1994 09:10:57 +0100 Organization: Digital Island Message-ID: <bkr.758880527@drdhh> References: <MS-C.758669225.1103527590.mrc@Ikkoku-Kan.Panda.COM> <2h9pui$d0k@geraldo.cc.utexas.edu> mycroft@colourbox.utexas.edu (Alex Currier) writes: >This is not an evidence of the low quality of the NeXT community or the >fringeness of NeXTStep, it is an evidence that the NeXT community (such as it >is here) is getting bored of hearing you go on about NeXTStep and some people >feel you're just beating your own personal dead horse. I WANT him around because he investigates kernel bugs instead of ignoring them. We can't help if some of you only use your machine to do wordprocessing or whatever - that's what these machines are built for. But please accept that other people want Unix when they buy Unix and complain if they get crashes instead of Unix. Unix is a registered bell etc... -- bkr@drdhh.hanse.de - Bjoern Kriews - Stormsweg 6 - D-22085 Hamburg [76] - FRG "Insisting on perfect safety is for people who don't have the balls to live in the real world." -- Mary Shafer, NASA Ames Dryden
Newsgroups: comp.sys.next.bugs From: cowboy@omega.csuohio.edu (Joe Rosenfeld) Subject: Re: 3.2 crashes Message-ID: <1994Jan19.115635.12255@news.csuohio.edu> Sender: news@news.csuohio.edu (USENET News System) Organization: Cleveland State University References: <MS-C.758669225.1103527590.mrc@Ikkoku-Kan.Panda.COM> <2h9pui$d0k@geraldo.cc.utexas.edu> <bkr.758880527@drdhh> Date: Wed, 19 Jan 1994 11:56:35 GMT Bjoern Kriews (bkr@drdhh.hanse.de) wrote: : mycroft@colourbox.utexas.edu (Alex Currier) writes: : >This is not an evidence of the low quality of the NeXT community or the : >fringeness of NeXTStep, it is an evidence that the NeXT community (such as it : >is here) is getting bored of hearing you go on about NeXTStep and some people : >feel you're just beating your own personal dead horse. : I WANT him around because he investigates kernel bugs instead of ignoring : them. I agree. Mark is treated so unfaily by NeXT users that I cannot believe it. His heart is in the right place. He wants NeXT to fulfill its potential. All this Mark Crispin bashing is almost mindless, like mob violence. Give the guy a break. Cowboy
Newsgroups: comp.sys.next.bugs From: fabien@free (Fabien Roy) Subject: Re: 3.2 crashes Message-ID: <1994Jan17.225238.1719@free.fdn.org> Sender: news@free.fdn.org Organization: Fabien Roy Electronic Engineering (F.R.E.E.) - Paris, France. References: <dhowland.758783682@kakwa.ucs.ualberta.ca> Date: Mon, 17 Jan 1994 22:52:38 GMT In article <dhowland.758783682@kakwa.ucs.ualberta.ca> dhowland@gpu.srv.ualberta.ca (Deborah Howland) writes: > Mark Crispin <mrc@Ikkoku-Kan.Panda.COM> writes: > > >I am on the trail of what seems to be an honest-to-goodness bug that caused > >the filesystem clobberage. The problem is in drawing up a reliable scenario > >to reproduce the problem. > > And the pain of reproducing the scenario, if you manage it. :) > > >Things to watch out for: > > 1) doing an upgrade instead of a full install from scratch > > 2) usage of the login window's restart button instead of double-command-` > > and halt > > 3) usage of the brightness keys while the system is starting (and before the > > window manager has run) > > Just to inject my own 2 cents regarding the substance of this thread > (and trying to side-step the flammage): I've done 1) and lived to tell > about it, can't say I've done 2) or 3). > > However: > > I was peacefully running a fairly stock black slab with 3.0 (no kernel > servers, in particular) when my external 1.2GB drive flamed out sort of > like Mark described, only more so: scores of media errors and bad > sectors, crashes, kernel panics, etc. This built up gradually over > about 5 days until my system was totally unbootable, unfsckable, and > unusable. > > The drive appeared to be physically toasted, as I was unable to get it > back even by trying a low-level format. I concluded the drive itself > had failed, rather than any software on it, so I sent it back to the > vendor for replacement under warranty. > > Several weeks later :(, I now have a new drive and am just wrapping up > a scratch install of NS 3.2. I have so far encountered no difficulties > except for a few blown apps recovered from the semi-fragged filesystems > I backed up on a DAT drive I bought following hard disk death (call me > stupid/lazy, but I'm lucky :). > > The moral of my story: confirmation of NS 3.2 stability after a scratch > install (touch wood), and a recommendation to make sure your disk is > physically okay. > > Sidebar: while my main drive was off being replaced, I ran a much- > reduced system off my internal 105MB drive, which had 3.2 on it, > upgraded from 3.1, upgraded from 3.0, upgraded from 2.1. No problems > to report after several weeks of use. > > Royce Howland > dhowland@gpu.srv.ualberta.ca (wife's account) Think it very easy to test the disk for bad reading blocks: as root type dd if=/dev/rsd0a of=/dev/null bs=512k and some time later: 812+1 records in 812+1 records out which says I have a wealthy 400Meg drive. --Fabien ___________________________________________________________________ Fabien Roy fabien@free.fdn.org NeXTmail OK
From: mark@ve6mgs.ampr.ab.ca (Mark Gregory Salyzyn) Newsgroups: comp.sys.next.bugs Subject: Re: Parallel port broken in 3.2 NS/I? Message-ID: <CJu6rH.4B6@ve6mgs.ampr.ab.ca> Date: Tue, 18 Jan 1994 17:47:39 GMT References: <1993Dec24.052014.261@corona.com> <1994Jan16.041005.406@corona.com> <2ham2h$oi2@news.acns.nwu.edu> <2hcgnu$14h@apollo.west.oic.com> Organization: VE6MGS Gateway dillon@apollo.west.oic.com (Matthew Dillon) writes: > PC parallel ports have a data-direction register, you can actually > *read* data from the data lines if you like. The bidirectionality > feature of the parallel port is usually used for file transfer between > two PC's or as a network link (under Linux, for example). You have to *find* these parallel ports. The original PC had an 8255 based parallel port for communicating with the printer. Later, in the 256K motherboard (yipes), they changed it to a set of '374 latches fixed for the purpose. Since then, you find ALL printer ports, minus ones that advertise otherwise, they are one direction only. The Linux network link, I suspect, uses the handshaking lines (4 bits) to transfer data *into* the port. LapLink certainly does if it finds the ports are not the 8255 based PIO. > The nominal handshaking of a printer port has NOTHING to do with > its bidirectionality. There are no timing considerations other then > writing to the IO and waiting for a flag that indicates acknowledgement. > Parallel ports do not work like serial ports, EVERY byte is handshaked. > It is not possible to overrun a printer by sending data two quickly. But it is, if the software times out on the flag to *get on with it*. But I do agree wholeheartedly that this is a *bug* with NeXTSTEP (I only notice losing characters on 3.1, have not yet on 3.2)! I suspect the problem is they are interrupting on the handshake signals and IRQ7 on the 8259A is a *garbage* interrupt line that catches interrupts from other devices that flash the interrupt line for too short of a period (line goes active, then back inactive *before* the IACK sequence is started). This assumption that `if I get an IRQ7, it *must* be from the printer acknowledge' must have been made by their programmers and thus the problem with *overrunning* the printer. A simple check of either the 8259, or the acknowledgment signal will solve this problem in their driver. Ciao -- Mark
From: feng@jedi.eng.uci.edu (Feng Liu) Newsgroups: comp.sys.next.bugs Subject: Re: Parallel port broken in 3.2 NS/I? Date: 19 Jan 1994 17:58:23 GMT Organization: University of California, Irvine Message-ID: <2hjsbv$9jd@news.service.uci.edu> References: <CJu6rH.4B6@ve6mgs.ampr.ab.ca> In article <CJu6rH.4B6@ve6mgs.ampr.ab.ca> mark@ve6mgs.ampr.ab.ca (Mark Gregory Salyzyn) writes: > dillon@apollo.west.oic.com (Matthew Dillon) writes: > > > PC parallel ports have a data-direction register, you can actually > I suspect the problem is they are interrupting on the handshake signals > and IRQ7 on the 8259A is a *garbage* interrupt line that catches interrupts > from other devices that flash the interrupt line for too short of a period > (line goes active, then back inactive *before* the IACK sequence is started). > This assumption that `if I get an IRQ7, it *must* be from the printer > acknowledge' must have been made by their programmers and thus the problem > with *overrunning* the printer. A simple check of either the 8259, or the > acknowledgment signal will solve this problem in their driver. > > Ciao -- Mark Is what you were discussing in any way relevant to the following problem that I have with parallel printer and PNI 1.6? Intel NS 3.2 Parallel port connected to HP IIIP Postscript Printer. Printer works well under any app, say TeXView, or directe lpr. Problem 1: once PNI 1.6 is loaded, if I click PRINT say under TeXView, it seems to take longer to get the print panel. Problem 2: If I terminate PNI 1.6 in any way (say, by using pnistat, or kill, or due to modem drop DTR), then when I click PRINT under TeXView or any other app, the spinning disk appears on the screen but the print panel never comes up. In fact, I can not even power off the machine if I do select logout and power-off, the app is still running with the spinning disk trying to get the print panel(of course I could just shut the power). I cannot print anything by lpr either. My print job just stays in the queue. Tryed various things, such as reset the printer, switch on and off the printer, clear the print queue, did not work. I had to use Process Moniter to kill TeXview, and then reboot the machine, then everything is fine again until you start and terminate PNI. PNI uses one serial port, how could it mess up the parallel port? I have an internal modem set to serial port 2, PNI set up to use cufb. Did anybody have the same problem? Any help or info is well come. Thank you. Feng Liu UC, Irvine
From: Ward_Travis@transarc.com Newsgroups: comp.sys.next.bugs Subject: ?: div-0 trap in kernel Date: Wed, 19 Jan 1994 13:10:55 -0500 Organization: Carnegie Mellon, Pittsburgh, PA Message-ID: <whDLSj70Bww8ALX_o8@transarc.com> Please let me know if this panic is to be expected (a known bug). If you can lend any advice on how to work around it, I'd be grateful. Configuration: 68030 Cube, NS 3.0, 12Mb RAM (mixed non-parity and parity ...) Problem occurs: when I try to BuildDisk onto the OD. The BuildDisk is being run with -NXHost <othersystem> The fault does not occur until after the build is started. The panic: Jan 8 23:39:23 rubik mach: called from pc 04073cc8 fp 10f4cf7c 4-args 0409920\ d 00000001 0000003c 00000008 Jan 8 23:39:23 rubik mach: called from pc 0401341e fp 10f4cfb4 4-args 0000003\ c 00000008 0409920d 040992d2 Jan 8 23:39:23 rubik mach: called from pc 04013262 fp 10f4cff4 4-args 0409920\ d 040992d2 00000000 00000000 Jan 8 23:39:23 rubik mach: called from pc 0401317a fp 10f4cffc 4-args 0000000\ 0 040a27e0 040a27e0 00000002 Jan 8 23:39:23 rubik mach: called from pc 040a27e0 fp 00000000 4-args 040a27e\ 0 00000002 00002304 80000000 Jan 8 23:39:23 rubik mach: last fp 0x0 Jan 8 23:39:23 rubik mach: panic: (Cpu 0) zero divide Thanks.. for ANY advice.. -- Ward C. Travis Pittsburgh PA USA "The fact is the sweetest dream that travis@transarc.com (412) 338 4388 labor knows." - R. Frost
From: blake015@mc.duke.edu(Denise Blakeley) Newsgroups: comp.sys.next.software,comp.sys.next.bugs Subject: AcChen crashing under NS 3.2 Moto Date: 19 Jan 1994 21:51:32 GMT Organization: Duke University; Durham, N.C., USA Message-ID: <2hka14$sel@news.duke.edu> I recently upgraded my standalone Turbo Color from NS 3.1 to 3.2. I didn't notice if this started happening right away or later, but now AcChen crashes right away upon launch. It gets as far as putting up the "Welcome to AcChen" window, and almost as fast as you can blink it all disappears. The "WELCOME!" sound is not played. I tried re-installing AcChen, but this made no difference. Can anyone tell me what's wrong before I go into severe withdrawal? Does AcChen not run under 3.2? Denise (AcChen Addict) Blakeley -- Denise Blakeley | PROGRAM, tr. v., An activity similar Duke Med Center Info Systems | to banging one's head against a wall, Durham, NC | but with fewer opportunities for (919) 286-6468 W | reward. blake015@mc.duke.edu | NeXTMail welcome!
From: lee@daruma.zoo.duke.edu (Lee Altenberg) Newsgroups: comp.sys.next.bugs Subject: NS/I 3.2 "RESTART" button broken Date: 20 Jan 1994 00:50:29 GMT Organization: Duke University Zoology Dept.; Durham, N.C. Distribution: world Message-ID: <2hkkgl$2l1@news.duke.edu> After upgrading a Gateway2000 4DX2/66V from NS/I 3.1 to 3.2, the login window's "RESTART" button fails to reboot the system. It starts the sequence, replacing the login window with the reboot monitor, but then the reboot monitor disappears and the login monitor comes back. I have to use Alt-NumLock to get the little NMI-like monitor to reboot. What is wrong and how do I fix it? Thanks, Lee Altenberg altenber@acpub.duke.edu
Newsgroups: comp.sys.next.bugs From: mark@ve6mgs.ampr.org (Mark G. Salyzyn) Subject: Parallel Drivers and IRQ7 (Re: Serial Drivers) Organization: ADEC Systems Inc. Distribution: na Date: Tue, 18 Jan 94 21:00:42 GMT Message-ID: <1994Jan18.210042.8682@ve6mgs.ampr.org> References: <1994Jan13.144903.346@corona.com> <2h6d2h$jrb@digdug.pencom.com> robin@pencom.com (Robin D. Wilson) writes: >This just sounds like bugs in the driver. How can your disk activity cause >the buffer on the printer to overflow? Technically, increased disk activity >should slow your transmission down, reducing the probability of buffer >overflow (due to a lower transmission rate). No, if the IRQ system is more busy, as it will be at peak disk activities, the chances of a `False IRQ' being handed to IRQ7 on the 8259 increase. I suspect that there is a bug in the handling of `False IRQ's from IRQ7 in the system. This may be a problem with more than just the parallel driver, as I know people that forgo the parallel port and use the Interrupt to run the sound system, bus mouse or some other hardware. Ciao -- Mark
From: mark@ve6mgs.ampr.ab.ca (Mark Gregory Salyzyn) Newsgroups: comp.sys.next.bugs Subject: Known Mux (Serial) Driver Bugs and Problems Message-ID: <CJv5uH.9HL@ve6mgs.ampr.ab.ca> Date: Wed, 19 Jan 1994 06:25:28 GMT Organization: VE6MGS Gateway The following problems have come to my attention for the updated Serial Driver: 1) The serial Bus Mouse Configuration program will not recognise that the Mux driver can replace the SerialPorts driver and will complain (or crash Configure.app). Since I do not have a serial mouse on my system, I will need to regroup to figure out how to test this and a workaround. (Suggestions, reduce the number of supported ports in SerialPorts to 1, and use the "Path #" Configuration parameters in the Instance?.table files under Mux to start placing ports at ttyb and so on. This requires *good* understanding of the Instance?.tables and editing, not for the weak or the Weary!). If anyone has dealt with this issue, please send me a note of their configuration as some people are interested. 2) My documentation on how to use the Mux program for each instance of a serial port is a bit weak. The concept of one port for each instance of the driver is a foreign concept for people that have gotten used to the two serial ports for the one instance of the NeXT supplied SerialPorts driver. Workaround? well, if I wrote the configure program, this *could* make more sense ... (time!!!) 3) I have noticed once in a while the driver will seem to act as if the receive interrupts are disabled on the serial ports. The only reason it works at all when this happens is because of a RLS (overun or parity) interrupt occurs and the port is scanned at that point. Throughput drops to 860ch/s rather than 1500ch/s when this happens. I have been trying desparately to find out why, but it's rare occurence has prevented me from getting to the root of it. I suspect it is because of my DigiBoard8, as I have not noticed this (yet) on any of the regular single interrupt per a port type boards. In any case, the driver STILL works better than the native serialPorts driver for uucp operation (tips and gettys may make little sense though :-( ) ... 4) There is an error in the localizable.strings file, Mux8 should be replaced with Mux. This is only a menu labelling issue. 5) when you run `make install', as root, the driver permissions may be inappropriate for regular users to inspect the configuration. Workaround: change the permissions manually to read for all, or be root when doing Configure.app ... 6) Manually editing the Instance?.table files is not for the casual observer. I *realy* need more time to write a suitable configuration inspector application to tidy this thing up ... I will gladly help those that desparatly need a fast serial port running if they need advice, patience may be the order of the day (especially if they are running a serial mouse and only have two ports on their system) ... I hope this gives some hope for the weary. Thanks for all the support I have received over this driver, I do hope that the patience continues since I only do writing for this driver during my spare time in the evenings after the kids are asleep (and my wife has stopped grumbling :-) Ciao -- Mark
Newsgroups: comp.sys.next.bugs From: Douglas B. Moore Subject: Re: 3.2 crashes Message-ID: <CJw0A5.6o6@news2.cis.umn.edu> Sender: news@news2.cis.umn.edu (Usenet News Administration) Organization: University of Minnesota, Twin Cities References: <1994Jan19.115635.12255@news.csuohio.edu> Distribution: World Date: Wed, 19 Jan 1994 17:15:44 GMT In article <1994Jan19.115635.12255@news.csuohio.edu> cowboy@omega.csuohio.edu (Joe Rosenfeld) writes: > Bjoern Kriews (bkr@drdhh.hanse.de) wrote: > : mycroft@colourbox.utexas.edu (Alex Currier) writes: > : >This is not an evidence of the low quality of the NeXT community or the > : >fringeness of NeXTStep, it is an evidence that the NeXT community (such as it > : >is here) is getting bored of hearing you go on about NeXTStep and some people > : >feel you're just beating your own personal dead horse. > > : I WANT him around because he investigates kernel bugs instead of ignoring > : them. > > I agree. Mark is treated so unfaily by NeXT users that I cannot > believe it. His heart is in the right place. He wants NeXT to > fulfill its potential. All this Mark Crispin bashing is almost > mindless, like mob violence. > > Give the guy a break. > > Cowboy It would be great if all Mark did was kernel debugging, but in this case there doesn't seem to be a bug for most of us. I've got 3.2 running on both a mono station and an ND turbo. None of the symptoms described are found here in my living room. I also have Marble Teleconnect doing it's kernel modifications along with those used by the ND. Don't tell me this operating system panics at the drop of a hat, I haven't had a panic window since last spring. We are all for getting the most out of our systems. Yes, there are things which can be improved. Let's keep looking for the legitimate bugs and leave the disgruntled mudslinging out of this pursuit. ---- Douglas Moore 172 East Sixth Street #2106 St Paul, MN 55101 612-227-3274 dmoore@maroon.tc.umn.edu <---NeXTMail ready
Newsgroups: comp.sys.next.bugs From: irving@Happy-Man.com (Irving_Wolfe) Subject: /etc/dumpdates (3.2 black) Message-ID: <1994Jan20.000755.3030@Happy-Man.com> Organization: Happy Man Corp, 4410 Pt Robinson, Vashon, WA 98070 206/463-9399 Date: Thu, 20 Jan 1994 00:07:55 GMT A copy of my report, their log 52886. As usual, there has been no acknowledgement. Like most of Next's bugs, even a rudimentary beta-test program (with appropriate responsiveness) would have kept them out of the release. But, then, perfect people don't need to test. (This is not to disparage 3.2, which seems lots better than 3.0 or 3.1.) I think you are shipping /etc/dumpdates as root-wheel-0644. It should be root-operator-0664. Dumps are normally performed by someone in group operator, not by root. (In our case, the someone is a pseudo-user named staffman.) I think the man page for dump has more info on this. It is the main reason that the group operator exists. (Yes, I know there's also some event notification that is sometimes set to contact anyone in group operator who's logged in, but that's not really applicable under Nextstep most of the time.) P.S.-- If there's any truth to Andrew Stone's comments on the net about your having shipped a dangerous version of indexing kit in 3.2, please let me know so I can take temporary steps to protect our databases. Please also send me the few replacement files that are needed a.s.a.p., to have it work safely again. -- Irving_Wolfe@Happy-Man.com 206/463-9399 x101 fax 206/463-9255 Happy Man Corp. 4410 SW Pt. Robinson Rd., Vashon, WA 98070-7399 In SOLID VALUE, we show intelligent investors under-priced stocks Printed Info Free: Send POSTAL address: Solid-Value@Happy-Man.com
Newsgroups: comp.sys.next.bugs From: cedman@princeton.edu (Carl Edman) Subject: Re: Help with dead? mail app 3.0 In-Reply-To: johannes@scribes.english.uiuc.edu's message of 15 Jan 1994 20:38:16 GMT To: johannes@scribes.english.uiuc.edu (John B. Friedman) Message-ID: <CEDMAN.94Jan18150149@capitalist.princeton.edu> Originator: news@nimaster Sender: news@Princeton.EDU (USENET News System) Organization: Princeton University References: <2h9k7o$36g@vixen.cso.uiuc.edu> Date: Tue, 18 Jan 1994 20:01:49 GMT In article <2h9k7o$36g@vixen.cso.uiuc.edu> johannes@scribes.english.uiuc.edu (John B. Friedman) writes: Netters, could I get some advice? I had a freeze the other day when my modem clicked off in midst of transmitting mail and cursor kept whirling. 3.0 Mail was not active then though had been working fine just 10 minutes before. I used Halt, rebooted, got a message there had been a disk segment error now fixed. Clicked on Mail app in dock. It would highlight for a second or two and then quit and take me back to Workspace menu. I tried to pull app out of dock and replace with Mail app from Next apps. Same thing. I logged in as root and replaced the app. I got the mail to open there but there was no mail in it, just "welcome new user" message. The modem is definitely passing mail and seems OK. It is just that the mail menu won't come up and the window open when mail icon clicked . Any thoughts on where to begin, what kind of error message to look for if any would be most appreciated. Thanks very much. John Friedman RTFFAQ. Carl Edman PS: That means, Read The Friendly Frequently Asked Question list almost certainly in your c.s.n.announce news spool.
Newsgroups: comp.sys.next.bugs From: jlemon@netcom.com (Jonathan Lemon) Subject: Re: Known Mux (Serial) Driver Bugs and Problems Message-ID: <jlemonCJwxpn.Dsp@netcom.com> Organization: Netcom - Online Communication Services (408 241-9760 guest) References: <CJv5uH.9HL@ve6mgs.ampr.ab.ca> Date: Thu, 20 Jan 1994 05:24:58 GMT In article <CJv5uH.9HL@ve6mgs.ampr.ab.ca> mark@ve6mgs.ampr.ab.ca (Mark Gregory Salyzyn) writes: >The following problems have come to my attention for the updated Serial Driver: > 1) The serial Bus Mouse Configuration program will not recognise > that the Mux driver can replace the SerialPorts driver and will > complain (or crash Configure.app). Since I do not have a serial > mouse on my system, I will need to regroup to figure out how to > test this and a workaround. (Suggestions, reduce the number of > supported ports in SerialPorts to 1, and use the "Path #" > Configuration parameters in the Instance?.table files under Mux to > start placing ports at ttyb and so on. This requires *good* > understanding of the Instance?.tables and editing, not for the > weak or the Weary!). If anyone has dealt with this issue, please > send me a note of their configuration as some people are interested. I'm currently running NS/I 3.2 with a serial mouse, driven by the SerialMouse + SerialPorts driver, and Mux on my modem line. Here's how: 1) copy the COM1,COM2 configuration files into Mux's top-level default configuration directory before building Mux. 2) build and install Mux 3) change the Default.table for the SerialPorts driver so that the 'Port Count' is '1' instead of '2'. (for good measure, I also edited down the 'IRQ Levels' to only show the IRQ that my mouse actually uses, although I don't think this is necessary) 3) Remove the 'Path:' lines from the SerialPorts Default.table for the port your mouse is not on. (IE: assuming your mouse is on /dev/ttya, remove the 'Path:' lines for /dev/ttyb). You may want to paste these lines into Mux's Default.table, see step 5). 4) the 'I/O Ports' line in the Mux Default.table is missing a trailing semicolon. Add it. There are also two 'Instance' lines. Remove one for good measure. 5) By default, Mux will start assigning lines from /dev/ttya. If, like me, your mouse is on that particular line, you will have to tell it to start with /dev/ttyb. The easiest way to do this is simply to copy all the 'Path' lines for /dev/ttyb from the SerialPorts Default.table, and then change them to reflect the correct PortCount. (IE: change "Path 1 ..." into "Path 0 ...") 6) fire up Configure, and add in the Mux driver. You should see the SerialPorts driver only taking up one IRQ level. 7) reboot, and keep your fingers crossed. If you've hosed your mouse, you can reboot with "config=Default", to hopefully get a workable configuration back. When I originally did this, I thought it would be possible to have Mux replace the SerialPorts driver, since they are both of 'Family = "Serial"'. No such luck. Configure complained that the SerialMouse driver needed SerialPorts. And it didn't work when I tried it anyway. Examining the Default.table, for the SerialMouse driver, I found the "Driver Dependencies" line and set it to "Mux". No go again. Configure still complains that it needs SerialPorts. (and it still doesn't work) Hmm. A quick check with check with 'strings' shows that the SerialMouse dependency on SerialPorts is hard-coded into Configure, of all places. (now, who's bright idea was this?) IMHO, more time should be spent getting Configure to work properly in cases like this, and less time putting in ridiculous 'Infomation' easter eggs into the app. H*ll, enough time may be saved to write a _REAL_ 1742 driver! I'm mostly happy with the driver at this point; I can at least drive the serial port (16550AFN) reliably at 19.2K on a V.32b/V.42b link without losing data. I haven't been able to get a squeak out of my terminal server when I set the line speed at 57600, although my computer will happily talk to my ZyXEL. Yesterday I didn't have any problems at 38400, but today I get the same results as 38400, so there's still some room for improvement. The Default.tables from my system for SerialPorts and Mux are included below. I have a serial mouse at 0x3F8, IRQ 4, /dev/ttya, and a modem at 0x2F8, IRQ 3, /dev/ttyb. Hope this helps someone! -- Jonathan uuencoded Default.tables for Mux and SerialPorts follow. Default.table for Mux ================================== cut here ================================== begin 644 Default.table M(E1I=&QE(B`](`D)(DUU>"!397)I86PB.PHB5F5R<VEO;B(@/2`)"2(Q+C`B M.PHB1')I=F5R($YA;64B(#T@"2)-=7@B.PHB1F%M:6QY(B`](`D)(E-E<FEA M;"(["B));G-T86YC92(@/2`)"2(P(CL*(E!O<W0M3&]A9"(@/2`)"2)0;W-T M3&]A9"(["@HB3&]C871I;VXB(#T@(E-Y<W1E;2!"87-E8F]A<F0B.PHB25)1 M($QE=F5L<R(@/2`B-"(["B))+T\@4&]R=',B(#T@(C!X,T8X+3!X,T9&(CL* M(D1-02!#:&%N;F5L<R(@/2`B(CL*(DUE;6]R>2!-87!S(B`]("(B.PHB5F%L M:60@25)1($QE=F5L<R(@/2`B.2`S(#0@-2`V(#<@,3`@,3$@,3(@,3,@,30@ M,34B.PH*(E!O<G0@0V]U;G0B(#T@(C$B.PHO*B!"875D(%)A=&5S/R`J+PHB M4W5P<&]R="!#5%,B(#T@(EE%4R(["0DO*B!!;&P@<&]R=',@<W5P<&]R="!H M87)D=V%R92!H86YD<VAA:VEN9R`J+PHB4W5P<&]R="!$:6%L:6XB(#T@(EE% M4R(["2\J($%L;"!P;W)T<R!S=7!P;W)T(&1I86QO=70@*'1T>60J*2`J+PHB M4W5P<&]R="!$:6%L;W5T(B`](")915,B.PDO*B!!;&P@<&]R=',@<W5P<&]R M="!D:6%L:6X@*&-U*BD@*B\*(E-E<G9E<B!.86UE(B`](")-=7@B.PHB1')I M=F5R(%9E<G-I;VXB(#T@(E!23T=204TZ375X(%!23TI%0U0Z<V5R:6%L9')I M=F5R+3,N,B!$159%3$]015(Z36%R:RY386QY>GEN($)524Q4.CDT+S`Q+S$S M(CL*"B\J(%!A=&@@;F%M97,@=7-E9"!F;W(@<&]R=',@*B\*(E!A=&@@,"() M"3T@(B]D978O='1Y8B(["B)0871H(#`@24XB"0D]("(O9&5V+W1T>61B(CL* M(E!A=&@@,"!/550B"0D]("(O9&5V+V-U8B(["B)0871H(#`@0U13(@D)/2`B M+V1E=B]T='EF8B(["B)0871H(#`@24X@0U13(@D)/2`B+V1E=B]T='ED9F(B B.PHB4&%T:"`P($]55"!#5%,B"3T@(B]D978O8W5F8B(["F1E ` end Default.table for SerialPorts ================================== cut here ================================== begin 644 Default.table M(E1I=&QE(B`](")3>7-T96T@4V5R:6%L(CL*(E9E<G-I;VXB(#T@(C$N,"([ M"B)$<FEV97(@3F%M92(@/2`B4V-C.#(P-3`B.PHB1F%M:6QY(B`](")397)I M86PB.PHB3&]C871I;VXB(#T@(E-Y<W1E;2!"87-E8F]A<F0B.PHB26YS=&%N M8V4B(#T@(C`B.PHB25)1($QE=F5L<R(@/2`B-"(["B))+T\@4&]R=',B(#T@ M(C!X,V8X+3!X,V9F(CL*(D1-02!#:&%N;F5L<R(@/2`B(CL*(DUE;6]R>2!- M87!S(B`]("(B.PHB5F%L:60@25)1($QE=F5L<R(@/2`B-"(["@HB4&]R="!# M;W5N="(@/2`B,2(["B\J($)A=60@4F%T97,_("HO"B)3=7!P;W)T($-44R(@ M/2`B6453(CL)"0DO*B!!;&P@<&]R=',@<W5P<&]R="!H87)D=V%R92!H86YD M<VAA:VEN9R`J+PHB4W5P<&]R="!$:6%L:6XB(#T@(EE%4R(["0DO*B!!;&P@ M<&]R=',@<W5P<&]R="!D:6%L;W5T("AT='ED*BD@*B\*(E-U<'!O<G0@1&EA M;&]U="(@/2`B6453(CL)"2\J($%L;"!P;W)T<R!S=7!P;W)T(&1I86QI;B`H M8W4J*2`J+PH*(E!O<W0M3&]A9"(@/2`B4&]S=$QO860B.PH*+RH@4&%T:"!N M86UE<R!U<V5D(&9O<B!P;W)T<R`J+PHB4&%T:"`P(@D)/2`B+V1E=B]T='EA M(CL*(E!A=&@@,"!)3B()"3T@(B]D978O='1Y9&$B.PHB4&%T:"`P($]55"() M"3T@(B]D978O8W5A(CL*(E!A=&@@,"!#5%,B"0D]("(O9&5V+W1T>69A(CL* M(E!A=&@@,"!)3B!#5%,B"0D]("(O9&5V+W1T>61F82(["B)0871H(#`@3U54 M($-44R()/2`B+V1E=B]C=69A(CL*(E-E<G9E<B!.86UE(B`](")397)I86Q0 M;W)T<R(["B)$<FEV97(@5F5R<VEO;B(@/2`B4%)/1U)!33I397)I86Q0;W)T M<R!04D]*14-4.G-E<FEA;&1R:79E<BTS+C(@1$5614Q/4$52.FUF;'EN;B!" =54E,5#I.3R!$051%(%-%5"`H+4(@=7-E9"DB.PI6 ` end
From: ccx009@rowan.coventry.ac.uk (Adam Bentley) Newsgroups: comp.sys.next.bugs,comp.sys.next.misc,comp.sys.next.software Subject: Sendmail on NS 3.0 Followup-To: poster Date: 20 Jan 1994 13:37:32 -0000 Organization: Coventry University Message-ID: <2hm1es$rd@rowan.coventry.ac.uk> IS there a version of sendmail for NeXT's (68040 Black h/w) that works? I have a lab of NS 3.0 machines and when I attempted to put a new sendmail.cf on the main gateway machine (one which functions fine on Suns,Decs,SGI's Sequents, etc.), it makes an absolute nonsense of the rules.... Where can I get a version of sendmail that works? thanks for any info. -- _ /-\dam ------------------------------------------------------------------------- FLESH: Adam Bentley (Fraggle), Systems/Networking, Coventry University. UK INET : adamb@rowan.coventry.ac.uk
From: mark@ve6mgs.ampr.ab.ca (Mark Gregory Salyzyn) Newsgroups: comp.sys.next.bugs Subject: cancel <CJx0sC.FGB@ve6mgs.ampr.ab.ca> Message-ID: <CJx268.Ep@ve6mgs.ampr.ab.ca> Date: Thu, 20 Jan 1994 07:01:19 GMT Control: cancel <CJx0sC.FGB@ve6mgs.ampr.ab.ca> Organization: VE6MGS Gateway cancel <CJx0sC.FGB@ve6mgs.ampr.ab.ca> in newsgroup comp.sys.next.bugs
From: mag@rz.uni-jena.de (Andreas Gutsch) Newsgroups: comp.sys.next.bugs Subject: problem with adaptec Date: 20 Jan 1994 17:01:48 GMT Organization: Friedrich-Schiller-University Jena, Germany Message-ID: <2hmdds$s04@hpux.rz.uni-jena.de> Hi, I have a problem with NSFIP 3.2, Compaq Prosignia 466 and ADAPTEC 1542B. 3.1 works fine, but with 3.2 the I always get a system panic during the installation. NS finds the Adaptec at IRQ11 say: registering sc0 and crashs with a system panic. The message is AHAController: out of ccbs. Thanks for help ! mail (no NeXTmail) to mag@rz.uni-jena.de
From: Charles William Swiger <infidel+@CMU.EDU> Newsgroups: comp.sys.next.misc,comp.sys.next.software,comp.sys.next.bugs Subject: Re: Sendmail on NS 3.0 Date: Thu, 20 Jan 1994 13:57:07 -0500 Organization: Fifth yr. senior, Chemistry, Carnegie Mellon, Pittsburgh, PA Message-ID: <chDhE3600WBO0386QX@andrew.cmu.edu> In-Reply-To: <2hm1es$rd@rowan.coventry.ac.uk> Excerpts from netnews.comp.sys.next.bugs: 20-Jan-94 Sendmail on NS 3.0 by Adam Bentley@rowan.coven > IS there a version of sendmail for NeXT's (68040 Black h/w) that works? > I have a lab of NS 3.0 machines and when I attempted to put a new > sendmail.cf on the main gateway machine (one which functions fine on > Suns,Decs,SGI's, Sequents, etc.), it makes an absolute nonsense of the > rules.... > > Where can I get a version of sendmail that works? V8.6.4 (or later) of sendmail is available: via anonymous FTP from FTP.CS.Berkeley.EDU in /ucb/sendmail It fixes all of the known security holes in sendmail. [Yes, there are security holes in the version of sendmail shipped with all versions of NEXTSTEP, at least up to 3.1 -- and I'll be seeing 3.2 soon, so I'll check when I get it.] -Chuck Charles William Swiger -- CMU...*splat*! | 1. You can't fly. --------------------------------------------+ 2. Cars are always real, even AMS & normal mail: infidel@cmu.edu | when they're not. Failing that: cs4w+@andrew.cmu.edu | 3. Police are not your friends. NeXTmail: chuck@cswiger.slip.andrew.cmu.edu | 4. Fire burns.
Newsgroups: comp.sys.next.bugs From: trapac@netcom.com (Transpacific Container) Subject: 3.2 DBKit and Oracle adaptors Message-ID: <trapacCJyFB1.1JG@netcom.com> Organization: NETCOM On-line Communication Services (408 241-9760 guest) Date: Fri, 21 Jan 1994 00:42:36 GMT We've encountered an application performance problem that appears to be related to 3.2 and DBKit. By simply relinking our 3.1 compiled application using 3.2 PB, I can slow execution of parts of it (initialization, DB access and some scrolling) down by about 4 times. Comparing the app as linked under 3.1 and 3.2, I found the following adaptor IDs in the executable code: 3.1 executable's adaptor ID: LIBRARY:OracleAdaptor PROJECT:oracle-34 DEVELOPER:mflynn BUILT:Mon Apr 12 16:22:56 PDT 1993 3.2 executable's adaptor ID: LIBRARY:OracleAdaptor6 PROJECT:oracle-41 DEVELOPER:mflynn BUILT:Tue Aug 24 20:04:31 PDT 1993 Additionally the 3.1 built app contains a additional file "realnames.dbmodel" that is not included in the 3.2 linked app. I assume that 3.2 DBKit changes made this file unnecessary. But who knows? Looking at the 3.2 Oracle adaptors, reveals that there are 4 files that all contain the same ID as shown above. This first lead me to believe that the adaptors are indentical. DBModeler however does not work with OracleAdaptor6. When used, DBModeler responds with one of two errors. If "OracleAdaptor6" is selected first, the following panel is shown: --------------------------------------- Database Oracle login panel not found! --------------------------------------- If after successfully using "OracleAdaptor", I attempt to use "OracleAdaptor6", the following error occurs: Jan 20 10:00:34 komodo_dragon DBModeler[547]: Error loading /NextLibrary/Adaptors/OracleAdaptor6.adaptor/OracleAdaptor6 A background, we have an Oracle7 server and have successfully been accessing it from NS (3.1 & 3.2). Setup for use with NeXTSTEP included: In NetInfo's root domain, orasrv at 1525. tcpctl running the orasrv process on the server. v6 compatibility mode (set by the catalog6.sql script). Any ideas or suggestions welcome. Keith Carpenter trapac@netcom.com
Newsgroups: comp.sys.next.bugs From: fabien@free.fdn.org (Fabien Roy) Subject: Re: AcChen crashing under NS 3.2 Moto Message-ID: <1994Jan20.202955.1166@free.fdn.org> Sender: news@free.fdn.org Organization: Fabien Roy Electronic Engineering (F.R.E.E.) - Paris, France. References: <2hka14$sel@news.duke.edu> Date: Thu, 20 Jan 1994 20:29:55 GMT In article <2hka14$sel@news.duke.edu> blake015@mc.duke.edu(Denise Blakeley) writes: > I recently upgraded my standalone Turbo Color from NS 3.1 to 3.2. I didn't > notice if this started happening right away or later, but now AcChen crashes > right away upon launch. It gets as far as putting up the "Welcome to AcChen" > window, and almost as fast as you can blink it all disappears. The "WELCOME!" > sound is not played. > > I tried re-installing AcChen, but this made no difference. > > Can anyone tell me what's wrong before I go into severe withdrawal? Does > c not run under 3.2? > > Denise (AcChen Addict) Blakeley > -- > Denise Blakeley | PROGRAM, tr. v., An activity similar > Duke Med Center Info Systems | to banging one's head against a wall, > Durham, NC | but with fewer opportunities for > (919) 286-6468 W | reward. > blake015@mc.duke.edu | NeXTMail welcome! Hi Denise, There is a way to know what happen to AcChen: 1) before launching AcChen open the console window (Cmd Shift C) and post the printing. 2) an alternate method is gdb (you must have the developer package installed) and type what is between " " without the " ": a) in a shell window type "gdb /Some_Path/AcChen.app/AcChen" Some_Path= the path to the directory where AcChen.app is located. b) type "run" c) when AcChen crashes type "where" and post the printing. d) type "q" to quit gdb Cheers --Fabien P.S. AcChen Addict also. ___________________________________________________________________ Fabien Roy fabien@free.fdn.org NeXTmail OK
Newsgroups: comp.sys.next.bugs,comp.sys.next.sysadmin From: jmh@info.polymtl.ca (Jean-Marc Heneman) Subject: swap space BIG! help!!! Message-ID: <1994Jan20.234551.927@vlsi.polymtl.ca> Sender: news@vlsi.polymtl.ca (USENET News System) Organization: Ecole Polytechnique de Montreal Date: Thu, 20 Jan 1994 23:45:51 GMT my swapfile : 111888 -rw------t 1 root 114507776 Jan 20 18:27 swapfile 116320 -rw------t 1 root 119111680 Jan 20 18:27 swapfile.front What can I do? HELP, I have no more diskspace... (please direct-mail me to jmh@info.polymtl.ca as soon as you can!) -- jmh jean-marc heneman = jmh@info.polymtl.ca if mail bounces, try jmh@aircanada.ca I prefer ASCII mail for now, but you can send me NextMail for attachment __ / /_/ __/ / /
Newsgroups: comp.sys.next.bugs,comp.sys.next.sysadmin From: jmh@info.polymtl.ca (Jean-Marc Heneman) Subject: Re: swap space BIG! help!!! Message-ID: <1994Jan21.004235.1607@vlsi.polymtl.ca> Followup-To: comp.sys.next.bugs,comp.sys.next.sysadmin Sender: news@vlsi.polymtl.ca (USENET News System) Organization: Ecole Polytechnique de Montreal References: <1994Jan20.234551.927@vlsi.polymtl.ca> Date: Fri, 21 Jan 1994 00:42:35 GMT I resolved the prob : just reboot! thanks and sorry! jean-marc Jean-Marc Heneman (jmh@info.polymtl.ca) wrote: : my swapfile : : 111888 -rw------t 1 root 114507776 Jan 20 18:27 swapfile : 116320 -rw------t 1 root 119111680 Jan 20 18:27 swapfile.front : What can I do? : HELP, I have no more diskspace... : (please direct-mail me to jmh@info.polymtl.ca as soon as you can!) : -- : jmh : jean-marc heneman = jmh@info.polymtl.ca : if mail bounces, try jmh@aircanada.ca : I prefer ASCII mail for now, but you can send me NextMail for attachment : __ : / /_/ : __/ / / -- jmh jean-marc heneman = jmh@info.polymtl.ca if mail bounces, try jmh@aircanada.ca I prefer ASCII mail for now, but you can send me NextMail for attachment __ / /_/ __/ / /
From: robert@steffi.demon.co.uk (Robert Nicholson) Newsgroups: comp.sys.next.bugs Subject: DL 3.1 problem Date: 21 Jan 1994 13:40:34 -0000 Organization: me organized? That's a joke! Message-ID: <2hom0i$20h@steffi.demon.co.uk> Why is it that when I create a brand new index on my Examples directory and I go searching for "bundles" I get a whole lot of filenames back but when I load each file the DL document window just beeps because it didn't find the string "bundles" anywhere? What constitutes are "Word" and "Literal" in DL? -- "You know what's wrong with you?" (Audrey Hepburn, Cary Grant) "Nothing" Charade, 1963 robert@steffi.demon.co.uk (ASCII for text only messages)
From: geom2@sfb256.iam.uni-bonn.de ( Michael Moellney ) Newsgroups: comp.sys.next.misc,comp.sys.next.software,comp.sys.next.bugs Subject: TeXview hangs after Upgrade to 3.2 Date: 21 Jan 1994 16:58:16 GMT Organization: Applied Math, University of Bonn, Germany Message-ID: <2hp1j8$7t@news.rhrz.uni-bonn.de> Hi! I upgraded an Intel from 3.1 to 3.2. NeXTTeX was installed in 3.1 and I changed some formats and don't wanted that to be erased. So I didn't installed NeXTTeX from the 3.2. I also use TeXMenu, if it matters. The Problem: Every time I try to print to a file by pressing save in the print-panel and select a filename, TeXview starts dvips. That's ok, but after dvips processed the dvi-file and generated a ps-file TeXview doesn't realize this and wait's (spinning wheel) until it is killed by me! Selecting 'Don't wait on dvips' doesn't change this behaviour. I installed the 3.2 NeXTSTEP NeXTTeX package on an other host in our net and copied only TeXview to my problem machine. No CHANGE! So did anybody have the same expirience? Any clues? Sometimes the Console Window of TeXview stops to show the output of dvips. Looks like this: [1] [2] [3 and nothing else (there are more sides and dvips processes them to the very end. You get the ps-file at full) Very seldom dvips isn't started and so doesn't produce output. Thank you for any hints! Michael
From: marcus@ursula.ee.pdx.edu (Marcus Daniels) Newsgroups: comp.sys.next.bugs Subject: I386 assembler doesn't support %= GCC notation Date: 22 Jan 1994 02:03:54 -0800 Message-ID: <2hqtma$ie5@ursula.ee.pdx.edu> Summary: I386 assembler barfs on labels with `=' If one creates a loop or conditional logic with several instructions within a single asm(), GCC provides a means to create unique labels for each usage -- namely %=. This works fine with the Motorola assembler, but not with Intel. It chokes on the `=' that gets embedded in the newly created symbol. Does anyone know a workaround (other than -B to a different assembler (front end)? Marcus Daniels
From: burton@het.brown.edu (Joshua W. Burton) Newsgroups: comp.sys.next.sysadmin,comp.sys.next.bugs Subject: Internal Error 3007 at login ??? Date: 22 Jan 1994 15:17:31 GMT Organization: Brown University Message-ID: <2hrg2b$c0g@cat.cis.Brown.EDU> I just finished rebuilding my filesystem on a new hard disk after a total failure of my boot disk. Unfortunately, I was missing one dumpfile in an incremental chain, so I had to restore some files one at a time from backup copies outside of my dump scheme. I had a lot of redundancy, and I believe I restored everything (finally!), but I am still suspicious of possible permissions problems that might be causing the following bizarre behavior: Our user accounts are joshua, debbie, me, and the root account. When joshua logs out, and debbie then logs in, the Workspace Manager comes up, all the right icons appear on the dock, and some of them (Mail.app and Edit) launch. The other ones set for launch (Stuart.app, Eyecon, and Backspace) are still in the launching process, highlighted, when the user gets an alert panel: "Internal Error 3007. Proceed at your own risk." The two options are Continue and Log Out, and no matter which one the user presses the screen goes blank gray (cursor still intact), and the only way to proceed is to restart the system with a nerve pinch. This happens ONLY when joshua logs out and debbie then logs in; the sequences joshua -> root -> debbie, joshua -> me -> debbie, and so on result in a normal login for debbie. Me is user 20, joshua is 21, and debbie is 22. Debbie owns all the files in her home directory, and her home directory is shown correctly in UserManager. The behavior is completely reproducible: whether joshua simply logs in and out, or does several hours of work, he always leaves the machine in a state that will hang debbie's login. We are running NS 2.1 on an '040 cube with a single brand new internal hard drive. Questions: (1) Does anyone have the remotest idea what is going on? (2) Where, in the on-line or the full paper documentation, are Workspace internal error numbers documented? The only other one I have ever seen is 3001, I think---from trying to overtax Display Postscript with a ridiculous number of fonts in one document. (3) For future troubleshooting purposes, where am I when the Workspace Manager is blown away (plain gray screen) but I still have a cursor? Where would I be if the cursor were gone? I've seen both conditions, and would like to be able to learn from the distinction. Any help would be _much_ appreciated. Please reply by posting or by mail to <burton@het.brown.edu>. I'll report again when I've solved it. 2.998 x 10^10 cm/sec: |================================================= It's not just a good | Joshua Burton (401)435-6370 burton@het.brown.edu idea. It's the LAW. |=================================================
From: burton@het.brown.edu (Joshua W. Burton) Newsgroups: comp.sys.next.bugs Subject: Internal Error 3007 at login ??? Date: 22 Jan 1994 15:18:44 GMT Organization: Brown University Message-ID: <2hrg4k$c0q@cat.cis.Brown.EDU> I just finished rebuilding my filesystem on a new hard disk after a total failure of my boot disk. Unfortunately, I was missing one dumpfile in an incremental chain, so I had to restore some files one at a time from backup copies outside of my dump scheme. I had a lot of redundancy, and I believe I restored everything (finally!), but I am still suspicious of possible permissions problems that might be causing the following bizarre behavior: Our user accounts are joshua, debbie, me, and the root account. When joshua logs out, and debbie then logs in, the Workspace Manager comes up, all the right icons appear on the dock, and some of them (Mail.app and Edit) launch. The other ones set for launch (Stuart.app, Eyecon, and Backspace) are still in the launching process, highlighted, when the user gets an alert panel: "Internal Error 3007. Proceed at your own risk." The two options are Continue and Log Out, and no matter which one the user presses the screen goes blank gray (cursor still intact), and the only way to proceed is to restart the system with a nerve pinch. This happens ONLY when joshua logs out and debbie then logs in; the sequences joshua -> root -> debbie, joshua -> me -> debbie, and so on result in a normal login for debbie. Me is user 20, joshua is 21, and debbie is 22. Debbie owns all the files in her home directory, and her home directory is shown correctly in UserManager. The behavior is completely reproducible: whether joshua simply logs in and out, or does several hours of work, he always leaves the machine in a state that will hang debbie's login. We are running NS 2.1 on an '040 cube with a single brand new internal hard drive. Questions: (1) Does anyone have the remotest idea what is going on? (2) Where, in the on-line or the full paper documentation, are Workspace internal error numbers documented? The only other one I have ever seen is 3001, I think---from trying to overtax Display Postscript with a ridiculous number of fonts in one document. (3) For future troubleshooting purposes, where am I when the Workspace Manager is blown away (plain gray screen) but I still have a cursor? Where would I be if the cursor were gone? I've seen both conditions, and would like to be able to learn from the distinction. Any help would be _much_ appreciated. Please reply by posting or by mail to <burton@het.brown.edu>. I'll report again when I've solved it. 2.998 x 10^10 cm/sec: |================================================= It's not just a good | Joshua Burton (401)435-6370 burton@het.brown.edu idea. It's the LAW. |=================================================
Newsgroups: comp.sys.next.bugs From: ctm@ardi.com (Clifford T. Matthews) Subject: Re: I386 assembler doesn't support %= GCC notation Message-ID: <CK1wKK.62w@cobra.cs.unm.edu> Sender: news@cobra.cs.unm.edu Organization: ARDI References: <2hqtma$ie5@ursula.ee.pdx.edu> Date: Sat, 22 Jan 1994 21:48:22 GMT In article <2hqtma$ie5@ursula.ee.pdx.edu> marcus@ursula.ee.pdx.edu (Marcus Daniels) writes: >If one creates a loop or conditional logic with several instructions >within a single asm(), GCC provides a means to create unique labels >for each usage -- namely %=. This works fine with the Motorola >assembler, but not with Intel. It chokes on the `=' that gets >embedded in the newly created symbol. Does anyone know a workaround (other >than -B to a different assembler (front end)? > >Marcus Daniels You might want to try the "f" and "b" notation. I don't recall the history of this stuff, but gas and other UNIX assemblers often allow you to use the integer labels with references to them requiring the suffix "f" or "b" to denote "forward" or "backward". The first integer match found by traversing the source code from the reference in the direction denoted by the suffix is what will be used. Example: blah blah jump 1f /* get to the common loop end stuff */ 2: blah blah test jump 2f /* exit loop */ 1: some more stuff jump 2b /* back to top of loop */ 2: out-of-here I believe I've used this notation on the intel side of things, but if it doesn't work, there's a chance that I haven't, or that I have, but I've replaced the Intel assembler with an unmodified gas (I don't remember doing so, but I do have a tendency to replace the NeXT tools with ones from the FSF where possible). I'm at home now and can't easily check. --Cliff ctm@ardi.com
Newsgroups: comp.sys.next.bugs From: thf@zelator.in-berlin.de (Thomas Funke) Subject: Re: AcChen crashing under NS 3.2 Moto Message-ID: <1994Jan21.140158.833@gamelan> Sender: thomas@gamelan (thomas) Organization: Disorganization References: <2hka14$sel@news.duke.edu> Date: Fri, 21 Jan 1994 14:01:58 GMT In article <2hka14$sel@news.duke.edu> blake015@mc.duke.edu(Denise Blakeley) writes: > Can anyone tell me what's wrong before I go into severe withdrawal? > Does AcChen not run under 3.2? It runs on my cube. -- ------------------------------------------------------------------ Thomas Funke ** Unix-Consultant ** thf@zelator.in-berlin.de Brooks's Law: Adding manpower to a late software project makes it later ------------------------------------------------------------------
From: sanguish@digifix.com Newsgroups: comp.sys.next.announce,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.bugs,comp.sys.next.programmer,comp.sys.next.marketplace Subject: Quick Guide to NEXTSTEP information on the Internet Date: 24 Jan 1994 15:32:36 -0500 Organization: Next Announcements Message-ID: <2i1b94$ba7@digifix.digifix.com> This post is made weekly, to help 'point' users to more NEXTSTEP information Topics include: comp.sys.next newsgroups related newsgroups comp.sys.next newsgroups mailing list ftp sites NeXTanswers comp.sys.next.* newsgroups -------------------------- Comp.Sys.Next.Advocacy This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. Comp.Sys.Next.Announce Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) The NEXTSTEP FAQs are posted here monthly as well. This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Comp.Sys.Next.Bugs A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT- specific groups as well. Comp.Sys.Next.Hardware Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. Comp.Sys.Next.Marketplace NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. Comp.Sys.Next.Misc For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! Comp.Sys.Next.Programmer Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. The NEXTSTEP programmer FAQs are posted here. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. Comp.Sys.Next.Software This is a place to talk about [third party] software products that run on NEXTSTEP systems. Comp.Sys.Next.Sysadmin Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. related Newsgroups ------------------ Comp.Soft-Sys.Nextstep Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. Comp.Lang.Objective-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. Comp.Object Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original Comp.Sys.Next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News ------------------------------------------- Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep-d next-advocacy-d next-announce-d next-bugs-d next-hardware-d next-marketplace-d next-misc-d next-programmer-d next-software-d next-sysadmin-d (For a full description, send mail saying LISTS to <digestif@antigone.com>). The subscription syntax is essentially the same as LISTSERV's. To subscribe, send a message to <digestif@antigone.com> saying: SUB Listname YourName Example: SUB next-hardware-d John Doe The ftp sites ------------- cs.orst.edu: The main site for North American submissions nova.cc.purdue.edu: Lots of older stuff, but very short on disk space ftp.informatik.uni-muenchen.de: In Germany. terra.stack.urc.tue.nl (Dutch NEXTSTEP User Group) and cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp.next.com: See the below ftp.next.com and NextAnswers@next.com ------------------------------------- From the document 1000_Help from ftp.next.com Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, or you can transfer them by anonymous ftp. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to NeXTanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO <your-address> If you have any problem with the system or suggestions for improvement, please send mail to NeXTanswers-request@NeXT.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to NeXTanswers-request@NeXT.com. Written by: Eric P. Scott eps@toaster.SFSU.EDU and Scott Anguish sanguish@digifix.com Additions from: Greg Anderson (Greg_Anderson@afs.com) and Michael Pizolato (Michael_Pizolato@afs.com)
Newsgroups: comp.sys.next.misc,comp.sys.next.software,comp.sys.next.bugs From: Robert_La_Ferla@hot.com Subject: Re: Sendmail on NS 3.0 Message-ID: <1994Jan26.212256.10498@hot.com> Sender: robertl@hot.com Organization: Hot Technologies References: <chDhE3600WBO0386QX@andrew.cmu.edu> Date: Wed, 26 Jan 1994 21:22:56 GMT Does this latest version of sendmail contain the changes NeXT made to support NetInfo? Robert La Ferla Hot Technologies NEXTSTEP ISV In article <chDhE3600WBO0386QX@andrew.cmu.edu> Charles William Swiger <infidel+@CMU.EDU> writes: > Excerpts from netnews.comp.sys.next.bugs: 20-Jan-94 Sendmail on NS 3.0 > by Adam Bentley@rowan.coven > > IS there a version of sendmail for NeXT's (68040 Black h/w) that works? > > I have a lab of NS 3.0 machines and when I attempted to put a new > > sendmail.cf on the main gateway machine (one which functions fine on > > Suns,Decs,SGI's, Sequents, etc.), it makes an absolute nonsense of the > > rules.... > > > > Where can I get a version of sendmail that works? > > V8.6.4 (or later) of sendmail is available: > > via anonymous FTP from FTP.CS.Berkeley.EDU in /ucb/sendmail > > It fixes all of the known security holes in sendmail. [Yes, there are > security holes in the version of sendmail shipped with all versions of > NEXTSTEP, at least up to 3.1 -- and I'll be seeing 3.2 soon, so I'll > check when I get it.] > > -Chuck > > > Charles William Swiger -- CMU...*splat*! | 1. You can't fly. > --------------------------------------------+ 2. Cars are always real, even > AMS & normal mail: infidel@cmu.edu | when they're not. > Failing that: cs4w+@andrew.cmu.edu | 3. Police are not your friends. > NeXTmail: chuck@cswiger.slip.andrew.cmu.edu | 4. Fire burns. >
From: elitman@nxstep.com (Eric Litman) Newsgroups: comp.sys.next.hardware,comp.sys.next.bugs Subject: DPT 2022 Failures Date: 27 Jan 1994 02:51:53 GMT Organization: Proxima, Inc. Message-ID: <2i7a89$nb1@hecate.umd.edu> Summary: DPT just can't seem to get it right. We've fielded over 120 machines (Intel Pro/GX) with DPT2022 controllers and various hard drives. We've run the gambit in firmware (from 5A through many engineering revs to the latest). We've spoken with just about everyone in a technical and managment position within DPT. And still, the intermittent failures we experience with these controllers are far from being solved. Anyone figured out the "working" combination? Currently, we have the best luck with the Quantum LPS drive series, but also use DEC DSP (fast, very fast), and Maxtor. -Eric Litman Proxima, Inc.
Newsgroups: comp.sys.next.bugs From: bruce@TotSysSoft.com (Bruce Gingery) Subject: Re: Internal Error 3007 at login ??? Message-ID: <1994Jan26.215153.1758@ToTSySSoft.com> Sender: news@ToTSySSoft.com Organization: Total System Software References: <2hrg4k$c0q@cat.cis.Brown.EDU> Date: Wed, 26 Jan 1994 21:51:53 GMT In article <2hrg4k$c0q@cat.cis.Brown.EDU> burton@het.brown.edu (Joshua W. Burton) writes: -> I just finished rebuilding my filesystem on a new hard -> disk after a total failure of my boot disk. -> Unfortunately, I was missing one dumpfile in an -> incremental chain, so I had to restore some files -> one at a time from backup copies outside of my dump -> scheme. I had a lot of redundancy, and -> I believe I restored everything (finally!), -> but I am still suspicious of possible permissions -> problems that might be causing the following bizarre -> behavior: -> -> Our user accounts are joshua, debbie, me, and the root -> account. When joshua logs out, and debbie then logs in, -> the Workspace Manager comes up, all the right icons appear -> on the dock, and some of them (Mail.app and Edit) launch. -> The other ones set for launch (Stuart.app, Eyecon, and -> Backspace) are still in the launching process, highlighted, -> when the user gets an alert panel: "Internal Error 3007. -> Proceed at your own risk." -> The two options are Continue and Log Out, and no matter -> which one the user presses the screen goes blank gray -> (cursor still intact), and the only -> way to proceed is to restart the system with a nerve pinch. -> -> This happens ONLY when joshua logs out and debbie then logs -> in; the sequences joshua -> root -> debbie, joshua -> me -> -> debbie, and so on result in a normal login for debbie. -> Me is user 20, joshua is 21, and debbie is 22. Debbie -> all the files in her home directory, and her -> home directory is shown correctly in UserManager. The -> behavior is completely reproducible: whether joshua simply -> logs in and out, or does several hours of work, he always -> leaves the machine in a state that will hang debbie's -> login. We are running NS 2.1 on an '040 cube with a -> single brand new internal hard drive. -> -> Questions: -> -> (1) Does anyone have the remotest idea what is going on? -> -> (2) Where, in the on-line or the full paper documentation, -> are Workspace internal error numbers documented? The -> only other one I have ever seen is 3001, I think--- -> from trying to overtax Display Postscript -> with a ridiculous number of fonts in one document. -> -> (3) For future troubleshooting purposes, where am I when -> the Workspace Manager is blown away (plain gray screen) -> but I still have a cursor? Where would I be if the -> cursor were gone? I've seen both conditions, -> and would like to be able to learn from the -> distinction. -> -> Any help would be _much_ appreciated. Please reply by -> posting or by mail to <burton@het.brown.edu>. I'll -> report again when I've solved it. Your problem is likely multiple in origin. At first I thought that it was likely related to a reported bug (in 3.1, and likely just went unreported before), where re-logins in Terminal.app without interleaving root would NOT properly modify all user annotations. As for all-gray but with a cursor, sounds like a dead but not exited Workspace or loginwindow, with Windowserver alive. If you can get a "remote" login, ie via network or null-modem as root, you can see just what is going on from commandline as root. Bruce Gingery bruce@TotSysSoft.com
From: hironobu@sra.CO.JP (Hironobu SUZUKI) Newsgroups: comp.sys.next.bugs Subject: join command NeXTSTEP for Intel 3.2 Message-ID: <HIRONOBU.94Jan27190802@sran265.sra.CO.JP> Date: 27 Jan 94 10:08:02 GMT Sender: news@sran265.sra.co.jp Organization: Software Research Associates, Inc., Japan /usr/bin/join still have bug under NS/I 3.2. (same under NS 3.0 on my Cube) Please type this command line on your NeXTSTEP. % join -j1 4 -j2 3 -t: /etc/passwd /etc/group Sorry, I can't send e-mail by NeXTBUG.app because my Intel machine don't have Ethernet Card. ----------------------------------------------------------------- Hironobu Suzuki Practical Software Engineering Laboratory Software Research Associates, Inc, Japan. 3-12, Yotsuya, Shinjuku-ku, Tokyo, 160, JAPAN. E-Mail: hironobu@sra.co.jp NeXT-Mail: hironobu@ext247.sra.co.jp
Newsgroups: comp.sys.next.misc,comp.sys.next.software,comp.sys.next.bugs From: kramer@fragile.termfrost.org (Mike Andrews) Subject: Re: Sendmail on NS 3.0 Message-ID: <CKAp8x.M5n@fragile.termfrost.org> Organization: Terminal Frost, Springfield OH References: <chDhE3600WBO0386QX@andrew.cmu.edu> <1994Jan26.212256.10498@hot.com> Date: Thu, 27 Jan 1994 15:48:33 GMT In <1994Jan26.212256.10498@hot.com> Robert_La_Ferla@hot.com writes: >Does this latest version of sendmail contain the changes NeXT made to >support NetInfo? >Robert La Ferla >Hot Technologies >NEXTSTEP ISV 8.6.4 doesn't, but 8.6.5 does. I haven't finished installing it yet, so does anyone know of any big gotchas here? I'm still running the stock 5.67 now... -- Mike Andrews "This guy's pretty bizarre, Gus." root@fragile.termfrost.org [NeXTmail OK] - Primus kramer@wittenberg.edu (school) kramer@mik.uky.edu (hometown) Bassists do it deeper
From: dhess@net.tamu.edu Newsgroups: comp.sys.next.bugs Subject: Strange things with inetd and lpd... Date: 27 Jan 1994 20:55:01 GMT Organization: Texas A&M University, College Station, TX Distribution: world Message-ID: <2i99n5$cfe@news.tamu.edu> Well this is a weird one and I was wondering if anyone else has experienced it or can recreate it. The simple thing is to kill and restart inetd. Then try to telnet into the machine. No matter which account, username, home directory or shell you try, all you get is standard messages from login and then the connection closes. I have been able to recreate this. Rlogin and exec'ing login still works. Rebooting the machine fixes it. lpd has the same sort of problem. I have Dots installed with a HP500C printer. The if filter for the printcap entry is a shell script (standard BSD setup) which invokes the Dots converter. If I kill lpd and restart it, the converter dies after writing a tiny bit to the output. I can run the converter by hand and it runs fine and completes with no problem. Rebooting once again will fix the problem. I thought maybe my environment and current directory from the shell where I invoked inetd and lpd might be causing a problem so I stripped down the environment and cd'd to / (like what the rc scripts would have on startup). No luck. Then I thought maybe it was signal masks setup in my bash shell so I made sure I started them from a Bourne shell that did not have a bash shell as an ancestor. No luck. I am baffled. I at least know that I can fix it with a reboot. However, I would *really* like to know what the hell is going on. Dave -- David K. Hess Network Analyst David-Hess@tamu.edu Computing and Information Services - Network Group (409) 845-0372 (work) Texas A&M University
Newsgroups: comp.sys.next.bugs From: glen@instep.wimsey.bc.ca Subject: printing DBTableViews Message-ID: <1994Jan28.224430.5277@instep.wimsey.bc.ca> Sender: glen@instep.wimsey.bc.ca Organization: InStep Mobile Communications Inc. Date: Fri, 28 Jan 1994 22:44:30 GMT I have a DBTableView layed out in a .nib file which prints fine. To solve an historical DBTableView printing bug, the columns are not editable but the entire table is marked as editable in the IB Inspector. If, in IB, I make the DBTableView uneditable, then editable again; then save the .nib file I get the following printing problem. Each left justified column prints fine. Each right justified column begins printing at the right of the field, leaving the field empty and extending into the field to its right. Each centered column begins printing at the centre of the column analogous to the right justify problem. Any suggestions? -- Glen Biagioni Software Developer glen@instep.wimsey.bc.ca (small NeXTmail accepted) 604 872-7116 604 872-7125 fax
From: ivo@next.agsm.ucla.edu (Ivo Welch) Newsgroups: comp.sys.next.hardware,comp.sys.next.bugs Subject: Re: DPT 2022 Failures Followup-To: comp.sys.next.hardware,comp.sys.next.bugs Date: 29 Jan 1994 00:05:21 GMT Organization: UCLA Microcomputer Support Office Message-ID: <2ic981$jlt@news.mic.ucla.edu> References: <2i7a89$nb1@hecate.umd.edu> I have experienced flakiness with the DPT 2012 adapters, too. Occasionally, even at bootup before anything starts. Shutting everything down for 10 seconds often cures the problem. Once up and running, they seem somewhat reliable (as long as <=2 hard disks are used!) at 5MB/sec. What I want to know is if there are more reliable alternatives to the DPT controller. Ivo Welch ivo@128.97.74.50 = next.agsm.ucla.edu Asst Prof of Finance iwelch@agsm.ucla.edu AGSM at UCLA
From: mow@marsu.tynet.sub.org (Markus Wenzel) Newsgroups: comp.sys.next.hardware,comp.sys.next.bugs Subject: Re: DPT 2022 Failures Date: 28 Jan 1994 23:32:05 +0100 Organization: Palumbian Research Labs Message-ID: <2ic3p5$uv@marsu.tynet.sub.org> References: <2i7a89$nb1@hecate.umd.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit elitman@nxstep.com (Eric Litman) writes: >We've fielded over 120 machines (Intel Pro/GX) with DPT2022 controllers and >various hard drives. We've run the gambit in firmware (from 5A through many >engineering revs to the latest). We've spoken with just about everyone in a >technical and managment position within DPT. And still, the intermittent >failures we experience with these controllers are far from being solved. >Anyone figured out the "working" combination? Currently, we have the best >luck with the Quantum LPS drive series, but also use DEC DSP (fast, very >fast), and Maxtor. Sounds weird... Are you sure you configured IRQ to edge-triggerd with your ECU? What does your typical SCSI chain look like? I currently use a DPT 2021 with a DEC DSP, a Fujitsu 2624, a Toshiba CD-ROM and a Sankyo QIC streamer. I had to remove a SyQuest drive from the chain because it definetly caused a total failure on the SCSI bus each time the DEC DSP was connected to. Furthermore, an actively terminated DEC at the end did not work, but a passively terminated Fujitsu at the end did work. After fiddling aroung, praying to Redwood and some other Voodoo stuff, it finally works great. It is not necessary the host adapter which is responsible for the failure. Regards, Markus. -- Marsu: "There really exist some people who live happily and satis- fied without computers." -- Frankie: "Oh no, they just emulate it!" ----- Markus Wenzel System administration, Consulting, Networking mow@marsu.tynet.sub.org on... NeXTSTEP / Unix / Novell / Windows NT IRC: Marsu Intel aside.
Newsgroups: comp.sys.next.bugs From: lloyd@world.std.com (Christopher Lloyd) Subject: cc never stops trying ... Message-ID: <CKF64s.EAI@world.std.com> Organization: The World Public Access UNIX, Brookline, MA Date: Sun, 30 Jan 1994 01:43:39 GMT // Maybe I'm bananas, maybe not ... /* This code won't compile for me. cc1obj sits and munches cycles for, what seems to be, forever. This is the smallest deriveable example I could come up which reproduces the apparent bug in cc. I've tried this on 3.2 black with -arch m68k and -arch i386 yielding the same lack of results. Can people try it on their configuration and post/mail me on what it does for them? Does it compile, or sit there forever? I have not reported this to NeXT, but will do so once I get some reports. Compile with, err, try to compile with: cc -O test.m -c Control-C to stop when you're sick of waiting. */ #import <objc/Object.h> @interface Blug : Object { unsigned int size; unsigned char *data; unsigned int pos; int *next; } @end @implementation Blug -foo:(const unsigned char *)bufData:(unsigned int)bufSize { int i; for(i=0;i<bufSize && pos<size;i++,pos++) while((pos>-1) && (bufData[i]!=data[pos])) pos=next[pos]; return self; } @end -- :: Christopher Lloyd :: Yrrid, Inc. :: lloyd@yrrid.com lloyd@world.std.com :: :: If I had known it was harmless, I would have killed it myself ::
Newsgroups: comp.sys.next.bugs From: lloyd@world.std.com (Christopher Lloyd) Subject: Re: cc never stops trying ... Message-ID: <CKF6Bx.GI2@world.std.com> Organization: The World Public Access UNIX, Brookline, MA References: <CKF64s.EAI@world.std.com> Date: Sun, 30 Jan 1994 01:47:55 GMT Oh, I know the example is riddled with it's own set of bugs, but that is no reason for the compiler to sit and spin. shrug, -- :: Christopher Lloyd :: Yrrid, Inc. :: lloyd@yrrid.com lloyd@world.std.com :: :: If I had known it was harmless, I would have killed it myself ::
Newsgroups: comp.sys.next.bugs From: c_avery@rsgreet.andi.org (Christopher Avery) Subject: CD-ROM on 3.2 Motorola Message-ID: <CKF3wH.Lr8@nextsrv1.andi.org> Keywords: CD-ROM bug Sender: usenet@nextsrv1.andi.org (usenet) Organization: Association of NeXTSTEP Developers International Date: Sun, 30 Jan 1994 00:55:27 GMT I upgraded from 3.1 to 3.2 and notice that if I boot with a NeXT CD-ROM everything on my scsi port works fine, but if I boot with the JANA CD-ROM the JANA disk does not show up on the workshelf, it does show up in the normal file list and is readable, but my removeable hard drive does not show up in either place and is not accessible. Since this is where I do backups, it is a good thing that it is accessible when a NeXT CD is loaded. I am not booting off the CD, but the internal hard disk. -- Christopher Avery reply to:cavery@andi.org (NeXTMail accepted)
From: glenn@rightbrain.com (Glenn Reid) Newsgroups: comp.sys.next.bugs Subject: Re: cc never stops trying ... Message-ID: <1476@rtbrain.rightbrain.com> Date: 30 Jan 94 08:58:02 GMT References: <CKF64s.EAI@world.std.com> Sender: glenn@rightbrain.com Christopher Lloyd writes [some sample code that wouldn't compile on his system] Since I'm still running 2.1 and I doubt many others are, I thought I'd post the results of this test on my system. In short, it compiled just fine: rtbrain> time cc -O test.m -c 0.5u 0.4s 0:01 71% 0+0k 3+17io 0pf+0w rtbrain> ls -lt total 3 -rw-r--r-- 1 glenn 1816 Jan 30 00:56 test.o -rw-r--r-- 1 glenn 1004 Jan 30 00:56 test.m -- Glenn Reid glenn@rightbrain.com Woodside, California
Newsgroups: comp.sys.next.bugs From: jlemon@netcom.com (Jonathan Lemon) Subject: Re: cc never stops trying ... Message-ID: <jlemonCKGCHx.Ay@netcom.com> Organization: Netcom - Online Communication Services (408 241-9760 guest) References: <CKF64s.EAI@world.std.com> <CKF6Bx.GI2@world.std.com> Date: Sun, 30 Jan 1994 16:58:45 GMT In article <CKF6Bx.GI2@world.std.com> lloyd@world.std.com (Christopher Lloyd) writes: > >Oh, I know the example is riddled with it's own set of bugs, but that is >no reason for the compiler to sit and spin. It appears to be a problem with the optimizer, at least at first glance. Compiling the code without the '-O' flag works fine, compilation under '-g' also works. However, if you specify both "-O" and "-fvolatile", then it works: localhost> time cc -c -O -fvolatile foo.m 0.210u 0.219s 0:00.82 51.2% 0+0k 0+18io 0pf+0w This is on an NS/I running 3.2. I agree that this shouldn't cause the compiler to spin, though. -- Jonathan
From: pkron@corona.com (Peter Kron) Organization: Corona Design, Inc., Seattle, WA Distribution: world Date: Sun, 30 Jan 1994 09:41:47 PST Message-ID: <1994Jan30.174147.930@corona.com> Newsgroups: comp.sys.next.bugs Subject: cc never stops trying ... References: <CKF64s.EAI@world.std.com> From: lloyd@world.std.com (Christopher Lloyd) > // Maybe I'm bananas, maybe not ... > > /* > > This code won't compile for me. cc1obj sits and munches > cycles for, what seems to be, forever. I duplicated it on 3.2 Intel. As someone mentioned, it is a problem with the optimizer, apparently related to the "pos>-1" test. Since pos is defined as *unsigned*, this test is useless. Redefining pos to *int*, compiles properly. korona> diff spinbug.m spinbug2.m 32c32 < unsigned int pos; --- > int pos; korona> /bin/time cc -O -Wall spinbug2.m -c 0.7 real 0.3 user 0.2 sys korona>
From: sanguish@digifix.com Newsgroups: comp.sys.next.announce,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.bugs,comp.sys.next.programmer,comp.sys.next.marketplace Subject: Quick Guide to NEXTSTEP information on the Internet Date: 30 Jan 1994 22:27:59 -0500 Organization: Next Announcements Message-ID: <2ihtrv$ihb@digifix.digifix.com> This post is made weekly, to help 'point' users to more NEXTSTEP information Topics include: comp.sys.next newsgroups related newsgroups comp.sys.next newsgroups mailing list ftp sites NeXTanswers comp.sys.next.* newsgroups -------------------------- Comp.Sys.Next.Advocacy This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. Comp.Sys.Next.Announce Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) The NEXTSTEP FAQs are posted here monthly as well. This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Comp.Sys.Next.Bugs A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT- specific groups as well. Comp.Sys.Next.Hardware Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. Comp.Sys.Next.Marketplace NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. Comp.Sys.Next.Misc For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! Comp.Sys.Next.Programmer Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. The NEXTSTEP programmer FAQs are posted here. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. Comp.Sys.Next.Software This is a place to talk about [third party] software products that run on NEXTSTEP systems. Comp.Sys.Next.Sysadmin Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. related Newsgroups ------------------ Comp.Soft-Sys.Nextstep Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. Comp.Lang.Objective-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. Comp.Object Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original Comp.Sys.Next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News ------------------------------------------- Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep-d next-advocacy-d next-announce-d next-bugs-d next-hardware-d next-marketplace-d next-misc-d next-programmer-d next-software-d next-sysadmin-d (For a full description, send mail saying LISTS to <digestif@antigone.com>). The subscription syntax is essentially the same as LISTSERV's. To subscribe, send a message to <digestif@antigone.com> saying: SUB Listname YourName Example: SUB next-hardware-d John Doe The ftp sites ------------- cs.orst.edu: The main site for North American submissions nova.cc.purdue.edu: Lots of older stuff, but very short on disk space ftp.informatik.uni-muenchen.de: In Germany. terra.stack.urc.tue.nl (Dutch NEXTSTEP User Group) and cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp.next.com: See the below ftp.next.com and NextAnswers@next.com ------------------------------------- From the document 1000_Help from ftp.next.com Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, or you can transfer them by anonymous ftp. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to NeXTanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO <your-address> If you have any problem with the system or suggestions for improvement, please send mail to NeXTanswers-request@NeXT.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to NeXTanswers-request@NeXT.com. Written by: Eric P. Scott eps@toaster.SFSU.EDU and Scott Anguish sanguish@digifix.com Additions from: Greg Anderson (Greg_Anderson@afs.com) and Michael Pizolato (Michael_Pizolato@afs.com)
Newsgroups: comp.sys.next.bugs From: jds@aplexus.jhuapl.edu (John D Stanhope) Subject: NS/i, is tip borken? Message-ID: <CKHxAy.C77@netnews.jhuapl.edu> Keywords: tip Sender: usenet@netnews.jhuapl.edu Organization: Johns Hopkins U. Applied Physics Lab Date: Mon, 31 Jan 1994 13:25:45 GMT Ever since I installed 3.2 I have been unable to get tip to transfer files larger than 20 to 30k reliably. It just drops characters here and there. And when I connect to work I get garbage on my screen now and then. None of this ever happened under 3.1. I reinstalled the system when I received 3.2, but copied over my old remote file, so I know the tip config is the same. Has anyone else experienced these problems, or have I just screwed something up somewhere? -- ---------------------------------------------------- Otto "Apes don't read philoshpy" Wanda "Yes they do, they just don't understand it" ----------------------------------------------------
From: wjabi@libra.arch.umich.edu (Wassim M. Jabi) Newsgroups: comp.sys.next.bugs,comp.sys.next.hardware Subject: The Original NeXTstation Color Printing Bulletin Date: 31 Jan 1994 21:21:30 GMT Organization: University of Michigan Engineering, Ann Arbor Distribution: usa Message-ID: <2ijsoqINN73u@srvr1.engin.umich.edu> I finally dug up the following bulletin from 1991: ***** Irregular Printing on NeXTstation Color ***** NeXT is committed to delivering high-quality products. We use the highest quality components and a careful manufacturing process to minimize any functionality problems. However, we do occasionally encounter malfunctions. Specifically, there have been a limited number of printing problems with NeXTstation Color computers. If your customers with NeXTstation Color computers have not yet experienced this problem, it is likely that the problem will not occur. Note that the probability of malfunction does increase under higher temperatures. The malfunction results in skipped pages, blank pages, and partially printed pages. When you consult the Console of Workspace Manager, you will see the following message: could not open printer device Device busy np0; spurious packet received, cmd = 87 np0: DMA Underrun; Reprinting Page np0; spurious packet received, cmd = 87 np0: DMA Underrun; Reprinting Page If you identify this problem, please contact NeXT Service Operations to request a replacement CPU board. We will replace two resistors to the current version of the color CPU board to solve the problem. An updated revision of the CPU board will be distributed with new products. ***** ***** Actually, now you have to contact BellAtlantic: Bell Atlantic Computer Technology Services (CTS) 2323 Industrial Parkway West Hayward, CA 94545 1-800-499-NeXT 8am-5pm PST (510) 732-3000 (510) 732-3051 (FAX) -- Wassim M. Jabi (313) 936-0229 Doctoral Program in Architecture, University of Michigan 2000 Bonisteel Boulevard Ann Arbor Michigan 48105-2069 wjabi@libra.arch.umich.edu NeXTMail-friendly
From: frank@lucas.medicine.nwu.edu Newsgroups: comp.sys.next.sysadmin,comp.sys.next.bugs Subject: Weird NewsGrazer problem Date: 31 Jan 1994 18:00:18 GMT Organization: Northwestern University, Evanston, IL USA Message-ID: <2ijgvi$238@news.acns.nwu.edu> Hi, There is a very weird problem for my NewsGrazer (Release 2.0, V72.3) on my NeXT blackbox. I have been using this version for several months and it's the first time I encountered this kind of problem. When I was reading comp.soft-sys.sas group, suddenly, NewsGrazer crashed. After that, I tried several times. Each time, it crashed immediately. Before the weird crash, I had been reading comp.soft-sys.sas for months without any problem. I also tried other news groups and NewsGrazer worked fine. I thought it might be the problem with that specific news group, so I logged in another machine with "nn" and it worked fine with comp.soft-sys.sas. I wonder what's wrong. Any suggestion and help would be very much appreciated. Frank Zhang Programmer Analyst Reingold ECG Center Northwestern University Internet: frank@lucas.medicine.nwu.edu
From: gad@eclipse.its.rpi.edu (Garance A. Drosehn) Newsgroups: comp.sys.next.bugs Subject: Re: Weird NewsGrazer problem Date: 31 Jan 1994 23:09:06 GMT Organization: Rensselaer Polytechnic Institute, Troy NY, USA Distribution: world Message-ID: <2ik32i$73q@usenet.rpi.edu> References: <2ijgvi$238@news.acns.nwu.edu> frank@lucas.medicine.nwu.edu writes: > There is a very weird problem for my NewsGrazer (Release 2.0, V72.3) on my > NeXT blackbox. I have been using this version for several months and it's > the first time I encountered this kind of problem. > > When I was reading comp.soft-sys.sas group, suddenly, NewsGrazer crashed. > After that, I tried several times. Each time, it crashed immediately. > Before the weird crash, I had been reading comp.soft-sys.sas for months > without any problem. I also tried other news groups and NewsGrazer worked > fine. I thought it might be the problem with that specific news group, so > I logged in another machine with "nn" and it worked fine with > comp.soft-sys.sas. I wonder what's wrong. This is usually caused by a bad article in the newsgroup. There's something about the header lines that NewsGrazer is not expecting. I've seen problems when there is no "Lines:" in the header info, or if there's a "Lines:" marker but the number in there is way off. I've also seen problems due to other kinds of unexpected headers, but I don't remember the exact nature of them. Sometimes the result is that newsgrazer will hang indefinitely, and other times it causes newsgrazer to crash immediately. In some of the cases where it crashes, it only crashes if you have newsgrazer sorting the articles based on subject fields. The work around is to go into your .newsrc file and change the line for the newsgroup you're having trouble with. Change it so it looks like you've read a few articles that you didn't read thru newsgrazer, and you should be OK. -- Garance Alistair Drosehn = gad@eclipse.its.rpi.edu ITS Systems Programmer (handles NeXT-type mail) Rensselaer Polytechnic Institute; Troy NY USA
From: wjabi@libra.arch.umich.edu (Wassim M. Jabi) Newsgroups: comp.sys.next.bugs,comp.sys.next.hardware,comp.sys.next.sysadmin,comp.sys.next.software Subject: PLEASE READ: BIZARRE EXPERIENCE WITH NeXT PRINTER Date: 30 Jan 1994 11:39:23 GMT Organization: University of Michigan Engineering, Ann Arbor Distribution: usa Message-ID: <2ig69bINNlsh@srvr1.engin.umich.edu> I have a bizarre experience to report on trouble with printing on the NeXT LaserPrinter I just bought from Sam Goldberger. I have a NeXTstation Color 25MHz: - 32MB RAM - 500MB External Fujitsu HD - Software Version 3.2 (Lightning5S) - Serial Number ABC 0006073 which I will call "LIBRA" for brevity. I bought a 400DPI NeXT LaserPrinter: - NeXT 400dpi Level II Printer - Serial Number: AAC 0015251 (February 1992 M) which I will call "NLP" for brevity. When I try to print, I often, but not always, get incomplete pages after which NLP tries to re-print two more times. Sometimes it fails and other times it succeeds. The console shows the following messages while that is happening: error At this point, LIBRA shows me a dialog box informing me that some or all of the requested pages may not have been printed. ***** The bizarre part begins here ******** I have conducted *extensive* testing of LIBRA with 3 other printers: 1) NeXT 400 DPI Printer: AAC 0000164 (April 1988 T) 2) NeXT 400 DPI Printer: AAC 0013319 (November 1991 M) 3) NeXT 400 DPI Printer: AAC 0016334 (March 1992 M) Please note that the third printer is very close to my printer in manufacturing date. It was also purchased from Mr. Goldberger. LIBRA prints with no problems whatsoever with these printers. I have tested a total of about 70-100 pages and it works fine. On the other hand, NLP was connected to: 1) A NeXTstation Color Turbo Software Version 3.1 (Lightning4H) 2) A NeXTstation Color Turbo Software Version 3.1 (Lightning4H) 3) An older NeXTDimension Cube Software Version 3.0 (Hyper3B) 4) NeXTstation color 25MHz Software Version 3.2 (Lightning5S) This is similar to libra. It is actually newer than LIBRA with ROM version 2.5 v 66 M (1991) while LIBRA is 2.4. NLP printed perfectly with all four computers. I only face problems when I connect NLP to LIBRA. I have used the same printer cable In all connections. I have moved the setup to four different locations with the same results. One of the four locations was in a totally different part of town. I tried different toner cartridges: no difference. This is driving me crazy because I can not point the finger at any component. LIBRA works with three other NeXT printers without any problems and NLP works with four other NeXT computers with no problems. I have so far printed about 150-200 testpages using PrintManager, WriteNow, Preview. The only time I get a failure is when libra is connected to NLP, why???? Does any one know of an incompatibility between an older 25MHz Color NeXTstation and a 400 DPI NeXT Laser Printer manufactured February 1992? But if that is the case how come LIBRA was able to print to the other Printer which is only a month newer than NLP???? Please help in any way you can. Thank you. --- Wassim M. Jabi (313) 936-0229 Doctoral Program in Architecture, University of Michigan 2000 Bonisteel Boulevard Ann Arbor Michigan 48105-2069 wjabi@libra.arch.umich.edu NeXTMail-friendly
From: mra@bink.mit.edu (Michael Raskin Andrews) Newsgroups: comp.sys.next.bugs Subject: /usr/etc/dump Date: 1 Feb 1994 13:31:49 GMT Organization: Massachvsetts Institvte of Technology Message-ID: <2illk5$m52@senator-bedfellow.MIT.EDU> Does anyone know why my /usr/etc/dump output is something like this: bink> /usr/etc/dump 1ufs /dev/rxt0 1200000 /dev/rsd0a DUMP: Date of this level 1 dump: Mon Jan 31 18:02:08 1994 DUMP: Date of last level 0 dump: Thu Jan 27 17:13:18 1994 DUMP: Dumping /dev/rsd0a (/) to /dev/rxt0 DUMP: mapping (Pass I) [regular files] DUMP: mapping (Pass II) [directories] DUMP: mapping (Pass II) [directories] DUMP: mapping (Pass II) [directories] swent_start: No such file or directory DUMP: estimated 14110 tape blocks on 0.00 tape(s). DUMP: dumping (Pass III) [directories] DUMP: dumping (Pass IV) [regular files] DUMP: DUMP: 14287 tape blocks on 1 tape(s) DUMP: DUMP IS DONE DUMP: level 1 dump on Mon Jan 31 18:02:08 1994 DUMP: Tape rewinding Note the "swent_start" error. I looked at the file "/usr/etc/dump," and found several instances of strings beginning with "swent," and actually one instance of "swent_start" in the program --- so the bug is coming from the program text itself. Any ideas? (Could you email me so I know a reply has been given? Thanks.) --- Michael
From: mra@bink.mit.edu (Michael Raskin Andrews) Newsgroups: comp.sys.next.bugs Subject: NewsGrazer Date: 1 Feb 1994 17:35:40 GMT Organization: Massachvsetts Institvte of Technology Message-ID: <2im3tc$qg5@senator-bedfellow.MIT.EDU> Speaking about NewsGrazer, does anyone have a manual for the program?
Newsgroups: comp.sys.next.bugs From: al@rbc.uucp (Al Davis) Subject: Re: Weird NewsGrazer problem Message-ID: <1994Feb1.160503.7702@rbc.uucp> Sender: al@rbc.uucp (Al Davis) Organization: Huh? References: <2ik32i$73q@usenet.rpi.edu> Date: Tue, 1 Feb 1994 16:05:03 GMT Garance A. Drosehn writes > frank@lucas.medicine.nwu.edu writes: > > ....... > > When I was reading comp.soft-sys.sas group, suddenly, NewsGrazer crashed. > > After that, I tried several times. Each time, it crashed immediately. > > ....... Garance A. Drosehn writes > This is usually caused by a bad article in the newsgroup. > .......... > The work around is to go into your .newsrc file and change the > line for the newsgroup you're having trouble with. Change it > so it looks like you've read a few articles that you didn't > read thru newsgrazer, and you should be OK. An easier way is to use rn to read the offending group. Once you are past the offending article you can go back to NewsGrazer.
Newsgroups: comp.sys.next.bugs From: obrooks@dale.ksc.nasa.gov (Oscar Brooks) Subject: test - ignore Message-ID: <1994Jan28.151500.16596@dale.ksc.nasa.gov> Sender: news@dale.ksc.nasa.gov Organization: NASA Date: Fri, 28 Jan 1994 15:15:00 GMT -- ------------------------------------------------ Oscar Brooks Mail Code: DL-DSD-24 Kennedy Space Center, Fla. 32899 Internet: obrooks@dale.ksc.nasa.gov <NeXTMail OK>
From: robert@steffi.demon.co.uk (Robert Nicholson) Newsgroups: comp.sys.next.bugs,comp.sys.next.hardware Subject: Re: The Original NeXTstation Color Printing Bulletin Date: 1 Feb 1994 10:42:16 -0000 Organization: me organized? That's a joke! Distribution: usa Message-ID: <2ilbm8$ri@steffi.demon.co.uk> References: <2ijsoqINN73u@srvr1.engin.umich.edu> wjabi@libra.arch.umich.edu wrote in comp.sys.next.bugs,comp.sys.next.hardware >If you identify this problem, please contact NeXT Service Operations to >request a replacement CPU board. We will replace two resistors to the >current version of the color CPU board to solve the problem. An updated >revision of the CPU board will be distributed with new products. >***** ***** Could anybody, NeXT? clarify specifically which serial numbers are effected by this problem? Can we keep a database of effected serial numbers ourselves? Anybody know if ABC00014894 will be effected? -- "You know what's wrong with you?" (Audrey Hepburn, Cary Grant) "What?" "Nothing" (Charade, 1963) (ASCII for text only messages)
From: martin@trefoil.demon.co.uk (Martin Higham) Newsgroups: comp.sys.next.bugs Subject: PasteUp Date: 1 Feb 1994 20:13:46 -0000 Organization: any offers? Distribution: world Message-ID: <2imd5q$36s@trefoil.demon.co.uk> I seem to be unable to install PasteUp onto my 3.2 system. It was fine under 3.1, but I upgraded my hard disk and upgraded to 3.2 Black at the same time. When I got round to reinstalling PasteUp I just got a normal document folder instead of the PasteUp Icon and the actual application within the folder appeared as a unix application icon. Has anybody else experienced similar problems ? Martin
From: wjabi@libra.arch.umich.edu (Wassim M. Jabi) Newsgroups: comp.sys.next.bugs Subject: Thank You, Bell Atlantic (Printing Problem Solved) Date: 2 Feb 1994 19:45:23 GMT Organization: University of Michigan Engineering, Ann Arbor Distribution: usa Message-ID: <2iovsjINN168@srvr1.engin.umich.edu> Great news. Today Airborne Express delivered the new NSC CPU board from Bell Atlantic/NeXT. I installed it in 30 minutes and tried printing 22 pages. Everything worked like a charm. I have new circuitry and a new printer that works!!! I paid NOTHING! I am one HAPPY person. -- Wassim M. Jabi (313) 936-0229 Doctoral Program in Architecture, University of Michigan 2000 Bonisteel Boulevard Ann Arbor Michigan 48105-2069 wjabi@libra.arch.umich.edu NeXTMail-friendly
Newsgroups: comp.sys.next.bugs,comp.sys.next.programmer From: gerald@kurt.in-berlin.de (Gerald Erdmann) Subject: Missing headerfile in 3.2 (busvar.h) Message-ID: <1994Feb2.080803.515@kurt.in-berlin.de> Sender: root@kurt.in-berlin.de (Operator) Date: Wed, 2 Feb 1994 08:08:03 GMT Hello together, the headerfile "bsd/dev/busvar.h" which was imported by "bsd/dev/m68k/busvar.h" doesn't exist! Is it a bug or a feature? Does anybody knows a workaround or bug fix? Thanx Gerald -------------------------------------------------------------- | GERALD ERDMANN | email: gerald @ kurt.in-berlin.de (NeXTmail welcome) | voice: +49 30 372 43 10 (Germany - Berlin) | crypt: pgp2 puplic key available
Newsgroups: comp.sys.next.bugs From: kelman@niehs.nih.gov Subject: Weird bug with NS 3.2 and NewsGrazer Message-ID: <kelman.34.0@niehs.nih.gov> Sender: postman@alw.nih.gov (AMDS Postmaster) Organization: National Institutes of Health Date: Thu, 3 Feb 1994 16:15:38 GMT I've been encountering a particularly annoying bug that I think is linked with NewsGrazer running under NS 3.2. I am using DialUpSlip from home to connect to an NNTP server. About 50% of the time now, at some point the connection hangs. Nothing can go out over the line, not even telnet or ftp. Then the weird part starts. I try to su to root. I don't even get the password prompt, and the session hangs. If I try to logout, the apps and dock disappear, but the background color remains, and I never get the login window. I can still cmd-cmd-~, and reboot from the monitor, and after that the system behaves. But I am confused by something that is preventing me from using su. David Kelman kelman@niehs.nih.gov
From: wolfgang@wi.WHU-Koblenz.de (Wolfgang Roeckelein) Newsgroups: comp.sys.next.bugs Subject: Re: NewsGrazer Date: 3 Feb 1994 09:37:03 GMT Organization: WHU Koblenz Message-ID: <2iqgjv$fmk@obelix.WHU-Koblenz.de> References: <2im3tc$qg5@senator-bedfellow.MIT.EDU> Speaking about NewsGrazer, what is the up-to-date version number? -- Dipl.-Wirtsch.-Inf. Voice: +49 261 6509 173 Wolfgang Roeckelein Fax: +49 261 6509 179 WHU Koblenz E-Mail: roeckelein@wi.whu-koblenz.de Burgplatz 2 (NeXTmail ok) D-56179 Vallendar Germany
Newsgroups: comp.sys.next.bugs From: me@canis.metaworks.de (Ralf Canis) Subject: Re: NewsGrazer Message-ID: <CKnJyu.1My@canis.metaworks.de> Sender: me@canis.metaworks.de (Ralf Canis) References: <2iqgjv$fmk@obelix.WHU-Koblenz.de> Date: Thu, 3 Feb 1994 14:23:18 GMT In article <2iqgjv$fmk@obelix.WHU-Koblenz.de> wolfgang@wi.WHU-Koblenz.de (Wolfgang Roeckelein) writes: > Speaking about NewsGrazer, what is the up-to-date version number? I think it's Release 2.0, Version 72.3. -- Ralf Canis NeXTMail welcome me@canis.metaworks.de 2:2456/50.16 FAX +49-261-701852 Origin: Brain not found: Retry, Panic, Ignore...
Newsgroups: comp.sys.next.bugs,comp.sys.next.software From: antoine@arrakis.osd.ulaval.ca (Antoine Gautier) Subject: Improv, TeXTView and print: wondering what broke in 3.2 - black Message-ID: <CKnxqq.14p@athena.ulaval.ca> Sender: news@athena.ulaval.ca Organization: CTI, Universite Laval Date: Thu, 3 Feb 1994 19:20:48 GMT Dear world, Here is the problem: I have a TeX document (foo.tex) with a few psfigs (.ps format). I need to email it to another NeXT, so I print-to-file it from TeXView (-> foo.ps). So far so good. Just to check, I open the PS foo.ps with the previewer, it crashes at the first psfig generated from Improv (the others, which come from Diagram, are OK). That is, the TeX document shows up on preview until the last line before the improv-figure. Any clue? btw, the comnsole fills up with cryptic messages, but I ain't no wizard: Feb 3 13:46:48 arrakis Preview[1024]: DPS client library error: PostScript program error, DPSContext 234e8 Feb 3 13:46:48 arrakis Preview[1024]: %%[ Error: invalidrestore; OffendingCommand: restore ]%% Feb 3 13:52:14 arrakis Preview[1025]: DPS client library error: PostScript program error, DPSContext 52e58 Feb 3 13:52:14 arrakis Preview[1025]: %%[ Error: invalidrestore; OffendingCommand: restore ]%% Feb 3 13:56:54 arrakis Preview[1048]: DPS client library error: PostScript program error, DPSContext 234e8 Feb 3 13:56:54 arrakis Preview[1048]: %%[ Error: invalidrestore; OffendingCommand: restore ]%% I tried recompiling TeXView and dvips, but that did nothing. Hummmm!!! Thks! --- +-------------------------------------------------------+ | Antoine Gautier | | Departement OSD, FSA | | Universite Laval, Quebec, PQ | | Internet: Antoine@arrakis.osd.ulaval.ca | +-------------------------------------------------------+
From: hsr@cs.Stanford.EDU (Scott Roy) Newsgroups: comp.sys.next.bugs Subject: getSelectedCells: dies in IB test mode Date: 4 Feb 1994 00:49:28 GMT Organization: Stanford University: Computer Science Department Distribution: world Message-ID: <2is62o$fan@Times.Stanford.EDU> NEXTSTEP Release 3.2 NeXTstation Turbo Color Hi folks, Using getSelectedCells: in IB test mode doesn't work. It seems that InterfaceBuilder defines a category that overrides this method and does bad things. Attached is some sample code that you should plug in as a custom view on an IB palette to replicate the problem. Unfortunately, I can't seem to find a workaround. :-( --- Scott Roy Department of Computer Science Stanford University // ----------------------------------------------------------------------- @interface BumMatrix : Matrix { } @end @implementation BumMatrix - initFrame: (const NXRect *) frameRect { if (self = [super initFrame: frameRect]) { [[[self setMode: NX_LISTMODE] setAutosizeCells: NO] setPrototype: [[ActionCell alloc] initTextCell: "hello"]]; [[self renewRows: 3 cols: 3] sizeToCells]; } return self; } - mouseDown: (NXEvent *) event { id selection; [super mouseDown: event]; selection = [self getSelectedCells: nil]; fprintf(stderr, "count = %d\n", [selection count]); [selection free]; return self; } @end
From: gt6963c@prism.gatech.EDU (John "Kzin" Rudd) Newsgroups: comp.sys.next.bugs,comp.sys.next.misc,comp.sys.next.sysadmin Subject: Formatting NeXTStep Floppies under NS/Intel Message-ID: <136098@hydra.gatech.EDU> Date: 4 Feb 94 01:06:37 GMT Followup-To: comp.sys.next.bugs Organization: Georgia Institute of Technology First, I'm sorry about the crosspost, I wasn't sure which of the three groups was most appropriate. I've been trying to make a floppy formatted for nextstep (instead of dos) on my NS/I system, but it always gives me the following error: block 2448 block_count 18 The 2448 might be replaced 2556 sometimes (out of 4 or 5 tries, I wrote down the second and last message on the console). I used the workspace "Disk->Scan for Disks" menu option in all of the cases. The first few times, I had my 5.25" drive hooked up, and activated in my cmos setup. The last time, I deactivated the 5.25" drive in cmos setup, but still had it physically hooked up. I'm really kind of miffed.. for one, I'd like to make a boot floppy in case something goes wrong. for two, I don't want to have to go back and forth from Dos to Unix file systems every time I do sneaker net transfers.. I want to be able to just use the unix file system. Thanks for anyhelp John -- John "Kzin" Rudd gt6963c@prism.gatech.edu (ex-kzin@cc.gatech.edu) ===========Intel: Putting the backward in backward compatable.============== DOOM ON YOU!
From: mow@marsu.tynet.sub.org (Markus Wenzel) Newsgroups: comp.sys.next.bugs Subject: Re: NS/i, is tip borken? Date: 2 Feb 1994 23:42:49 +0100 Organization: Palumbian Research Labs Message-ID: <2ipa99$qf@marsu.tynet.sub.org> References: <CKHxAy.C77@netnews.jhuapl.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Keywords: tip jds@aplexus.jhuapl.edu (John D Stanhope) writes: >Ever since I installed 3.2 I have been unable to get tip to >transfer files larger than 20 to 30k reliably. It just drops >characters here and there. And when I connect to work I get >garbage on my screen now and then. None of this ever happened >under 3.1. >I reinstalled the system when I received 3.2, but copied over >my old remote file, so I know the tip config is the same. >Has anyone else experienced these problems, or have I just >screwed something up somewhere? Sounds like the bogus serial driver of NS/I 3.2 (the second 3.1 driver did a better job). Just get Mark Salyzyn's Mux driver via ftp, it works like charm at any speed up to 2000cps. But be careful with the installation if you use a serial mouse - then you may only replace the second serial port with the Mux driver, because the mouse driver wants the original NeXT serial driver (hardcoded). Regards. Markus. -- Marsu: "There really exist some people who live happily and satis- fied without computers." -- Frankie: "Oh no, they just emulate it!" ----- Markus Wenzel System administration, Consulting, Networking mow@marsu.tynet.sub.org on... NeXTSTEP / Unix / Novell / Windows NT IRC: Marsu Intel aside.
From: Randy Antler Newsgroups: comp.sys.next.bugs Subject: having problems with system panics? Date: 4 Feb 1994 05:42:11 GMT Organization: CERFnet Distribution: usa Message-ID: <2isn7j$nfe@news.cerf.net> Is anyone else having problems with system panics while using slip under NS 3.2? I had been running Luis Makamos (sp?) slip software under NS 3.1 for many months without a single problem. After I upgraded to 3.2...WHAMO! My system panics virtually everyday! :((( I have done some limited experimentation with bringing the slip i/f up and down and there is DEFINATELY a correlation between the slip connection going down and a system panic occuring. I've since set my inactivity timeout to a huge value to see what happens. But I'd like to return to the good old days of stability under 3.1 (and 3.0 for that matter). Any comments or recomendations? -- randy@pilot.com (home address) NeXTMAIL randy@nacm.com (work address) Welcome! *Out* and _proud_ of it! (finger me for my public key!) ________________________________________________________________________ "We are here! We are here! We are here!"
From: mra@bink.mit.edu (Michael Raskin Andrews) Newsgroups: comp.sys.next.bugs Subject: DarkForest trouble Date: 4 Feb 1994 13:50:38 GMT Organization: Massachvsetts Institvte of Technology Message-ID: <2itjre$8n4@senator-bedfellow.MIT.EDU> After playing with "DarkForest.app", the system caused itself to reboot, with the following put it /usr/adm/messages: Feb 4 08:42:23 bink mach: panic: (Cpu 0) iinactive Feb 4 08:42:23 bink mach: NeXT ROM Monitor 3.3 v74 Feb 4 08:42:23 bink mach: panic: NeXT Mach 3.0: Wed Jul 29 19:43:28 PDT 1992; root(rcbuilder):mk-127.15/BUILD/RELEASE_M68K Feb 4 08:42:23 bink mach: Feb 4 08:42:26 bink syslogd: going down on signal 15 Feb 4 08:39:04 bink syslogd: restart Any ideas?
From: Charles William Swiger <infidel+@CMU.EDU> Newsgroups: comp.sys.next.bugs Subject: Re: having problems with system panics? Date: Fri, 4 Feb 1994 11:52:58 -0500 Organization: Fifth yr. senior, Chemistry, Carnegie Mellon, Pittsburgh, PA Distribution: usa Message-ID: <QhIbpeW00WBMM3xRts@andrew.cmu.edu> In-Reply-To: <2isn7j$nfe@news.cerf.net> Excerpts from netnews.comp.sys.next.bugs: 4-Feb-94 having problems with system.. by Randy Antler@??? > [...description munched...] > > Any comments or recomendations? Perhaps you should email a bug report to Mr. Mamakos so that he knows that there may be a problem and can try to fix it? -Chuck Charles William Swiger - WhiteLight Systems | "All the world's a stage, and" --------------------------------------------+ "we are merely players...." AMS & normal mail: infidel@cmu.edu | NeXTmail: chuck@cswiger.slip.andrew.cmu.edu | "Semper ubi sub ubi."
Newsgroups: comp.sys.next.bugs From: peter@corsica (Peter Eisch) Subject: Re: netboot clients bug Message-ID: <CKH3EL.16p@news.cis.umn.edu> Sender: news@news.cis.umn.edu (Usenet News Administration) Organization: University of Minnesota Hospital and Clinic, Laboratory Information Services References: <CK9CqC.2z6@ra.nrl.navy.mil> Distribution: w Date: Mon, 31 Jan 1994 02:36:01 GMT Did you set up the client files with the 'newclient' command or try to do it manually? If you did this with newclient it's an anomaly that I've never come across tho I've been creating netboots almost monthly since 2.x. Can you find any problems with the /dev directory (or others) for that client? peter David Norton (norton@cmf.nrl.navy.mil) wrote: : If you set up a net boot client using the newclient program, you copy : files from the /usr/template/client dir. If you look at the executables : in the /usr/template/client/etc directry, they are links to : ./../usr/etc/..... These files include catman, chmod, init, etc. about : 30 in all. Once installed, they are over in /clients/<machine>/etc where : the link ../../usr/etc/...... points to no where. : This bug has been on the CD (black hardware) since the very first CD. : Hasn't anyone else reported this? : thanks : -dave -- Always looking for a handy place to nap... peter@tahiti.umhc.umn.edu (Peter Eisch) peter.a.eisch@uwrf.edu
From: Randy Antler Newsgroups: comp.sys.next.bugs Subject: Re: having problems with system panics? Date: 4 Feb 1994 21:23:31 GMT Organization: CERFnet Distribution: usa Message-ID: <2iueck$i5s@news.cerf.net> References: <QhIbpeW00WBMM3xRts@andrew.cmu.edu> In article <QhIbpeW00WBMM3xRts@andrew.cmu.edu> Charles William Swiger <infidel+@CMU.EDU> writes: |Excerpts from netnews.comp.sys.next.bugs: 4-Feb-94 having problems with |system.. by Randy Antler@??? | |> [...description munched...] |> |> Any comments or recomendations? | |Perhaps you should email a bug report to Mr. Mamakos so that he knows |that there may be a problem and can try to fix it? | |-Chuck | In an email message mark@ve6mgs.ampr.ab.ca (Mark Gregory Salyzyn) writes: I don't believe the second part of this, the 3.1 machine would panic under any serial load. You need my replacement serial driver (Mux) available via ftp, or by dropping me a note. Ciao -- Mark -- randy@pilot.com (home address) NeXTMAIL randy@nacm.com (work address) Welcome! *Out* and _proud_ of it! (finger me for my public key!) ________________________________________________________________________ "We are here! We are here! We are here!"
Newsgroups: comp.sys.next.bugs From: mark@ve6mgs.ampr.ab.ca (Mark Gregory Salyzyn) Subject: Re: having problems with system panics? Message-ID: <CKsDMy.6vM@ve6mgs.ampr.ab.ca> Organization: VE6MGS Gateway References: <QhIbpeW00WBMM3xRts@andrew.cmu.edu> <2iueck$i5s@news.cerf.net> Date: Sun, 6 Feb 1994 04:54:33 GMT Randy Antler writes: >mark writes: >>In an email message mark@ve6mgs.ampr.ab.ca (Mark Gregory Salyzyn) writes: >>I don't believe the second part of this, the 3.1 machine would panic under >>any serial load. > >>You need my replacement serial driver (Mux) available via ftp, or by >>dropping me a note. We have since found that Randy has *black* hardware, which kind of makes my advice `moot'. However, the symptoms are *close* to what would happen to me under white. Perhaps there is some simularity, but I was lead to believe that black hardware (040) was reliable to high serial speeds. Ciao -- Mark
Newsgroups: comp.sys.next.bugs From: frank@fnbc.com (Frank Mitchell) Subject: Strange DO behavior Message-ID: <1994Feb2.173015.19690@fnbc.com> Sender: news@fnbc.com Organization: First National Bank Of Chicago, Chicago IL, USA Date: Wed, 2 Feb 94 17:30:15 GMT I have a method in a protocol declared something like: - thisMethod:(bycopy in id)firstObj second:(bycopy in id)secondObj third:(bycopy in id)thirdObj; (The objects are tiny data holders; since they're input only, no reason not to copy them over.) To save a little encoding/decoding time, I decided to send thirdObj as nil whenever it is superfluous. However, when I actually sent the message, I got a NX_remoteInternalException and the Console error : Feb 2 10:34:46 cato syslog: encodeObjectBycopy: object 0 of class (null pointer) does not implement encodeUsing: Doesn't this strike anyone else as odd? Shouldn't the "copied" version of nil just be nil? Has anybody seen this before? I haven't had time to confirm that the "bycopy" keyword is the culprit, so perhaps this posting is a bit premature. It's definitely cropping up when I send the method. -- Frank Mitchell email(work): frank@fnbc.com (NeXTmail) (home): frank@gagme.chi.il.us "I would not live in any country in which I was the president." -- Tom Brokaw
From: mark@ve6mgs.ampr.ab.ca (Mark Gregory Salyzyn) Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs Subject: IOAddressRanger? Summary: Where is IOAddressRanger in the library? Keywords: IOAddressRanger Message-ID: <CKt71o.ELJ@ve6mgs.ampr.ab.ca> Date: Sun, 6 Feb 1994 15:29:47 GMT Followup-To: comp.sys.next.programmer Organization: VE6MGS Gateway In an attempt to use the various DriverKit tools associated with inspection, I found that there is *no* Library associated with the IOAddressRanger (setRangeLength,setAddressLimits,setStartAddress,init et al). I would realy prefer not to have to write all my own methods to handle a Configuration.app, and was wondering if someone knew if the Library was *hiding* somewhere (I have nm'd&string'd all the files in /lib and /usr/lib, and have perused NeXTanswers up to #1473, I have not checked if there are any new NeXTanswers for the past month). I don't want to re-invent the wheel, but I may have to anyways since I want to extend the capabilities of the IOAddressRanger to handle multiple address ranges. Ciao -- Mark
Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs,comp.sys.next.sysadmin From: thf@zelator.de (Thomas Funke) Subject: Sybase Administrator buggy under 3.x, Solution included. Message-ID: <1994Feb6.120754.878@gamelan.uucp> Sender: thomas@gamelan.uucp (thomas) Organization: Disorganization Date: Sun, 6 Feb 1994 12:07:54 GMT If you have the Sybase limited server, as distributed with 2.0, you'll notice a severe bug under NS3.x: When using the Administrator-App to start the server, the App will start 2 Terminal windows (with the same commands), starting the server twice. Of course this eventually will lead to a crash. Nevertheless, some hacking shows: There is a workaround available ! Get the 'binpatch' program and apply the following patch to the binary '/usr/sybase/Administrator' cmp -l Administrator.orig Administrator 2651 141 116 2652 377 161 2653 5 116 2654 0 161 2655 16 116 2656 42 161 As usual: No guarantee, use it with your own risk. -- ------------------------------------------------------------------ Thomas Funke ** Unix-Consultant ** thf@zelator.de Brooks's Law: Adding manpower to a late software project makes it later ------------------------------------------------------------------
Newsgroups: comp.sys.next.bugs From: smith@nextone (Howard Smith) Subject: libg++ CursesWindow problem Message-ID: <1994Feb6.213006.3515@alw.nih.gov> Keywords: libg++ Sender: postman@alw.nih.gov (AMDS Postmaster) Organization: National Institutes of Health Date: Sun, 6 Feb 1994 21:30:06 GMT Im currently having some kind of problem using the libg++ CursesWindow class. For some reason the CursesWindow destructor fails if the CursesWindow is a subwindow (ie. the curses delwin() call is failing). I wrote a simple chunk of code to test the normal curses(3X) delwin call, and it seems to work alright. The problem must be pretty serious because the debugger refuses to work after the Bus Error message. Anyone else seen this problem? System: NStation Mono 3.2 void addrec(DBM *db,CursesWindow w) { datum name,value; char keystr[128]; char valstr[128]; CursesWindow w1(w,3,40,2,25,'a'); w1.clear(); w1.box('|','-'); w1.move(1,1); w1.addstr("key:"); w1.refresh(); /* function fails here when ~CursesWindow dtor is called? */ } -- Howard C. Smith National Institute of Environmental Health Sciences 111 T.W. Alexander Drive Research Triangle Park, NC 27709 smith@nextone.niehs.nih.gov (919) 541-7594 /// I have seen the future and it is Channel 3, or was it 4? \\\
Newsgroups: comp.sys.next.bugs,comp.sys.next.software From: gemoe@proximus.north.de (Gerhard Moeller) Subject: Re: Improv, TeXTView and print: wondering what broke in 3.2 - black Content-Type: text/plain; charset=ISO-8859-1 References: <CKnxqq.14p@athena.ulaval.ca> Mime-Version: 1.0 Organization: Gerhard Moeller, German NeXT User Group, Oldenburg. Date: Sat, 5 Feb 1994 10:00:29 GMT Message-ID: <1994Feb5.100029.20544@proximus.north.de> Content-Transfer-Encoding: quoted-printable I had the same experience, TeXview.app crashed with an "segmentation fault" instantly - the TeX document compiled and displayed fine under 3.1... -- +---------------------------< principiis obsta! >---------------------------+ N Gerhard Moeller, Hetzg. 32/11, A-1030 Wien (Austria) [*: 02/21/1968] N e Private: gemoe@proximus.north.de Phone (voice): +43-1-7151531 e X Uni: Gerhard.Moeller@arbi.Informatik.Uni-Oldenburg.DE NeXTmail X T Z-Net: Gerhard.Moeller@uniol.zer encouraged! T +-> NoGeNUG - Northern German NeXT User Group: NoGeNUG@proximus.north.DE <-+
Newsgroups: comp.sys.next.bugs From: jlemon@netcom.com (Jonathan Lemon) Subject: NS/I tape driver woes Message-ID: <jlemonCKtt11.A3w@netcom.com> Keywords: SCSI Tape Archive Viper Organization: Netcom - Online Communication Services (408 241-9760 guest) Date: Sun, 6 Feb 1994 23:24:37 GMT I just got around to hooking up my tape drive to my NS/I 3.2 machine and found out that it won't read/write the tape. In particular, whenever I try using tar/cpio/dd to read/write the tape, it always gives me an I/O error. All the 'mt' commands work fine, however. I'm pretty sure that it isn't a problem with the drive or the media, since the drive works just fine under DOS. Particulars: Archive Viper 150 SCSI tape drive Adaptec 1742 SCSI controller The tape is at LUN 4; LUN 3 is my CD-ROM, LUN 0, 1, 2 are disks. I checked NextAnswers, and couldn't find anything other than the SCSITape patch for 3.1. (which I assume I don't need since I'm running 3.2) Any pointers, hints, or solutions would be appreciated. -- Jonathan
Control: cancel <CKt71o.ELJ@ve6mgs.ampr.ab.ca> Newsgroups: comp.sys.next.bugs From: mark@ve6mgs.ampr.ab.ca (Mark Gregory Salyzyn) Subject: cancel <CKt71o.ELJ@ve6mgs.ampr.ab.ca> Message-ID: <CKu2o5.Htu@ve6mgs.ampr.ab.ca> Organization: VE6MGS Gateway Date: Mon, 7 Feb 1994 02:52:48 GMT cancel <CKt71o.ELJ@ve6mgs.ampr.ab.ca> in newsgroup comp.sys.next.bugs
From: sanguish@digifix.com Newsgroups: comp.sys.next.announce,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.advocacy,comp.sys.next.software,comp.sys.next.bugs,comp.sys.next.programmer,comp.sys.next.marketplace Subject: Quick Guide to NEXTSTEP information on the Internet Date: 6 Feb 1994 23:40:06 -0500 Organization: Next Announcements Message-ID: <2j4gn6$909@digifix.digifix.com> This post is made weekly, to help 'point' users to more NEXTSTEP information Topics include: comp.sys.next newsgroups related newsgroups comp.sys.next newsgroups mailing list ftp sites NeXTanswers comp.sys.next.* newsgroups -------------------------- Comp.Sys.Next.Advocacy This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. Comp.Sys.Next.Announce Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) The NEXTSTEP FAQs are posted here monthly as well. This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Comp.Sys.Next.Bugs A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT- specific groups as well. Comp.Sys.Next.Hardware Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. Comp.Sys.Next.Marketplace NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. Comp.Sys.Next.Misc For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! Comp.Sys.Next.Programmer Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. The NEXTSTEP programmer FAQs are posted here. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. Comp.Sys.Next.Software This is a place to talk about [third party] software products that run on NEXTSTEP systems. Comp.Sys.Next.Sysadmin Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. related Newsgroups ------------------ Comp.Soft-Sys.Nextstep Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. Comp.Lang.Objective-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. Comp.Object Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original Comp.Sys.Next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News ------------------------------------------- Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep-d next-advocacy-d next-announce-d next-bugs-d next-hardware-d next-marketplace-d next-misc-d next-programmer-d next-software-d next-sysadmin-d (For a full description, send mail saying LISTS to <digestif@antigone.com>). The subscription syntax is essentially the same as LISTSERV's. To subscribe, send a message to <digestif@antigone.com> saying: SUB Listname YourName Example: SUB next-hardware-d John Doe The ftp sites ------------- cs.orst.edu: The main site for North American submissions nova.cc.purdue.edu: Lots of older stuff, but very short on disk space ftp.informatik.uni-muenchen.de: In Germany. terra.stack.urc.tue.nl (Dutch NEXTSTEP User Group) and cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp.next.com: See the below ftp.next.com and NextAnswers@next.com ------------------------------------- From the document 1000_Help from ftp.next.com Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, or you can transfer them by anonymous ftp. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to NeXTanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO <your-address> If you have any problem with the system or suggestions for improvement, please send mail to NeXTanswers-request@NeXT.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to NeXTanswers-request@NeXT.com. Written by: Eric P. Scott eps@toaster.SFSU.EDU and Scott Anguish sanguish@digifix.com Additions from: Greg Anderson (Greg_Anderson@afs.com) and Michael Pizolato (Michael_Pizolato@afs.com)
Newsgroups: comp.sys.next.bugs From: jlc@alien.att.com (John L. Chmielewski) Subject: Re: NS/I tape driver woes Organization: AT&T Date: Mon, 7 Feb 1994 05:15:22 GMT Message-ID: <CKu99v.Ds9@cbnewse.cb.att.com> Keywords: SCSI Tape Archive Viper References: <jlemonCKtt11.A3w@netcom.com> Sender: news@cbnewse.cb.att.com (NetNews Administrator) In article <jlemonCKtt11.A3w@netcom.com> jlemon@netcom.com (Jonathan Lemon) writes: > > I just got around to hooking up my tape drive to my NS/I 3.2 machine >and found out that it won't read/write the tape. > > In particular, whenever I try using tar/cpio/dd to read/write the tape, >it always gives me an I/O error. All the 'mt' commands work fine, however. >I'm pretty sure that it isn't a problem with the drive or the media, since >the drive works just fine under DOS. > > Particulars: Archive Viper 150 SCSI tape drive > Adaptec 1742 SCSI controller > > The tape is at LUN 4; LUN 3 is my CD-ROM, LUN 0, 1, 2 are disks. > > I checked NextAnswers, and couldn't find anything other than the SCSITape >patch for 3.1. (which I assume I don't need since I'm running 3.2) > > Any pointers, hints, or solutions would be appreciated. The problem is that the tape driver needs to be set to a fixed block size of 512 bytes for many cartridge tape drives, otherwise "i/o errors" is returned. I wrote setmtd.c, for NeXT hardware in 1991, to set the SCSI tape driver to a fixed block size of 512 bytes. It is still needed for NEXTSTEP 3.2, NeXT or Intel hardware. After compiling and running it, you should be able to read and write your tape drive. The program needs to be run after each reboot, so I added it to /etc/rc.local. The program has been around a long time and it has been improved upon by several other people. If you check the archives, you should find it and an improved version in a collection of scsi programs called scsitools.tar.Z. Here is the program, modified to run under NEXTSTEP 3.2: ------------------- start of setmtd.c ---------------------- /* * setmtd - set SCSI tape driver to a fixed block size * by John L. Chmielewski * Tue Feb 19, 1991 * Modified: Mon Aug 30, 1993 by jlc for NEXTSTEP 3.1 * cc -o setmtd -arch m68k -arch i386 setmtd.c */ #include <fcntl.h> #include <sys/types.h> #include <dev/scsireg.h> #define RSTDEVICE "/dev/rst0" #define BLOCKSIZE 512 main() { int fd, size = BLOCKSIZE; if ((fd = open(RSTDEVICE, O_RDWR)) < 0) { perror(RSTDEVICE); exit(1); } if (ioctl(fd, MTIOCFIXBLK, &size) < 0) { perror("ioctl"); exit(1); } (void) close(fd); return 0; } ------------------- end of setmtd.c ---------------------- John jlc@attmail.com
From: ranilla@polar.etsiig.uniovi.es (Ranilla Pastor) Newsgroups: comp.sys.next.bugs Subject: problems with HSD SCAN-X Color Message-ID: <1994Feb7.203431.1391@polar> Date: 7 Feb 94 20:34:31 +0100 Organization: Universidad de Oviedo About one and a half year ago I bought a HSD SCAN-X Color for my 68040 NeXT Cube. Currently I am transfering my work to a Intel 486 based running MICROSOFT WINDOWS 3.1 environment, because my Next Cube lost the login window and now I can't use it. Therefore, I would like to use the scan on the 486 Pc. Unfortunately I haven't the drivers software for the 486 Pc and I can't use the scanner. Does someone know about drivers software for the HSD SCAN-X Color on the Intel 486? or Does soemone know about the technical details of the scanner (instruction code, class, ADF, etc), or how to get these informations? Thanks in advance, Ranilla PD: Best way to reply to me is email PD: Excuse my (non-existent :-) english and my poor dictionary. -------------------------------------------------------------------------------- Jose Ranilla Pastor Phone: +34 9 5182129 Area de Ciencias de la Computacion e I.A. Fax: +34 9 5338538 E.T.S.I.I de Gijon EMail: ranilla@polar.etsiig.uniovi.es Campus de Viesques s/n Gijon \,,,/ Spain (o o) ----------------------------------------------------oOO--(_)--OOo---------------
From: richard%runner.uucp@usc.edu (Richard Ruth) Newsgroups: comp.sys.next.bugs Subject: Re: Sybase Administrator buggy under 3.x, Solution included. Date: 7 Feb 1994 18:09:25 -0800 Organization: runner Sender: richard@runner.uucp@usc.edu Distribution: world Message-ID: <2j6s8l$763@runner.uucp> References: <1994Feb6.120754.878@gamelan.uucp> Thomas Funke writes ... > Nevertheless, some hacking shows: There is a workaround available ! > > Get the 'binpatch' program and apply the following patch to the > binary '/usr/sybase/Administrator' > ... Anyone else try this? and, where do you get 'binpatch' for NeXT hardware? Thanks -- Richard richard%runner.uucp@usc.edu (ok to send NeXT Mail)
From: philipp@res.enst.fr (Philippe-Andre Prindeville) Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs Subject: NS/fip and SCSI "synchronous" transfers Date: 7 Feb 1994 17:42:54 GMT Organization: Ecole Nationale Superieure des Telecommunications, Paris FRANCE Distribution: world Message-ID: <2j5uiu$3l1@enst.enst.fr> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit I recently picked up a copy of Oliver Unter Ecker's DriverPerformance.app from ORST and tried it on my machine, wondering if I would really see as good numbers as the data sheet for my Maxtor MXT540S drive announced (it is supposedly capable of several MB/s transfers). Well, the transfer rates I saw were 600KB/s to 850KB/s -- far short of what I was expecting. Then I remembered something about Synchronous transfers not being supported by NeXT drivers (I have an Adaptec 1542CF which is supposedly supported). Curious, I thought. Why did NeXT ignore synchronous transfers? Granted that my disk controller is ISA (and not VL or EISA), I should still be getting better performance that what I'm seeing. Is it because of the PC's slow DMA (I'm using a 16bit channel)? Is it because the transfers are asynchronous? Or is it the Mach (BSD) filesystem that just isn't fast? -Philip
Newsgroups: comp.sys.next.bugs From: mckelvey@fafnir.com (James W. McKelvey) Subject: Compress warning! Message-ID: <CKsDH6.F96@fafnir.com> Sender: mckelvey@fafnir.com (James W. McKelvey) Organization: Fafnir Software, Saugus, CA, USA Date: Sun, 6 Feb 1994 04:51:06 GMT Do NOT use the "compress" workspace feature if you care about your data. I've used it dozens of times; but when I recently tried to compress a large directory, the directory simply disappeared! And with no error messages (even in the console). I've lost several hours worth of work, plus the cost of downloading the original data in the first place. Thanks a lot, NeXT. -- What I did must have been right, or an angel would have stopped me, just as the angel stopped Abraham in the Bible. -- Albert Fish, the cannibal Jim McKelvey mckelvey@fafnir.com
Newsgroups: comp.sys.next.bugs From: mckelvey@fafnir.com (James W. McKelvey) Subject: Re: Missing headerfile in 3.2 (busvar.h) Message-ID: <CKtLJ6.HGC@fafnir.com> Sender: mckelvey@fafnir.com (James W. McKelvey) Organization: Fafnir Software, Saugus, CA, USA References: <1994Feb2.080803.515@kurt.in-berlin.de> Date: Sun, 6 Feb 1994 20:42:41 GMT In article <1994Feb2.080803.515@kurt.in-berlin.de> gerald@kurt.in-berlin.de (Gerald Erdmann) writes: > Hello together, > > the headerfile "bsd/dev/busvar.h" which was imported by > "bsd/dev/m68k/busvar.h" doesn't exist! > > Is it a bug or a feature? Does anybody knows a workaround or bug fix? > I just reported this to NeXT yesterday. I worked around it by using the 3.0 version of bsd/dev/m68k/busvar.h (which doesn't import a bsd/dev/busvar.h). I suspect it's a bug. It's in 3.1 also. If they wanted to remove the capability, they would have removed both of the busvar.h's. Or maybe the mistake was in not removing both of them? -- What I did must have been right, or an angel would have stopped me, just as the angel stopped Abraham in the Bible. -- Albert Fish, the cannibal Jim McKelvey mckelvey@fafnir.com
From: wmorse@erasure-sl.cc.emory.edu (William Morse) Newsgroups: comp.sys.next.bugs Subject: Re: Compress warning! Date: 9 Feb 1994 01:24:29 GMT Organization: Emory University, Dept of Math and CS Distribution: world Message-ID: <2j9e0dINNhv4@emory.mathcs.emory.edu> References: <CKsDH6.F96@fafnir.com> In article <CKsDH6.F96@fafnir.com> mckelvey@fafnir.com (James W. McKelvey) writes: > Do NOT use the "compress" workspace feature if you care about your > data. I've used it dozens of times; but when I recently tried to > compress a large directory, the directory simply disappeared! And with > no error messages (even in the console). I've lost several hours worth > of work, plus the cost of downloading the original data in the first > place. Thanks a lot, NeXT. > -- > What I did must have been right, or an angel would > have stopped me, just as the angel stopped Abraham > in the Bible. -- Albert Fish, the cannibal > Jim McKelvey mckelvey@fafnir.com ----- This can happen if you try to compress a directory after you have just started another compression on the same directory. In other words, if you push the compress button twice in a row. The same can happen in a decompress. I think this is in the release notes... William Consult. / ITD
Newsgroups: comp.sys.next.bugs From: mark@ve6mgs.ampr.ab.ca (Mark Gregory Salyzyn) Subject: Mux V1.3 (V1.2, V1.1 and Mux8 V1.0) fix for DCD handling Message-ID: <CKxxuu.DLo@ve6mgs.ampr.ab.ca> Organization: VE6MGS Gateway Date: Wed, 9 Feb 1994 04:59:16 GMT James Cassidy <jcassidy@quark.uwaterloo.ca> has pointed out a problem in my Serial Mux driver that may require an immediate fix for those that are programming to use the driver. I had the driver check for DCD for every entry point except /dev/cu??. This is incorrect and does not match up with the way the SerialPorts driver handles this issue. The fix is the following: Replace the line in Mux_reloc.tproj/Mux.m: #define MUX_LOCAL(x) (minor(x)&128) with: #define MUX_LOCAL(x) ((minor(x)&(128|64))!=64) and that should fix it! Thanks, and I am thankful for the feedback! Ciao -- Mark
Newsgroups: comp.sys.next.bugs From: wfc@cl.cam.ac.uk (William Clocksin) Subject: bug in code generation? Message-ID: <1994Feb9.141945.6457@infodev.cam.ac.uk> Sender: news@infodev.cam.ac.uk (USENET news) Organization: U of Cambridge, England Date: Wed, 9 Feb 1994 14:19:45 GMT Bug? I have a C program which ran correctly under 2.1 and 3.0, but will not run under 3.2. Given float y, unsigned int p.b and int k, where p.b is a bit field, y + p.b * k returns the incorrect value under 3.2, but int m = p.b; y + m * k returns the correct value. This is so basic to code generation that I cannot believe the compiler is broken. Can anyone else duplicate this? ---------------------------------------------------- Dr W.F. Clocksin, Assistant Director of Research, Computer Laboratory, University of Cambridge Pembroke Street, Cambridge CB2 3QG, U.K. Internet: wfc@CL.cam.ac.uk (NeXTMail is acceptable) Tel UK: (0223) 334628. Tel (Int'l): +44 223-334628. Fax UK: (0223) 334678. Fax (Int'l): +44 223-334678.
Newsgroups: comp.sys.next.bugs From: wfc@cl.cam.ac.uk (William Clocksin) Subject: definitely a bug in code generation Message-ID: <1994Feb9.152209.10365@infodev.cam.ac.uk> Sender: news@infodev.cam.ac.uk (USENET news) Organization: U of Cambridge, England Date: Wed, 9 Feb 1994 15:22:09 GMT The following short program will demonstrate a bug in code generation under NS 3.2: struct { unsigned int nlines : 3; unsigned int spacing : 4; unsigned int subtype : 2; unsigned int size : 2; unsigned int haspref : 1; unsigned int hidden : 1; } flags; main() { float y = 1701.0; int k = -3; float x; flags.spacing = 2; x = y + flags.spacing * k; printf("x = %f\n", x); } Under 3.2, it will print 4294968832.000000 instead of 1695.000000. It works OK under previous releases. It also works OK if you set k = 3, getting (correctly) 1707. It also works OK if you assign flags.spacing to a temp variable and use the temp variable in an expression. Sorry if this is a known bug or in a FAQ. ---------------------------------------------------- Dr W.F. Clocksin, Assistant Director of Research, Computer Laboratory, University of Cambridge Pembroke Street, Cambridge CB2 3QG, U.K. Internet: wfc@CL.cam.ac.uk (NeXTMail is acceptable) Tel UK: (0223) 334628. Tel (Int'l): +44 223-334628. Fax UK: (0223) 334678. Fax (Int'l): +44 223-334678.
Newsgroups: comp.sys.next.bugs From: wfc@cl.cam.ac.uk (William Clocksin) Subject: Re: definitely a bug in code generation Message-ID: <1994Feb9.160012.13268@infodev.cam.ac.uk> Sender: news@infodev.cam.ac.uk (USENET news) Organization: U of Cambridge, England References: <1994Feb9.152209.10365@infodev.cam.ac.uk> Date: Wed, 9 Feb 1994 16:00:12 GMT Just to improve on my own article: In article <1994Feb9.152209.10365@infodev.cam.ac.uk> wfc@cl.cam.ac.uk (William Clocksin) writes: > > The following short program will demonstrate a bug in code generation under NS 3.2 [stuff deleted] A shorter program to demonstrate the bug is: main() { unsigned int f = 3; int k = -3; float y = 1701.0; printf("%f\n", y + f * k); } So multiplying an unsigned int by a negative int and then adding that to a float will return trash. How many apps out there do this? It works OK if you don't add the float! Sorry again if this is a known bug or in a FAQ. ---------------------------------------------------- Dr W.F. Clocksin, Assistant Director of Research, Computer Laboratory, University of Cambridge Pembroke Street, Cambridge CB2 3QG, U.K. Internet: wfc@CL.cam.ac.uk (NeXTMail is acceptable) Tel UK: (0223) 334628. Tel (Int'l): +44 223-334628. Fax UK: (0223) 334678. Fax (Int'l): +44 223-334678.
From: averill@bela.music.washington.edu (Ron Averill) Newsgroups: comp.sys.next.bugs Subject: System panics in 3.1 Date: 9 Feb 1994 17:58:15 GMT Organization: University of Washington Message-ID: <2jb87n$5a6@news.u.washington.edu> We've been getting system panics on a regular basis since upgrading to 3.1. With 3.0 they happened about once a year. One of the machines in our network is running 3.0 while the others run 3.1. Could this cause the problem? Any ideas? Ron Averill University of Washington School of Music Computer Center raverill@u.washington.edu
From: jason_fosback@psca.com (Jason Fosback) Newsgroups: comp.sys.next.misc,comp.sys.next.bugs Subject: DOOM problem: STBAR not found Date: 9 Feb 1994 23:38:58 GMT Organization: McCaw Cellular Communications, Inc. Message-ID: <2jbs6i$ipd@ftp-p.mccaw.com> Like many people, I grabbed the latest version of Doom off the archives yesterday. I was very excited that it was FAT, but I was more excited that I could get all of the levels by copying the DOS doom.wad file into the NeXTSTEP app wraper. Well, I replaced doom1.wad with the full doom.wad (calling it doom1.wad). Things looked promising, and Doom reported that I had a registered version. But then, I got the following error during the status bar initialization: W_GetNumForName: STBAR Not Found! What's the scoop with this? The only thing I can think of is that the DOS version of Doom is 1.1, and the NeXTSTEP version is 1.2. Anyone know what the scoop is? Thanks! -jason ____________________________________________________________ Jason Fosback, Systems Analyst | No sir, I didn't like it --- Paradigm Systems Corp --- | -R&S Internet: jason.fosback@psca.com | Star Trek: NeXT mail: jason.fosback@psca.com | The NeXT Generation...
From: robert@steffi.demon.co.uk (Robert Nicholson) Newsgroups: comp.sys.next.bugs Subject: cmsg cancel <2jbrpo$38g@steffi.demon.co.uk> Control: cancel <2jbrpo$38g@steffi.demon.co.uk> Date: 9 Feb 1994 23:35:45 -0000 Organization: me organized? That's a joke! Message-ID: <2jbs0h$3dr@steffi.demon.co.uk> <2jbrpo$38g@steffi.demon.co.uk> was cancelled from within trn. -- "You know what's wrong with you?" (Audrey Hepburn, Cary Grant) "No, what?" "Nothing" (Charade, 1963) (ASCII for text only messages)
From: jeffreyh@scalable.com (Jeffrey Heller) Newsgroups: comp.sys.next.bugs Subject: getwd syscall looks broken under posix Date: 09 Feb 1994 19:50:47 GMT Organization: Scalable Technologies Inc. Distribution: world Message-ID: <JEFFREYH.94Feb9115047@frodo.scalable.com> I porting some code to the NeXT using the posix compiler flag I have found that the getwd call almost always fails. Be careful to use the getpwd call when using the -posix flag. The posix mode seems to have some gotcha's. Anyone else find things to look out for when using it? -Jeff -- =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= =- Jeffrey Heller -= =- Scalable Technologies, Inc. -= =- 1975 West El Camino Real, Suite 102, Mountain Vaiw, Ca 94040 -= =- Phone: (415) 965-5873 Email: jeffreyh@scalable.com -= -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=--=
From: gcolello@biosphere.Stanford.EDU (Greg Colello) Newsgroups: comp.sys.next.bugs Subject: Another C Code Generation Error? Date: 10 Feb 1994 04:45:32 GMT Organization: Stanford University Distribution: world Message-ID: <2jce5c$b71@nntp2.Stanford.EDU> Can some of you please check whether the following program gets a segmentation fault on their machine before I report it as a bug to NeXT? It would be too difficult to explain what it's doing. I just need to know if it works or not. We write large and small C programs all the time and have never had a problem like this. It produces a segmentation fault on all of our Next's running 3.0. We have not loaded 3.2 yet. It would be nice to know if it works under 3.2. It runs fine on our Sun running SunOS Release 4.1.3 (GENERIC) with ANSI C (acc) or our DEC workstation. We have traced the code with gdb and got nonsense results. It appears the machine code or memory map generated is faulty in some way. I suspect there is a bug in the program that DEC and SUN are more permissive about. I don't know. This is one of those problems that changing the number of lines in the C file can change its behavior, but only on the NeXT's. Anyway to test this, which takes about 3 minutes to setup, you will need to create a subdirectory to put three very small data files and one C source file into. The files are listed below preceeded by their name. The compile line is commented at the top of the C source file. After compilation type "limit stacksize 5M" to eliminate this as a possible problem. Type "svc" to run the program. You should get the following messages on your terminal: terminal messages: --------------------------------------- filename=setup comment char: '#' obs fln: 'file1' onx = 1 ony = 2 sim fln: 'file2' snx = 6 sny = 2 number of data points in the observation file: 1 Segmentation fault --------------------------------------- The files are listed below preceeded by their name: filename = "setup" (cut between lines and paste) --------------------------------------- # file1 1 2 file2 6 2 --------------------------------------- filename = "file1" (cut between lines and paste) --------------------------------------- 187.5 1.0 --------------------------------------- filename = "file2" (cut between lines and paste) --------------------------------------- 1987 7 6 00 00 187 1987 7 7 00 00 188 --------------------------------------- filename = "svc.c" (cut between lines and paste) --------------------------------------- #include <time.h> #include <fcntl.h> #include <stdio.h> #include <stdlib.h> #include <string.h> #include <math.h> #define TRUE 1 #define FALSE 0 #define MAXF 10000 #define LBUF 30 /* -------------------------------------- program svo Reads the input file whose name is given as a command line parameter In that file, two file names are given, one for the observation file, one for the simulation file, and for each file the index of the x and y columns. Then the simulation file is interpolated, one point for each of the observed points. Compile Sun: acc -o svo svo.c Compile NEXT: cc -ansi -Wall -o svo svo.c Usage: svo -------------------------------------- */ void main (int argc, char *argv[]) { char chr; char ofln[200], sfln[200], cmt; int onx,ony,snx,sny; double ox[MAXF], oy[MAXF], sx[MAXF], sy[MAXF], sipol; char oxbuf[LBUF], oybuf[LBUF], sxbuf[LBUF], sybuf[LBUF]; int ioxb = 0, ioyb = 0, isxb = 0, isyb = 0; int iod = 0, isd = 0, nol = 0, nsl = 0; int first = TRUE, inword = FALSE, skip = FALSE, jf = 0, nl = 0; FILE *stp, *ofp, *sfp; printf("filename=%s\n","setup"); stp = fopen("setup","r"); if (stp == NULL) { printf("File could not be opened: %s\n","setup"); exit(1); } fscanf(stp,"%c", &cmt); fscanf(stp,"%s %d %d", ofln, &onx, &ony); fscanf(stp,"%s %d %d", sfln, &snx, &sny); fclose(stp); printf("comment char: '%c'\n", cmt); printf("obs fln: '%s' onx = %3d ony = %3d\n", ofln, onx, ony); printf("sim fln: '%s' snx = %3d sny = %3d\n", sfln, snx, sny); printf("\n"); /* now read in the observed file */ ofp = fopen(ofln, "r"); if (ofp == NULL) { printf("File could not be opened: %s\n",ofln); exit(1); } /* Read the observed file line by line and fill the arrays */ while ((chr = getc(ofp)) != EOF) { if (first) { first = FALSE; if (chr == cmt) { skip = TRUE; } else if (chr != '\n') { skip = FALSE; } } if (!skip) { if (chr == '\n') { first = TRUE; inword = FALSE; ++nol; oxbuf[ioxb] = '\0'; oybuf[ioyb] = '\0'; if ((jf < onx) || (jf < ony)) { printf("missing observed field on line = %5d \n", nol); return; } ox[iod] = atof(oxbuf); /* multiply by .4 for kf2 files only. Otherwise just atof */ /* oy[iod] = atof(oybuf) * .4; */ oy[iod] = atof(oybuf); /*printf("%5d %10.5f %10.5f\n",iod,ox[iod],oy[iod]);*/ ++iod; jf = 0; ioxb = 0; ioyb = 0; } else if ((chr == ' ') || (chr == '\t')) { inword = FALSE; } else if (inword == FALSE) { inword = TRUE; ++jf; if (jf == onx) { oxbuf[ioxb] = chr; ++ioxb; } if (jf == ony) { oybuf[ioyb] = chr; ++ioyb; } } else if (inword == TRUE) { if (jf == onx) { oxbuf[ioxb] = chr; ++ioxb; } else if (jf == ony) { oybuf[ioyb] = chr; ++ioyb; } } } if (skip && (chr == '\n')) { first = TRUE; skip = FALSE; } } nol = iod; printf("\nnumber of data points in the observation file: %d\n",nol); fclose(ofp); /* now read in the simulated file */ sfp = fopen(sfln, "r"); if (sfp == NULL) { printf("File could not be opened: %s\n",sfln); exit(1); } /* Read the simulated file line by line and fill the arrays */ first = TRUE; inword = FALSE; jf = 0; while ((chr = getc(sfp)) != EOF) { if (first) { first = FALSE; if (chr == cmt) { skip = TRUE; } else if (chr != '\n') { skip = FALSE; } } if (!skip) { if (chr == '\n') { first = TRUE; inword = FALSE; ++nsl; sxbuf[isxb] = '\0'; sybuf[isyb] = '\0'; if ((jf < snx) || (jf < sny)) { printf("missing simulated field on line = %5d \n", nsl); return; } sx[isd] = atof(sxbuf); sy[isd] = atof(sybuf); /*printf("%5d %10.5f %10.5f\n",isd,sx[isd],sy[isd]);*/ ++isd; jf = 0; isxb = 0; isyb = 0; } else if ((chr == ' ') || (chr == '\t')) { inword = FALSE; } else if (inword == FALSE) { inword = TRUE; ++jf; if (jf == snx) { sxbuf[isxb] = chr; ++isxb; } if (jf == sny) { sybuf[isyb] = chr; ++isyb; } } else if (inword == TRUE) { if (jf == snx) { sxbuf[isxb] = chr; ++isxb; } else if (jf == sny) { sybuf[isyb] = chr; ++isyb; } } } if (skip && (chr == '\n')) { first = TRUE; skip = FALSE; } } nsl = isd; printf("number of data points in the simulation file: %d\n\n",nsl); fclose(sfp); /* Now for each observed point find the pair of simulated points it is sandwiched in, and interpolate. */ for (iod = 0; iod < nol; ++iod) { for (isd = 0; sx[isd] < ox[iod]; ++isd) ; if (isd == 0) { isd = 1; } sipol = ((sy[isd]-sy[isd-1])/(sx[isd]-sx[isd-1]))*(ox[iod]-sx[isd]) + sy[isd]; /* printf("%3d %10.5f %3d %5.2f %5.2f %6.5f %6.5f %6.5f %6.5f\n", iod,ox[iod],isd,sx[isd-1],sx[isd],oy[iod ],sy[isd-1],sy[isd], sipol);*/ printf("%e %e %e\n",ox[iod],oy[iod],sipol); } } --------------------------------------- ----------------------------------------------------------------- Greg Colello Carnegie Institution, Department of Plant Biology Stanford University gcolello@biosphere.stanford.edu (NeXT mail OK)
From: jacobsen@arundel.doit.wisc.edu Newsgroups: comp.sys.next.bugs Subject: Re: Another C Code Generation Error? Date: 10 Feb 1994 05:49:42 GMT Organization: Division of Information Technology Message-ID: <2jchtm$gtv@news.doit.wisc.edu> References: <2jce5c$b71@nntp2.Stanford.EDU> In article <2jce5c$b71@nntp2.Stanford.EDU> gcolello@biosphere.Stanford.EDU (Greg Colello) writes: > Can some of you please check whether the following program gets a > segmentation fault on their machine before I report it as a bug to NeXT? > It would be too difficult to explain what it's doing. I just need to know > if it works or not. We write large and small C programs all the time and > have never had a problem like this. <<<test program and instructions deleted>>> > Greg Colello > Carnegie Institution, Department of Plant Biology > Stanford University > gcolello@biosphere.stanford.edu (NeXT mail OK) I ran your program on my NeXTStation Color (NS3.2) with no problems: arundel> svo filename=setup comment char: '#' obs fln: 'file1' onx = 1 ony = 2 sim fln: 'file2' snx = 6 sny = 2 number of data points in the observation file: 1 number of data points in the simulation file: 2 1.875000e+02 1.000000e+00 7.000000e+00 arundel> 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: 25 MHz Processor active: 0 System type: 3 Board revision: 0x0 Primary memory available: 16.00 megabytes. Default processor set: 55 tasks, 91 threads, 1 processors Load average: 0.20, Mach factor: 0.78 arundel> It looks like this was fixed somewhere between 3.0 and 3.2. --Erik
Newsgroups: comp.sys.next.bugs From: wfc@cl.cam.ac.uk (William Clocksin) Subject: Retraction: not bug in code generation! Message-ID: <1994Feb10.092405.22373@infodev.cam.ac.uk> Sender: news@infodev.cam.ac.uk (USENET news) Organization: U of Cambridge, England Date: Thu, 10 Feb 1994 09:24:05 GMT The behaviour that I posted as a bug (and fortunately did not officially report yet) is not a bug. According to Appendix A of K&R para 6.5, combining signed and unsigned operands converts the signed operand to unsigned. I had thought it did not convert, but performed signed arithmetic yielding a signed result. I must have been thinking of some other language! ---------------------------------------------------- Dr W.F. Clocksin, Assistant Director of Research, Computer Laboratory, University of Cambridge Pembroke Street, Cambridge CB2 3QG, U.K. Internet: wfc@CL.cam.ac.uk (NeXTMail is acceptable) Tel UK: (0223) 334628. Tel (Int'l): +44 223-334628. Fax UK: (0223) 334678. Fax (Int'l): +44 223-334678.
Organization: The American University - University Computing Center Date: Wed, 9 Feb 1994 20:30:43 EST From: David Ritz <DR8831A@auvm.american.edu> Message-ID: <94040.203043DR8831A@auvm.american.edu> Newsgroups: comp.sys.next.bugs Subject: Any problems with HP Laserjet Series 4M printers ? Has anyone had any problems with Next Step 3.2 running on an Intel DX2 66 Mhz p rocessor and a parallel port HP Laser jet Series 4M ? Please reply to dr8831a@american.edu
Newsgroups: comp.sys.next.bugs From: emarinos@marcon.ka.sub.org (Evstathios Marinos) Subject: Re: definitely a bug in code generation Message-ID: <1994Feb10.075901.20822@marcon.ka.sub.org> Sender: emarinos@marcon.ka.sub.org (Evstathios Marinos) Organization: Marinos EDV - Consulting References: <1994Feb9.160012.13268@infodev.cam.ac.uk> Date: Thu, 10 Feb 1994 07:59:01 GMT In article <1994Feb9.160012.13268@infodev.cam.ac.uk> wfc@cl.cam.ac.uk (William Clocksin) writes: \\ Just to improve on my own article: \\ \\ In article <1994Feb9.152209.10365@infodev.cam.ac.uk> wfc@cl.cam.ac.uk (William \\ Clocksin) writes: \\ > \\ > The following short program will demonstrate a bug in code generation under \\ NS 3.2 \\ [stuff deleted] \\ \\ A shorter program to demonstrate the bug is: \\ \\ main() \\ { \\ unsigned int f = 3; \\ int k = -3; \\ float y = 1701.0; \\ printf("%f\n", y + f * k); \\ } \\ \\ So multiplying an unsigned int by a negative int and then adding that to a \\ float will return trash. How many apps out there do this? \\ \\ It works OK if you don't add the float! \\ \\ Sorry again if this is a known bug or in a FAQ. \\ \\ ---------------------------------------------------- \\ Dr W.F. Clocksin, Assistant Director of Research, \\ Computer Laboratory, University of Cambridge \\ Pembroke Street, Cambridge CB2 3QG, U.K. \\ Internet: wfc@CL.cam.ac.uk (NeXTMail is acceptable) \\ Tel UK: (0223) 334628. Tel (Int'l): +44 223-334628. \\ Fax UK: (0223) 334678. Fax (Int'l): +44 223-334678. It is not a bug in 3.2. If it worked in prior version you had luck, because that was a bug according to the C Language Definition. The statement "float + int * unsigned int" is processed in the following way: First: "int * unsigned int". In this special case C says, if one operand is unsigned int, the other operand is converted also to unsigned int and the result is also unsigned int. So your example result is not -9, it is 4294967287. Second: "float + unsigned int" In this case unsigned int is converted to a float, because one of the operands is float. So you are actually doing "1701.0 + 4294967287.0". A correct result, but not the result you expected. To get the expected result you have to do a cast: "float + (int) (int * unsigned int)" Best regards Stathis -- Marinos EDV - Consulting | NEXTSTEP Software and Consulting Services Evstathios Marinos | Phone : +49 721 37 71 78 Gartenstr. 2 | Fax : +49 721 37 71 79 76133 Karlsruhe (GERMANY) | E-Mail: emarinos@marcon.ka.sub.org
From: eric@skatter.usask.ca Newsgroups: comp.sys.next.bugs Subject: Re: definitely a bug in code generation Date: 9 Feb 1994 21:08:58 GMT Organization: University of Saskatchewan Message-ID: <2jbjda$7uo@tribune.usask.ca> References: <1994Feb9.160012.13268@infodev.cam.ac.uk> From article <1994Feb9.160012.13268@infodev.cam.ac.uk>, by wfc@cl.cam.ac.uk (William Clocksin): > Just to improve on my own article: > > In article <1994Feb9.152209.10365@infodev.cam.ac.uk> wfc@cl.cam.ac.uk (William > Clocksin) writes: >> >> The following short program will demonstrate a bug in code generation under > NS 3.2 > [stuff deleted] > > A shorter program to demonstrate the bug is: > > main() > { > unsigned int f = 3; > int k = -3; > float y = 1701.0; > printf("%f\n", y + f * k); > } > > So multiplying an unsigned int by a negative int and then adding that to a > float will return trash. How many apps out there do this? > > It works OK if you don't add the float! > > Sorry again if this is a known bug or in a FAQ. > > ---------------------------------------------------- > Dr W.F. Clocksin, Assistant Director of Research, > Computer Laboratory, University of Cambridge > Pembroke Street, Cambridge CB2 3QG, U.K. > Internet: wfc@CL.cam.ac.uk (NeXTMail is acceptable) > Tel UK: (0223) 334628. Tel (Int'l): +44 223-334628. > Fax UK: (0223) 334678. Fax (Int'l): +44 223-334678. Before everyone madly starts checking their compiler for bugs.... There is no bug here. There's only a misunderstanding of arithmetic type conversions in C. From the ANSI C Standard section on arithmetic conversions: ....Otherwise, if either operand is unsigned int, the other is converted to unsigned int. The example expression is: unsigned int f = 3; int k = -3; float y = 1701.0; printf("%f\n", y + f * k); f * k is (unsigned int) * (int), which gets converted to (unsigned int) * (unsigned int) This is equivalent to: (unsigned int)3 * (unsigned int)-3 Which, on a 32 bit machine is: 3 * 2147483645 And not the 3 * -3 that Dr. Clocksin expects. -- Eric Norum eric@skatter.usask.ca Saskatchewan Accelerator Laboratory University of Saskatchewan Saskatoon, Canada. NeXTMail accepted.
From: spb@darkwing.uoregon.edu (Steven Berry) Newsgroups: comp.sys.next.misc,comp.sys.next.bugs Subject: Re: DOOM problem: STBAR not found Followup-To: comp.sys.next.misc,comp.sys.next.bugs Date: 10 Feb 1994 18:04:46 GMT Organization: University of Oregon Message-ID: <2jdsvu$lou@pith.uoregon.edu> References: <2jbs6i$ipd@ftp-p.mccaw.com> Jason Fosback (jason_fosback@psca.com) wrote: : Like many people, I grabbed the latest version of Doom off the archives : yesterday. I was very excited that it was FAT, but I was more excited : that I could get all of the levels by copying the DOS doom.wad file into : the NeXTSTEP app wraper. : Well, I replaced doom1.wad with the full doom.wad (calling it doom1.wad). : Things looked promising, and Doom reported that I had a registered : version. But then, I got the following error during the status bar : initialization: : W_GetNumForName: STBAR Not Found! : What's the scoop with this? The only thing I can think of is that the DOS : version of Doom is 1.1, and the NeXTSTEP version is 1.2. : Anyone know what the scoop is? : Thanks! : -jason : ____________________________________________________________ : Jason Fosback, Systems Analyst | No sir, I didn't like it : --- Paradigm Systems Corp --- | -R&S : Internet: jason.fosback@psca.com | Star Trek: : NeXT mail: jason.fosback@psca.com | The NeXT Generation... Fuzzo The problem is that you're a gomer. G! Spiffo
Newsgroups: comp.sys.next.bugs From: thf@zelator.de (Thomas Funke) Subject: Re: Sybase Administrator buggy under 3.x, Solution included. Message-ID: <1994Feb9.235137.1522@gamelan.uucp> Sender: thomas@gamelan.uucp (thomas) Organization: Disorganization References: <2j6s8l$763@runner.uucp> Date: Wed, 9 Feb 1994 23:51:37 GMT In article <2j6s8l$763@runner.uucp> richard%runner.uucp@usc.edu (Richard Ruth) writes: > Thomas Funke writes > ... > > Nevertheless, some hacking shows: There is a workaround available ! > > > > Get the 'binpatch' program and apply the following patch to the > > binary '/usr/sybase/Administrator' > > > ... > > Anyone else try this? and, where do you get 'binpatch' for NeXT > hardware? Many people have requested 'binpatch', so I post it once more: -------- cut here --------------------------- ------------ cut here -------------------- --- ------------------------------------------------------------------ Thomas Funke ** Unix-Consultant ** thf@zelator.de Brooks's Law: Adding manpower to a late software project makes it later ------------------------------------------------------------------ -- NewsGrazer, a NeXTstep(tm) news reader, posting -- M>UQR=&8P7&%N<VE[7&9O;G1T8FQ<9C!<9FUO9&5R;B!/:&QF<SM]"EQM87)G M;#$R,`I<;6%R9W(Q,C`*7'!A<F1<='@Q,34R7'1X,C,P-%QT>#,T-39<='@T M-C`X7'1X-3<V,%QT>#8Y,3)<='@X,#8T7'1X.3(Q-EQT>#$P,S8X7'1X,3$U M,C!<9C!<8C!<:3!<=6QN;VYE7&9S,C1<9F,P7&-F,"!);B!A<G1I8VQE(#PR M:C9S.&PD-S8S0')U;FYE<BYU=6-P/B!R:6-H87)D)7)U;FYE<BYU=6-P0'5S M8RYE9'4@*%)I8VAA<F0@4G5T:"D@=W)I=&5S.EP*/B!4:&]M87,@1G5N:V4@ M=W)I=&5S7`H^("XN+B!<"CX@/B!.979E<G1H96QE<W,L('-O;64@:&%C:VEN M9R!S:&]W<SH@5&AE<F4@:7,@82!W;W)K87)O=6YD(&%V86EL86)L92`A7`H^ M(#X@7`H^(#X@("!'970@=&AE("=B:6YP871C:"<@<')O9W)A;2!A;F0@87!P M;'D@=&AE(&9O;&QO=VEN9R!P871C:"!T;R!T:&5<"CX@/B`@(&)I;F%R>2`G M+W5S<B]S>6)A<V4O061M:6YI<W1R871O<B=<"CX@/B!<"CX@+BXN7`H^(%P* M/B!!;GEO;F4@96QS92!T<GD@=&AI<S\@(&%N9"P@=VAE<F4@9&\@>6]U(&=E M="`G8FEN<&%T8V@G(&9O<B!.95A4(%P*/B!H87)D=V%R93]<"EP*36%N>2!P M96]P;&4@:&%V92!R97%U97-T960@)V)I;G!A=&-H)RP@<V\@22!P;W-T(&ET M(&]N8V4@;6]R93I<"EP*+2TM+2TM+2T@8W5T(&AE<F4@+2TM+2TM+2TM+2TM M+2TM+2TM+2TM+2TM+2TM7`I<"@I[>UQ.1T1O8W5M96YT-3$P(&)I;G!A=&-H M+F,*-#@P-B!-)UE6,#A-*EA`,2XF4T%`5"M,.&`V+$)02#PN)R0B+2I')$)1 M.$0T,"8F5S!(8"0B8"0Z+"Q&251?*$`Q"DU11$(R)B9?+$HB)#(Q*%0Y+"8C M/$(G)2,I/#XL+"8K8%PF0E%)7%X_,"@L*B<T2E1*+29#,BDL2C<[2U0*34$P M2"4H)3)@(B-"4%@L)B0H)4I@(F`L)R-)3%PH+5LT2#`M1SLE0%,H)#(L.C`L M1RI13#$H+D)<8#0G,`I-*"E*3310-B9$,"\V,R%(5CDZ8%Y#,$HB(DE(5#Q4 M(RPR/"PV+"5463M<6R@B/2Q&.TE`53DQ*S,P-2Q&"DU;+4)2/20B.$`U43DK M-4Y7/"PN6"@[+E4J64Q41C0F1"(Q5$8\,4,E.3Y`*"]$3S<J1C\M/%9>/S(V M+D<*34,A3SE@7%=:(3Y0(DA>)RY@1"8K7T!./5-)4%([+6!"(4TG1B,Y1%0_ M3$Y0-4Q7.E,Y3STX3"HL-C4N1PI-6S<K2$(\/R@J-2TP/R\O(CHN2$0A2%Q7 M7S]+5D\F5S96,4U222,F/"96)E`H-4U".28R-E%'4DQ(-R,X"DTF-D%<33U- M7#A;4S$Q04\F+31!)B)"4F`H(5U=/24P5S4Q148D+4(Y/T`J/40U/34]350W M42<E15@D)CD*3350+4HY450H7D`U)C8H24Q,+R<F)S<O2"@M04XP/%!1,5%4 M/%%(-28P,B-$*#$X*#-1.T<Y)C]/23PT)@I-)R8R6%PI5#DY3")5)CDS4RL\ M129".E(M24HF,30R-3!`2$)'33HC0$5%5BI@,45&45TG24!%75A"8#XB"DTC M4B-(7CHT.3%1*T91*6!`1",F-T`J*ETD/C(P*SPX-4<F34`L-4LW,#-(1"@I M0CLH2SLU,$5-5"%,2D4*32)7.EY`/CY&.U4I55HU5$U((C1(/BI94C8Z031( M/5M115@K1E<W)D([8%9:,#0\.CA2,$A*.D$]7$9&/PI-/%@_345=0#LZ)C4X M05%3-UPW0CTF-S1(/F`]1U%413E15EDU(TY#)C55/B8U/B(A4$DH-UM*(S,N M-CI."DU+-R\\*3$\*"8L63<W6T)+0DL].SP](D9<.#@[*TQ0*E!!8#I01"8] M32)72B@N*#@V,U8Q53$T*2I-0R(*33<J2UI=3BI*(C]$*RPA45E#*$TF.B)` M+$0P*"@L/C(P,%!1,T`]-R0W0$$W)#8P.TI")#DI8$$R)BM!40I-.$4I1#5( M.#Q1.291)E$E14,E)28F5BL\(RLH+2@X.%Q,1F!/*BHC8"<G,#0A*E(H,UQ( M(5M10BM$.R4E"DTE4B-41"$X*3(M(S=((5`F3U(A1$PF4EPL*#P\.%M00$5: M+5-/(2)1-S)=*"P\,#0[1#]'-E0H3%9@)E(*32)2,B,L+R@K+#Q`8$Y@5$!2 M6F`C(R,W2#0H+2DK*BPM,$$E6$TD-B(B(TQ(+S@I8#A;(3$Q6"Q`7"PC5`I- M/59>6$`H."\B15-.,4XZ.SU>.%DG-C-,/3@[2$DI3BXN3CY+1%\F.B8K(3M7 M2B$_.T<P-D$A,2<U+24D"DTD-%DP+#0T(BA+,CHJ*E$_.2A$/#HL53-16#0[ M*$(S8$$P*"DY8%PW,D]!52U&0U1#1E17.$=63$(P)#8*325:.5E*0B(A(4Y8 M62PG/SY`,#T[3#`L(RY*0UM`*B1/0$TI+"Y85U`P-TXI(D%00"A11RH_8"4A M,STJ.@I-)%I+8"$B.31+1UY?7D5?-$A!*CU,.")/(V`^1#,F/#12)&`G7$I= M220X,3I@)3=03S\]1"PH)$`V)F`C"DTJ6CQ@25A@8")`-U(E,$%.(4%&23PT M02DX,T$C.5I854PA2RTI12E$2B0M1U=%-"<M5B-((S]`*D!10#P*33!4+"(Q M,C5)*TI1(DTN4E!%4S!(-#554BY@+C-#(2XZ55,A338D3THB2"4B5%@Y2E\P M22<V7$=")4=4)0I-1RP\*$%@64@T-"<R*"]`5"Y;4$$C.E$Q,"$G1#0L+$\P M)EA6*"1@(6`J-D9`0CLJ6#!-348L+#DV+B(B"DTQ2CE@8#=.8%)$/DA>8#1/ M5$`X+BLB,"DI/TQ'(5A942DT,B8Q*C!"-%DC*4-904%`)ETD(E!(7C%<+#`* M34(Z8"0Q-$DK0F`X.$8\1E18.#%%*"HT1"PV*B<T7#@P0T@E1"DQ72%2-#PB M0C<J,4L])T4K1"5@4C!8)0I-1$)/*BDH1"\N,3@C-S\[04!?64-905=3*BPT M/#U`."5#6%!$,C=2.5)&(E)+235)6CHX+B,P.2LD+"%2"DT^0BHX*3`E34(\ M3#!?,EE84%)//B4G,$DS,U!@)$`W-3DN)%5%4#LB+D98,R%@2F!8,4U<24HP M3R$D)BT*32<F*4,H+5XX44,N+"@T,B-#*$M+6R<B53PB6C-(0C$I,%)$441" M*DE(,F`M-C`O.2-&,S,N.E9!4"<T*`I-5%1)8#D[+B(E-U!,1$!'0"-0231, M+21-26`E2R\](SHB-4E00$E-0D-2,5-,*34]8$I&(D56.C%(4UPE"DU0,$E@ M14`J0RE(8$10.$LV(CQ14$!$4$50*#1%)RE`/$PY/3,U.BHT(5Q?+#A$4DDZ M339+/T(R3$`L)44*33)'33I465TV*T=-748B)#M(2#4S358I)$<]7S0I-5LI M.D(_6DLN)DXI+44])BQ12$DN4TU,.%U?62E`*`I-(20K)55;(T8C-R-853M9 M8"PM,#)#+B0D7&!1,3173#I5.D4V-%M@*2HN444D+TXQ.E(X0$M=0T%:521? M"DTT1"1:-R4B3C9"+4TE7EE:,SDK5%5(-2-8+%%:1V`R12Y7,T8C)EPH(S50 M6SA:6T<J-#Y,,C!.22I$7CH*32<M3CHD*"0[629@(R]<."<O0#%23CM#-2M! M+DHB0"H\8$,Z,%\F4DU@0#E*,5-00BTI,S,C*C`Q1$PB)`I-*$$A1DXR)CH[ M5U0I-%U,*SY',CQ&-TTC324B4%<O,C(M+&`C6TY=6T4K1RQ>*BI`(B5!*&!! M-EXX*#!`"DU4*SPJ-$)"(BTY-S970U)2,R)@,T4R)2PL,TDA1T-:54A9(3Q) M1UM-0E$D,BQ.7%U%,B\I0D->0F!`-3,*32E<.E1%,B4V43(I(TL\)&!144!) M6$8K+E0C3S8B+$0M*T8J-B8Z62,A4B-34F!5(RM#*5(P0RXF)B8A-`I-(4%! M15@S*4DV/"$[.$<^3BE8.R$[234F)$@T4CU@6"Y!,D`W)%XM-$)8.4A(4%(V M.T)@*5\N*B$Z,E4E"DT\(R,N/5Y,/#5>4EHT0%A"*"9=,#1=02%*4ULV1BDD M+4DE230\6"LP4BA,55M:,3Y"+"(K-S`N(T@K/E0*34<Z6BTG-RU'4RY*.4Y- M(EM+0#M?3R@I8#]64S`J05%//CLE5UHD+$8F0#8K/40N0U@^-%4]7%A!32Q: M50I-*T<U4#HV12PX5#$Q2E1:8$Q72$Q&0T0D4E!62E$F73HR2F!9/5Q:12$F M7B@I4$$L2"8P1D%<0"0U.UM6"DTF)C!`-BE4+BY3,B,K3"@S0R8A-41*0U9, M+4(C6S5:04A<24]`4F`D-DE-(4XS7B(R15Y)-$`R34E'-#@*34](+40V7BXZ M5$%(-$LC,BL^-$`G*E=/,TX^,$U=2DH]1E$Z5#A@/SD\7B4Z520]2RY#4"M2 M12Y),S8]*`I-44]@/4,[-518+40](BA!,"$B)318,$,X2E,\4#@U5"(E7F!@ M(U(Q24XP44!?028Y6%`G-R<\/DA%-#<D"DU;/$HM2"0O0#0U23E@6T=$+$DQ M02,R)"XM-S4B12<L,#,J)"@Q-4,B)5E9+D$X7R$F4#!01B,])C4G6T\*33\E M7T=:*20B/#HR14,W(3Y,43A#7"\T1U%063XW/UQ*-RU`22TX+"XC.2E4(E); M6$M-6$96/3<G2"<T5PI-5D5/,30I(3A%0#E53%PR2$HS*E4S,3U3,%TM43U; M2#PC4#E37RQ043A/+S!$231$,SM>2$)@*CH[2U<J"DU52E\E+4-+,C`V5U=) M.5Y%6B<P*%0]+D,O3S1"-2]73$A`)"U00#(I63526R\]540[0E0H/BY`."4I M0D$*32,L4SHL64-<1E,T/S%=+C0B+UE&0#E53$%.,S<W/3Y@0C\M)R=@44$B M-%Y=(D0L.#!!)D9,+%\J-S,M*PI-6$DL-T9.64,F+U<H(4)65UPP75M602E< M,4LF+S=@2SHS2"PV8%HE8#-*02(^.T$P4U,P-#0P7B$R)#-""DTY4BDA-%4H M2TQ!0UU=0SDW0#$D4RQ$+%9404!9*$%,6U1"+EM@,C$B,$(Q3"Q!03)$,C$K M5$(Q0R1!2SH*33!"-4$X(C8L1$(T0EPB24`H0CDO6$9".3@T.5X\1T`W4B)$ M,UE1.R0P0DE72$8]6CA"4DQ@,D!&,B)!,0I-3#%.)SQ`/2PD.#$B+B@Q)#8H M,28^*#$H1B@Q*DXH,2Q6*#$N7B@Q,"8H-3(N*#4T-B@U-CXH-3=>(T(Q"DU@ M4"HS4%0J4DA31E1<4T=:1$--/T8C2CU<4T9.,%@Z.&!8.DI<4T971%-.14Q3 M3")8,T]@*%-3)#A34R@*34A34RM!,20H(51?-C!5.RA04U<]3%-3(R%76U$D M."-%/%-;22A8+T]!4UX]8#-=/RQ#7BY>(CXD.F`Z+@I-2B-?."4D*#E),"D\ M,41=/RE%6"\U)6`R+20I15U81B1-1"HY7B,R.R(G8$A%)DA,620H4#$E)B91 M(58Y"DTL,#\^32=.)U`T,4HA5#$B+2)1)3TD,2E-)#(]+#A-0R4T-3,U)#4W M42512BE$+C8A)#M-8#`]0U$A548*345@.4A9(TE+32(O3U!'-2\Q)2,Y528D M*%4V25Y))SHA32A&044D0BM5)$!.7"5>,R5@.C5!(4(W+24K+`I-42A:/%DD M15`P)%%=.2A;)S%5*SI.)$M'+213*R5$3S)52$0F-29%+E5$3ULJ)%%2."A? M)CDH64\A0#]9"DTF*%T_6"103UDY(U!:)B(\+54A5$$C(SY*8"-'(31822U$ M6"=6.%PI1C1:5DXD6E991EXR74%"52TI(D,*32%`*U<Y*#]:.29$-DHT4#@J M.$LZ6258*2$Q14=%1CDZ)24D6S!%5E1(-3A$+40H754P-TI2*#Q:,45/4`I- M-B4H,T4U*%%,)2E'634H4$4Q-30A.315.34T6CDV*4=@)35@*34Q)%4W/2DM M,4PG4#4Q35%)-3)1-3LA"DU!130Z455%/%DU-4!1-CX]+#4[.4A5+4=%-3@V M,3DV1R4V4D!16$@[-24D/#4U/4-!13X^55%%73$Q*#T*34(U2DY%-4M20$99 M6T(R*BL],CHM13`C0%XF2C-9,#HU635:.$DU1R=4-48_039.0STV*D51-3E7 M23E")@I-(3=.3DDV6C\E52E-*59(2$DE3"I954TP8#53(D$W,#U.)T\G1352 M-3XD)3M,,4DH*C%5.2Q%.5)!0"I$"DTA5U0X72A'0E$U5SXQ-CXE3B9/)CXX M2T,Y-38G6C8S-U$E6"=),"%)751<4RE!-RI%2596359@/T(W42(*32Q&(RM- M5%,M)5E;/T$G(S)!5SM"7B9@,2U&)#4]1B0Y348D*DPI)#\]05%@-C4F.DU8 M124^.3!`+"8L30I-3451+EE)7#1-*4A7744E*4(Q)TY91BTN*48P)BA&-2=% M1C%`4$8P*R)&-"U:0CPT13947$56-2I,5C=3"DU1*2I-)C8Y0TU$448U2$Q< M/CA53"U&*5Y&.5M'+58]6U8F15DY25U<7C8^1C5'4RXZ-D`D)5A;7R4F3E,* M35U&0$@J-E5<5C9,/TDV1"121D8M7$HK+SU535Y%03E2-"9*(CU)4R$M24=+ M2CHL2DXF7%TF0BI%)B<C,@I-3C138$%%+"DV1E,G14914%HT2RPU0#\N6493 M,C%&5#%&04\Z34<O.%5'+SA85E1<(499.BA66T1%23@_"DT\-EI<+3E`2RI$ M05%>1EM154<_6T(F7"1)5E]<64(I0RU615`Z-SI303<X5SI'(4%625<T*#5" M3E4W)"X*32%7)3\I,311*B(\-3U7)BE$,5LQ4$<E25I')2E@5RE&*"<H1#E7 M*$A%4"(A5D<H3R57+$XL)T1575I(5PI-15DJ.4U7+SM)1$Q?35%)(5U:.DHU M)S$G/5HS3U`W,D%*1TDP*5<U-#E7-R8Y6CT[751-/2U:03]-4#E!"DT]45(G M(DI%74%*-BLN1"XM3DE93$E7,S%=5S@H/%8]4RU73TA"-S]-24<].E!7/RI: M1SQ+*5`J(4U5,B,*-"%23B4I44I$1#9.*DU1-BPE6DHN75<J,557*"@*8`I] M"JQ]7'!A<F1<='@Q,34R7'1X,C,P-%QT>#,T-39<='@T-C`X7'1X-3<V,%QT M>#8Y,3)<='@X,#8T7'1X.3(Q-EQT>#$P,S8X7'1X,3$U,C!<9C!<8C!<:3!< M=6QN;VYE7&9S,C1<9F,P7&-F,"!<"EP*+2TM+2TM+2TM+2TM(&-U="!H97)E M("TM+2TM+2TM+2TM+2TM+2TM+2TM7`I<"EP*+2TM7`HM+2TM+2TM+2TM+2TM M+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM M+2TM+2TM+2U<"E1H;VUA<R!&=6YK92`@*BH@56YI>"U#;VYS=6QT86YT("HJ M('1H9D!Z96QA=&]R+F1E7`I<"D)R;V]K<R=S($QA=SI<"B`@("`@("`@061D M:6YG(&UA;G!O=V5R('1O(&$@;&%T92!S;V9T=V%R92!P<F]J96-T(&UA:V5S M(&ET(&QA=&5R7`HM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM E+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2U<"@I]"BTM ` -- ------------------------------------------------------------------ Thomas Funke ** Unix-Consultant ** thf@zelator.de Brooks's Law: Adding manpower to a late software project makes it later ------------------------------------------------------------------
Newsgroups: comp.sys.next.bugs From: alastair@farli.otago.ac.nz (Alastair Thomson) Subject: Re: definitely a bug in code generation Message-ID: <CL0yD6.2z@news.otago.ac.nz> Sender: usenet@news.otago.ac.nz (News stuff) Organization: University of Otago Date: Thu, 10 Feb 1994 20:03:06 GMT In article <1994Feb9.152209.10365@infodev.cam.ac.uk> wfc@cl.cam.ac.uk (William Clocksin) writes: > > The following short program will demonstrate a bug in code generation under NS 3.2 > [stuff deleted] > A shorter program to demonstrate the bug is: > > main() > { > unsigned int f = 3; > int k = -3; > float y = 1701.0; > printf("%f\n", y + f * k); > } Hmmm, interesting, I have had similar problems with the maths in NS. It usually fix it if I disable optimisation :-( In this case, these are the results I get from the program farli:[~] 52 % cc -O qqq.c -o qqq farli:[~] 53 % qqq 4294968832.000000 farli:[~] 54 % cc qqq.c -o qqq farli:[~] 55 % qqq 4294968988.000000 -- Alastair Thomson, | Phone +64-3-479-8347 Computer Science Department, | Fax +64-3-479-8529 University of Otago, | Dunedin, | alastair@farli.otago.ac.nz New Zealand | NeXTmail Welcome
From: elitman@proxima.com (Eric A. Litman) Newsgroups: comp.sys.next.bugs Subject: posting test - please ignore Date: 10 Feb 1994 14:40:23 -0600 Organization: Proxima, Inc. Sender: daemon@cs.utexas.edu Message-ID: <CL0uKr.4KD@proxima.com> this is only a test. -- Eric Litman Proxima, Inc. vox: (703) 506.1661 Systems Engineer McLean, VA elitman+@proxima.com echo '[q]sa[ln0=aln256%Pln256/snlbx]sb3135071790101768542287578439snlbxq'|dc
Newsgroups: comp.sys.next.bugs From: antoine@arrakis.osd.ulaval.ca (Antoine Gautier) Subject: Re: Improv, TeXTView and print:summary Message-ID: <CL1LI6.rz@athena.ulaval.ca> Sender: news@athena.ulaval.ca Organization: CTI, Universite Laval References: <1994Feb5.100029.20544@proximus.north.de> Date: Fri, 11 Feb 1994 04:22:53 GMT A while back, I wrote: >Here is the problem: I have a TeX document (foo.tex) with a few >psfigs (.ps format). I need to email it to another NeXT, so I >print-to-file it from TeXView (- foo.ps). So far so good. Just >to check, I open the PS foo.ps with the previewer, it crashes at >the first psfig generated from Improv (the others, which come >from Diagram, are OK). That is, the TeX document shows up on preview >until the last line before the improv-figure. Any clue? > >I tried recompiling TeXView and dvips, but that did nothing. >Hummmm!!! > Well, apparently I was not the only one having problems wth the - 3.2 upgrade. Thanks to Tom Marchioro , I got in touch with Tomas G. Rokicki, the author of TeXview. The recommendation is to upgrade to the latest versions of TeXview and dvips. these are in /pub From labrea.stanford.edu, ftp dvips5528.tar.Z and texview.tar.Z, compile, install, and go. Thks Tom, Tom, and good luck to all who had similar problems. Antoine --- +-------------------------------------------------------+ | Antoine Gautier | | Departement OSD, FSA | | Universite Laval, Quebec, PQ | | Internet: Antoine@arrakis.osd.ulaval.ca | +-------------------------------------------------------+
From: M.Crawford@dcs.shef.ac.uk (Malcolm Crawford) Newsgroups: comp.sys.next.bugs Subject: Compress warning! Date: 11 Feb 1994 11:04:12 -0600 Organization: UTexas Mail-to-News Gateway Sender: daemon@cs.utexas.edu Message-ID: <9402111643.AA22276@dcs.shef.ac.uk> > Do NOT use the "compress" workspace feature if you care > about your data. I've used it dozens of times; but when > I recently tried to compress a large directory, the > directory simply disappeared! And with no error messages > (even in the console). I've lost several hours worth of > work, plus the cost of downloading the original data in > the first place. Thanks a lot, NeXT. > I use Workspace's compress frequently, and the only problem I ever have is that the original hangs around as a .nfs file... I wonder if the following may be relavant...? Have fun, mmalcolm. From: Malcolm Crawford <M.Crawford> Date: Sat, 12 Jun 93 18:25:03 GMT To: next-managers@edu.stolaf, UK-NeXT-Users@uk.ac.york.ohm Subject: compress <-> quota interaction "problem" Here's a warning from our system administrator which may be of interest to those of you enforcing quotas. The basic message -- use compress with care if you are over your hard quota! I'd be interested to hear of any fixes for this. Have fun, Malcolm. Date: Fri, 11 Jun 93 14:42:09 BST From: Dave Mitchell <D.Mitchell> To: all Subject: dangers of compression Here is an IMPORTANT warning to anyone who uses the UNIX compress utility (either directly, or via the NeXT Workspace manager etc): The compress command has a bug in it, that causes it to truncate files to zero length if you are over quota, and the disk is non-local. i.e., given a file f, "compress f" does the following: it creates a new file f.Z. It then tries to write data to it. This fails because you are over quota, so f.Z remains zero-length. compress fails to notice this error (due to arcane reasons involving delayed writes over NFS), and happily goes ahead and deletes the original file f. This bug seems to be present on all versions of SunOS and NEXTSTEP within the department. So, be VERY careful when compressing files while over quota. Dave.
From: nlau@sct.edu (Neric Lau) Newsgroups: comp.sys.next.bugs Subject: HELP NEEDED ON RESEARCH Message-ID: <2jh21t$pa2@st6000.sct.edu> Date: 11 Feb 94 22:49:33 GMT Organization: Southern College of Technology, Atlanta I am writing a research paper on why Next computer fail on the workstation market. anyone outthere can send me the information about the good and bad thingon NeXT. Please send the information to nlau@st6000.sct.edu Neric Lau
From: robert@steffi.demon.co.uk (Robert Nicholson) Newsgroups: comp.sys.next.bugs,comp.sys.next.advocacy Subject: Re: HELP NEEDED ON RESEARCH Date: 12 Feb 1994 01:20:30 -0000 Organization: me organized? That's a joke! Message-ID: <2jhasu$l6@steffi.demon.co.uk> References: <2jh21t$pa2@st6000.sct.edu> nlau@sct.edu (Neric Lau) wrote in comp.sys.next.bugs >I am writing a research paper on why Next computer fail on the workstation >market. anyone outthere can send me the information about the good and bad thingon NeXT. >Please send the information to nlau@st6000.sct.edu > >Neric Lau Perhaps you should ask somebody to tell you what cross-posting is before you research anything. Why you used two different inappropriate subject headers is beyond me. As for your original question. I put it down to two things. Lack of market focus and inability (lack of adequate resources) to perform marketing. -- "You know what's wrong with you?" (Audrey Hepburn, Cary Grant) "No, what?" "Nothing" (Charade, 1963) (ASCII for text only messages)
From: t146678@lehtori.cc.tut.fi (Tuominen Juha) Newsgroups: comp.sys.next.bugs Subject: Print panel blows itself away when using Options-panel Date: 12 Feb 1994 13:39:49 +0200 Organization: Tampere University of Technology, Computing Centre Distribution: world Message-ID: <2jif65$2ei@cc.tut.fi> I attached a HP PaintJet XL300 color printer to a PC running NSi3.2 at Friday. It printed some quite nice pics but rastered images looked somewhat awful. I tried to fix the problem using Print panel's Options menu and there it was: Print quality (which was set to normal (meaning draft?)). I choosed Best and clicked OK. All right, the printing panel crashed. At this point I was using Preview. I decided to use ImageViewer and the same thing happened. The applications didn't crash (why would they), but Print panel blows itself away as soon as the OK button of the Options menu is clicked. I didn't try any other driver so I don't know if this is an OS bug or a driver bug. And I'm not going to find it out during the weekend - motion sensors are very hard to fool (the PC is located my workinplace). I would appreciate if someone out there would try to choose the printer type to PaintJet XL300 using PrinterManager.app and try if it crashes in the same manner. In that way could be found out if the problem is in driver/OS or in our PC (or in harddisk). -Juha -- Unauthorised copying, public performance, broadcasting, hiring or rental of this mail prohibited. All the names have been changed to protect the innocent.
From: dgc@ccrwest.org (David G. Cantor) Newsgroups: comp.sys.next.bugs Subject: Problem with initex Date: 12 Feb 1994 19:18:32 -0800 Organization: IDA Center for Communications Research Message-ID: <2jk668$6fq@ccrwest.ccrwest.org> Summary: Unable to use .fmt files Keywords: tex initex fmt I recently upgraded to 3.2. Now I am unable to use initex to make format files. Whenever I attempt to use one that I have just created I get the diagnostic: (Fatal format file error; I'm stymied) Format files made one year ago work without any problem. Does anyone know what might cause this? dgc David G. Cantor Center for Communications Research 4320 Westerra Court San Diego, CA 92121 dgc@ccrwest.org
From: dgc@ccrwest.org (David G. Cantor) Newsgroups: comp.sys.next.bugs Subject: tex Date: 12 Feb 1994 20:00:19 -0800 Organization: IDA Center for Communications Research Message-ID: <2jk8kj$76m@ccrwest.ccrwest.org> Summary: Older version of tex works with newly created .fmt files Keywords: tex initex .fmt In my previous posting I noted that .fmt tex files created with initex were being rejected by tex. I restored the tex that was on the version 3.0 cdrom. This seems to have solved the problem. dgc David G. Cantor Department of Mathematics University of California Los Angeles, CA 90024
From: vamp@gothic.acs.csulb.edu (VampLestat) Newsgroups: comp.sys.next.bugs,comp.sys.next.misc Subject: Q: MouseX on NS 3.2 ? Date: 13 Feb 1994 06:27:39 GMT Organization: Me, organized? You gotta be kidding. Message-ID: <2jkh8r$29v@garuda.csulb.edu> Has anyone successfully run MouseX under NS 3.2? We had it running fine on a 2.1 machine, and after we installed 3.2 (from scratch), it seemed to stop working. It could just be something that wasn't reinstalled during the upgrade, but I was wondering if any others had a problem with MouseX under 3.2, or if if MouseX was running fine under 3.2? -- _O_ Ryan L. Watkins : : : : : :: ::: :::: ::::: | vamp@csulb.edu : : : : : :: ::: :::: ::::: | aka VampLestat finger vamp@beach.csulb.edu for pgp public key
Newsgroups: comp.sys.next.bugs,comp.sys.next.sysadmin From: mgb@moksha.uucp (Michael Branton) Subject: freezing windows Message-ID: <CL52GK.6zC@moksha.uucp> Sender: mgb@moksha.uucp (Michael Branton) Organization: Totally Disorganized Date: Sun, 13 Feb 1994 01:21:56 GMT Here's a strange one: On a NeXT turbo slab, we experience all windows freezing their positions. You can launch an app, but once it opens a window, you can't move it. Two entries in the message log seem like they might have some bearing on the problem. One occurs during boot, the other within 30 seconds after boot complete : . . . Feb 12 17:22:59 ferrari mach: Disk Label: swapdisk Feb 12 17:22:59 ferrari mach: Disk Capacity 247MB, Device Block 1024 bytes Feb 12 17:22:59 ferrari mach: sc: MESSAGE REJECT RECEIVED Feb 12 17:22:59 ferrari last message repeated 20 times Feb 12 17:22:59 ferrari mach: Generic SCSI Device as sg0 at sc0 target 7 lun 7 . . . Feb 12 17:23:53 ferrari loginwindow[313]: loginwindow: could not find WindowServ-er port! At the same time, we're getting printer errors: Feb 12 17:45:26 ferrari npd[187]: PrintSpooler couldn't stat /usr/spool/NeXT/MIG: No such file or directory Feb 12 17:48:01 ferrari Server:MIG[1718]: Driver class not set in printer entry Feb 12 17:50:46 ferrari loginwindow[526]: loginwindow: Workspace exited ts 0 cd 0 rc 0 sv 0 ss 0. /usr/spool/NeXT/MIG is there, as one can cd to the directory and look around. Anybody seen anything like this, particularly the frozen window positions ? --- -Michael mgb@thoth.stetson.edu moksha!mgb@bliss.stetson.edu NeXT mail welcome ! -- -Michael mgb@thoth.stetson.edu -- -Michael
Newsgroups: comp.sys.next.bugs From: bugni@cs.washington.edu (Paul F Bugni) Subject: Compile problems w/ term112 on NSFIP Message-ID: <1994Feb13.174811.26787@beaver.cs.washington.edu> Sender: news@beaver.cs.washington.edu (USENET News System) Organization: Computer Science & Engineering, U. of Washington, Seattle References: <1994Feb7.203431.1391@polar> Date: Sun, 13 Feb 94 17:48:11 GMT I'm trying to install term on my NSFIP and am running into the following compiling problems. Has anyone met these with success or do you have suggestions as to what I can do to fix the problem. Compile errors: cc -O -DLOGIN_SHELL -funsigned-char -c lib.c lib.c:173: illegal statement, missing `;' after `)' lib.c:174: illegal statement, missing `;' after `)' *** Exit 1 Stop. *** Exit 1 Stop. I looked into this and it is the following function where the complier screams specifically where i've added the comment 'the following is line #173': #define BUFF_INC_SIZE 1024 int add_to_buffer(struct Buffer *B, un_char c) { if (B->size >= B->alloced - 2) { DEBUG_LL (stderr, "growing buffer: add_to_buffer %d, %d\n", B->size, B->alloced); if (!B->data) B->data = malloc( B->alloced + BUFF_INC_SIZE ); else B->data = realloc(B->data, B->alloced + BUFF_INC_SIZE); B->alloced += BUFF_INC_SIZE; if (B->end > B->start) { /* curses. need to move data. */ int i; char *p, *q; /* the followin is line #173 */ p = (caddr_t)B->data + B->alloced - 1; q = (caddr_t)B->data + B->alloced - BUFF_INC_SIZE - 1; for (i = B->alloced - BUFF_INC_SIZE - B->end; i > 0; --i) *p-- = *q--; B->end += BUFF_INC_SIZE; } } B->data[B->start++] = (c); B->start == B->alloced ? B->start = 0 : 0; B->size++; return 0; } I've followed the Makefile instructions to set CC... # if use a NeXT machine, then replace CC with this line... CC=cc -DNeXT # CPP=$(CC) -E # if useing a NeXT then use.. CPP=cc -E -DNeXT DEBUGFLAGS= $(DEBUG) -O I man(ed) curses and it showed: SYNOPSIS cc [ flags ] files -lcurses -ltermcap [ libraries ] So I don't think the mentioned -ltermcap is the problem. Any suggestions would be appreciated. Thanks, Paul.
Newsgroups: comp.sys.next.bugs From: volker@abulafia.in-berlin.de (Volker Safran) Subject: Re: Q: MouseX on NS 3.2 ? Message-ID: <1994Feb13.144327.781@abulafia.in-berlin.de> Sender: volker@abulafia.in-berlin.de Organization: Volker Safran, Interprint, Berlin, Germany References: <2jkh8r$29v@garuda.csulb.edu> Date: Sun, 13 Feb 1994 14:43:27 GMT In article <2jkh8r$29v@garuda.csulb.edu> vamp@gothic.acs.csulb.edu (VampLestat) writes: > Has anyone successfully run MouseX under NS 3.2? We had it running > fine on a 2.1 machine, and after we installed 3.2 (from scratch), > it seemed to stop working. It could just be something that wasn't > reinstalled during the upgrade, but I was wondering if any others > had a problem with MouseX under 3.2, or if if MouseX was running > fine under 3.2? Hmmm, did you check the permissions of /dev/vid0? crw-rw-rw- 1 root wheel 13, 0 May 5 1993 /dev/vid0 CIAO Volker -- ************************************************************ * Volker Safran, FB13, TU Berlin, PHONE: +49 30 4542303 * * EMail: FAX: +49 30 4537157 * * volker@abulafia.in-berlin.de (NeXTMail very welcome)* * safran@fb3-s7.math.TU-Berlin.DE (no NeXTMail, sorry) * ************************************************************
From: mow@marsu.tynet.sub.org (Markus Wenzel) Newsgroups: comp.sys.next.misc,comp.sys.next.bugs Subject: Re: DOOM problem: STBAR not found Date: 12 Feb 1994 18:33:03 +0100 Organization: Palumbian Research Labs Message-ID: <2jj3sf$12l@marsu.tynet.sub.org> References: <2jbs6i$ipd@ftp-p.mccaw.com> <2jdsvu$lou@pith.uoregon.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit spb@darkwing.uoregon.edu (Steven Berry) writes: >Jason Fosback (jason_fosback@psca.com) wrote: >: W_GetNumForName: STBAR Not Found! >: What's the scoop with this? The only thing I can think of is that the DOS >: version of Doom is 1.1, and the NeXTSTEP version is 1.2. >: Anyone know what the scoop is? >The problem is that you're a gomer. >G! >Spiffo Very informative answer. Could you please explain what a "gomer" is? I found it neither in Webster nor in my dictionary. -- Markus Wenzel System administration, Consulting, Networking mow@marsu.tynet.sub.org on... NeXTSTEP / Unix / Novell / Windows NT IRC: Marsu Veni, vidi, NeXTSTEPi.
Newsgroups: comp.sys.next.bugs,comp.sys.next.misc From: mperry@netcom.com (Michael Perry) Subject: Re: Q: MouseX on NS 3.2 ? Message-ID: <mperryCL7Go5.7J2@netcom.com> Organization: NETCOM On-line Communication Services (408 241-9760 guest) References: <2jkh8r$29v@garuda.csulb.edu> Date: Mon, 14 Feb 1994 08:24:04 GMT vamp@gothic.acs.csulb.edu (VampLestat) writes: >Has anyone successfully run MouseX under NS 3.2? Yes, I use it on a Color Station. It works fine except for one application which fails when it tries to allocate 128 entries in the color table. Michael Perry mperry@netcom.com
From: dave@sashimi.wwa.com (David Vrona) Newsgroups: comp.sys.next.bugs Subject: need help with uucp connection Date: 14 Feb 1994 10:56:56 -0600 Organization: Computing Engineers Inc. Home of WorldWide Access (SM) Message-ID: <2joago$49k@sashimi.wwa.com> Hi all, I have a problem with a Next machine that polls my site for news and email. It seems that the uucp session dies after a successful login. I'm not a Next user so I don't know how to solve this problem. I was hoping to get some help from the net. Any and all tips would be greatly appreciated. I hope I am posting to the right group. BEGIN L.sys ENTRY wwa Any DIR 38400 cufb\ "" ATDT95555555\ ogin: uXXXXXX assword: YYYYYY BEGIN DEBUGGIN SESSION OUTPUT opus:1# /usr/lib/uucp/uucico -r1 -x6 -swwa root wwa (2/1-11:39-1476) DEBUG (Local Enabled) finds (wwa) called getto: call no. cufb for sys wwa Using DIR to call Opening /dev/cufb login called wanted """" got: that send "ATDT95555555" wanted "ogin:" ATDT95555555 CONNECT 38400 This is WorldWide Access (SM) Lake County, IL Node enter "bbs" to enter the bulletin board system enter "guest" to get more information or proceed if you have shell access \007Connected - 14400/ARQ/V42b [8/N/1] on termi1E 2/1/1994 11:41:38 current users = 9 login:got: that send "uXXXXXX" wanted "assword:" uXXXXXX Password:got: that send "YYYYYY" root wwa (2/1-11:39-1476) SUCCEEDED (call to wwa ) imsg looking for SYNC< NO CARRIER root wwa (2/1-11:41-1476) TIMEOUT (wwa) opus:2# END DEBUGGING SESSION OUTPUT -- David Vrona N9QNZ +1 708 367 1870 (voice) Internet: dave@wwa.com Computing Engineers Inc. +1 708 367 1871 (data) +1 708 367 1872 (fax) Home of WorldWide Access (SM), Internet services for the Chicagoland area. Send e-mail to info@wwa.com for information, or call +1 708 367 1871 (data).
From: f92-hst@byse.nada.kth.se (Henrik Stehl) Newsgroups: comp.sys.next.bugs Subject: Keyboard problem installing 3.2 Date: 14 Feb 1994 17:14:05 GMT Organization: Royal Institute of Technology, Stockholm, Sweden Message-ID: <2jobgt$oup@news.kth.se> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit I get the following error message trying to install NS/FIP 3.2: PS2Keyboard/DoResend: Unexpected RESEND from keyboard It occurs at the "This is a complete installation. Press 1 to cont" prompt. At the same time the keyboard hangs and nothing I type will appear on the screen. I have a Mylex 486/66 with an Adaptec 1542B SCSI adapter and a Compaq QVision 1024/E graphics card in my system, and there is a Conner HD and a NEC CD-ROM hooked up to the Adaptec. There is nothing special about the keyboard and NS/FIP 3.1 worked fine on my system. I have not tried installing 3.1 and upgrading. If you have any clue as to what's the cause of this, please mail me. Henrik Stahl Royal Institute of Technology Stockholm, Sweden f92-hst@nada.kth.se
From: sanguish@digifix.com Newsgroups: comp.sys.next.announce,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.bugs,comp.sys.next.programmer,comp.sys.next.marketplace Subject: Quick Guide to NEXTSTEP information on the Internet Date: 14 Feb 1994 14:45:33 -0500 Organization: Next Announcements Message-ID: <2jokct$gqf@digifix.digifix.com> This post is made weekly, to help 'point' users to more NEXTSTEP information Topics include: comp.sys.next newsgroups related newsgroups comp.sys.next newsgroups mailing list ftp sites NeXTanswers comp.sys.next.* newsgroups -------------------------- Comp.Sys.Next.Advocacy This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. Comp.Sys.Next.Announce Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) The NEXTSTEP FAQs are posted here monthly as well. This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Comp.Sys.Next.Bugs A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT- specific groups as well. Comp.Sys.Next.Hardware Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. Comp.Sys.Next.Marketplace NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. Comp.Sys.Next.Misc For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! Comp.Sys.Next.Programmer Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. The NEXTSTEP programmer FAQs are posted here. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. Comp.Sys.Next.Software This is a place to talk about [third party] software products that run on NEXTSTEP systems. Comp.Sys.Next.Sysadmin Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. related Newsgroups ------------------ Comp.Soft-Sys.Nextstep Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. Comp.Lang.Objective-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. Comp.Object Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original Comp.Sys.Next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News ------------------------------------------- Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep-d next-advocacy-d next-announce-d next-bugs-d next-hardware-d next-marketplace-d next-misc-d next-programmer-d next-software-d next-sysadmin-d (For a full description, send mail saying LISTS to <digestif@antigone.com>). The subscription syntax is essentially the same as LISTSERV's. To subscribe, send a message to <digestif@antigone.com> saying: SUB Listname YourName Example: SUB next-hardware-d John Doe The ftp sites ------------- cs.orst.edu: The main site for North American submissions nova.cc.purdue.edu: Lots of older stuff, but very short on disk space ftp.informatik.uni-muenchen.de: In Germany. terra.stack.urc.tue.nl (Dutch NEXTSTEP User Group) and cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp.next.com: See the below ftp.next.com and NextAnswers@next.com ------------------------------------- From the document 1000_Help from ftp.next.com Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, or you can transfer them by anonymous ftp. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to NeXTanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO <your-address> If you have any problem with the system or suggestions for improvement, please send mail to NeXTanswers-request@NeXT.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to NeXTanswers-request@NeXT.com. Written by: Eric P. Scott eps@toaster.SFSU.EDU and Scott Anguish sanguish@digifix.com Additions from: Greg Anderson (Greg_Anderson@afs.com) and Michael Pizolato (Michael_Pizolato@afs.com)
Newsgroups: comp.sys.next.bugs From: preuss@netcom.com (Peter Preuss) Subject: Re: Compile problems w/ term112 on NSFIP Message-ID: <preussCL8IzB.14A@netcom.com> Organization: NETCOM On-line Communication Services (408 241-9760 guest) References: <1994Feb7.203431.1391@polar> <1994Feb13.174811.26787@beaver.cs.washington.edu> Date: Mon, 14 Feb 1994 22:11:35 GMT The problem is NeXT's funny way of rearranging include files from release to release (their software developers must really be bored, PA-RISC, PowerPC, and SPARC ports all done, held back just by marketing :) Because the proper include file is not found, caddr_t is not recognized in p = (caddr_t)B->data + B->alloced - 1; q = (caddr_t)B->data + B->alloced - BUFF_INC_SIZE - 1; Changing "caddr_t" to "char *" permits compile. Please submit better fixes to the term maintainer. Peter. Paul F Bugni (bugni@cs.washington.edu) wrote: : I'm trying to install term on my NSFIP and am running into the following : compiling problems. Has anyone met these with success or do you have : suggestions as to what I can do to fix the problem. : Compile errors: : cc -O -DLOGIN_SHELL -funsigned-char -c lib.c : lib.c:173: illegal statement, missing `;' after `)' : lib.c:174: illegal statement, missing `;' after `)' : *** Exit 1 : Stop. : *** Exit 1 : Stop. : I looked into this and it is the following function where the complier screams : specifically where i've added the comment 'the following is line #173': : #define BUFF_INC_SIZE 1024 : int add_to_buffer(struct Buffer *B, un_char c) : { : if (B->size >= B->alloced - 2) { : : DEBUG_LL (stderr, "growing buffer: add_to_buffer %d, %d\n", : B->size, B->alloced); : if (!B->data) : B->data = malloc( B->alloced + BUFF_INC_SIZE ); : else : B->data = realloc(B->data, B->alloced + BUFF_INC_SIZE); : B->alloced += BUFF_INC_SIZE; : if (B->end > B->start) { /* curses. need to move data. */ : int i; char *p, *q; /* the followin is line #173 */ : p = (caddr_t)B->data + B->alloced - 1; : q = (caddr_t)B->data + B->alloced - BUFF_INC_SIZE - 1; : for (i = B->alloced - BUFF_INC_SIZE - B->end; i > 0; --i) : *p-- = *q--; : B->end += BUFF_INC_SIZE; : } : } : B->data[B->start++] = (c); : B->start == B->alloced ? B->start = 0 : 0; : B->size++; : return 0; : } : I've followed the Makefile instructions to set CC... : # if use a NeXT machine, then replace CC with this line... : CC=cc -DNeXT : # CPP=$(CC) -E : # if useing a NeXT then use.. : CPP=cc -E -DNeXT : DEBUGFLAGS= $(DEBUG) -O : I man(ed) curses and it showed: : SYNOPSIS : cc [ flags ] files -lcurses -ltermcap [ libraries ] : So I don't think the mentioned -ltermcap is the problem. : Any suggestions would be appreciated. : Thanks, : Paul. -- preuss@futon.sfsu.edu San Francisco State University
Newsgroups: comp.sys.next.bugs From: bugni@cs.washington.edu (Paul F Bugni) Subject: Re: Compile problems w/ term112 on NSFIP Message-ID: <1994Feb14.224146.20612@beaver.cs.washington.edu> Sender: news@beaver.cs.washington.edu (USENET News System) Organization: Computer Science & Engineering, U. of Washington, Seattle References: <1994Feb7.203431.1391@polar> <1994Feb13.174811.26787@beaver.cs.washington.edu> Date: Mon, 14 Feb 94 22:41:46 GMT >I'm trying to install term on my NSFIP and am running into the following >compiling problems. Has anyone met these with success or do you have >suggestions as to what I can do to fix the problem. > >Compile errors: > >cc -O -DLOGIN_SHELL -funsigned-char -c lib.c >lib.c:173: illegal statement, missing `;' after `)' >lib.c:174: illegal statement, missing `;' after `)' >*** Exit 1 >Stop. >*** Exit 1 >Stop. > It turned out that all that needs to be changed is replace the following two occurences of caddr_t with char* in lib.c > p = (caddr_t)B->data + B->alloced - 1; > q = (caddr_t)B->data + B->alloced - BUFF_INC_SIZE - 1; Thankyou for all the response.
Newsgroups: comp.sys.next.bugs From: Robert La Ferla <Robert_La_Ferla@hot.com> Subject: Re: need help with uucp connection Message-ID: <1994Feb14.215429.563@hot.com> Sender: robertl@hot.com Organization: Hot Technologies References: <2joago$49k@sashimi.wwa.com> Date: Mon, 14 Feb 1994 21:54:29 GMT Check your modem configuration. I also recommend using Taylor UUCP. It's available in source form for free from ftp.uu.net:/networking/uucp/taylor. Like most of the UNIX software that NeXT ships, UUCP isn't up to date. Taylor UUCP is and will give you better performance and easier maintenance. Robert La Ferla Hot Technologies NEXTSTEP ISV and Consulting In article <2joago$49k@sashimi.wwa.com> dave@sashimi.wwa.com (David Vrona) writes: > Hi all, > > I have a problem with a Next machine that polls my site for news and email. It > seems that the uucp session dies after a successful login. I'm not a Next user > so I don't know how to solve this problem. I was hoping to get some help from > the net. > > Any and all tips would be greatly appreciated. I hope I am posting to the > right group. > > > > BEGIN L.sys ENTRY > > wwa Any DIR 38400 cufb\ > "" ATDT95555555\ > ogin: uXXXXXX assword: YYYYYY > > BEGIN DEBUGGIN SESSION OUTPUT > > opus:1# /usr/lib/uucp/uucico -r1 -x6 -swwa > root wwa (2/1-11:39-1476) DEBUG (Local Enabled) > finds (wwa) called > getto: call no. cufb for sys wwa > Using DIR to call > Opening /dev/cufb > login called > wanted """" > got: that > send "ATDT95555555" > wanted "ogin:" > ATDT95555555 > CONNECT 38400 > > This is WorldWide Access (SM) > Lake County, IL Node > > enter "bbs" to enter the bulletin board system > enter "guest" to get more information > or proceed if you have shell access > > > \007Connected - 14400/ARQ/V42b [8/N/1] on termi1E > > 2/1/1994 11:41:38 current users = 9 > login:got: that > send "uXXXXXX" > wanted "assword:" > uXXXXXX > Password:got: that > send "YYYYYY" > root wwa (2/1-11:39-1476) SUCCEEDED (call to wwa ) > imsg looking for SYNC< > NO CARRIER > root wwa (2/1-11:41-1476) TIMEOUT (wwa) > opus:2# > > > END DEBUGGING SESSION OUTPUT > -- > David Vrona N9QNZ +1 708 367 1870 (voice) Internet: dave@wwa.com > Computing Engineers Inc. +1 708 367 1871 (data) +1 708 367 1872 (fax) > Home of WorldWide Access (SM), Internet services for the Chicagoland area. > Send e-mail to info@wwa.com for information, or call +1 708 367 1871 (data).
Newsgroups: comp.sys.next.bugs From: kurt@frsvnsvn.irvine.ca.us (Kurt Werle) Subject: Re: Compile problems w/ term112 on NSFIP Message-ID: <1994Feb15.210923.17788@frsvnsvn.irvine.ca.us> Organization: Little to None References: <1994Feb7.203431.1391@polar> <1994Feb13.174811.26787@beaver.cs.washington.edu> Date: Tue, 15 Feb 1994 21:09:23 GMT In article <1994Feb13.174811.26787@beaver.cs.washington.edu>, Paul F Bugni <bugni@cs.washington.edu> wrote: >I'm trying to install term on my NSFIP and am running into the following >compiling problems. Has anyone met these with success or do you have >suggestions as to what I can do to fix the problem. > >Compile errors: > >cc -O -DLOGIN_SHELL -funsigned-char -c lib.c ... errors caused by not knowing what caddr is. I will submit my fix to the term dude, but meanwhile all you need to do is #include <sys/types.h> Which I did using, at the end of the "includes.h" file, #else #include <sys/types.h> after the #if !defined(NeXT) statement... The program kicks!! Kurt p.s. This worked for me under NS 3.1
Newsgroups: comp.sys.next.bugs From: bruce@TotSysSoft.com (Bruce Gingery) Subject: Re: freezing windows Message-ID: <1994Feb15.214557.4180@ToTSySSoft.com> Sender: news@ToTSySSoft.com Organization: Total System Software References: <CL52GK.6zC@moksha.uucp> Date: Tue, 15 Feb 1994 21:45:57 GMT In article <CL52GK.6zC@moksha.uucp> mgb@moksha.uucp (Michael Branton) writes: -> -> Here's a strange one: On a NeXT turbo slab, we experience -> all windows freezing their positions. You can launch an -> app, but once it opens a window, you can't move it. Two -> entries in the message log seem like they might have some -> bearing on the problem. One occurs during boot, the other -> within 30 seconds -> after boot complete : I don't match any of the error messages you mention, but under 3.1 (black) I've occasionally had that problem under the following condition... 1. Primary swapfile has HIWAT set but is "prefer"red 2. Secondary swapfile was not mounted, even though specified and secondary disk was mounted at boot in /etc/fstab When the primary swapfile fills, EVERYTHING freezes until additional swap space can be found, i.e. through deallocation of tasks, BUT you can't get to most of the controls to shut down any GUI tasks, so must depend on background to free some space. The Workspace runs slower and slower, able to open new windows until nearly the last moment. If you have a disk activity light showing, or can hear the seeking, it's going WILD! Seems to be worse with swap compression "turned -on" in swaptab, but this is subjective... I can't prove this. Bruce Gingery bruce@TotSysSoft.com
Newsgroups: comp.sys.next.bugs From: finkenst@enmu.edu (Tim Finkenstadt) Subject: Re: DOOM problem: STBAR not found Sender: news@math.enmu.edu Message-ID: <1994Feb15.235738.6003@math.enmu.edu> Date: Tue, 15 Feb 1994 23:57:38 GMT References: <2jbs6i$ipd@ftp-p.mccaw.com> Organization: Mathematical Sciences Department In article ipd@ftp-p.mccaw.com, jason_fosback@psca.com (Jason Fosback) writes: > Like many people, I grabbed the latest version of Doom off the archives > yesterday. I was very excited that it was FAT, but I was more excited > that I could get all of the levels by copying the DOS doom.wad file into > the NeXTSTEP app wraper. > > Well, I replaced doom1.wad with the full doom.wad (calling it doom1.wad). > Things looked promising, and Doom reported that I had a registered > version. But then, I got the following error during the status bar > initialization: > > W_GetNumForName: STBAR Not Found! > > What's the scoop with this? The only thing I can think of is that the DOS > version of Doom is 1.1, and the NeXTSTEP version is 1.2. > > Anyone know what the scoop is? > > Thanks! > > -jason > ____________________________________________________________ > Jason Fosback, Systems Analyst | No sir, I didn't like it > --- Paradigm Systems Corp --- | -R&S > Internet: jason.fosback@psca.com | Star Trek: > NeXT mail: jason.fosback@psca.com | The NeXT Generation... We've had the same problem here with copying the 1.1 full version into the 1.2 wad file... we think it has to be the 1.2 v 1.1.. anyone else have the same problem? --- Timothy Finkenstadt | finkenst@math.enmu.edu Eastern New Mexico University | finkenst@ziavms.enmu.edu
Newsgroups: comp.sys.next.bugs From: Jolly alias Patrick Stein Subject: BUGS in AM.1.14 -> AM.1.15 Sender: news@informatik.uni-muenchen.de (News System) Message-ID: <CLCDGt.4J0@informatik.uni-muenchen.de> Date: Thu, 17 Feb 1994 00:02:52 GMT Organization: Institut fuer Informatik der Universitaet Muenchen AM.1.15 / 2/17/94 HiHo :) sorry there were old files in the am package ( uucpbug ... ) AM is now Available via ftp from: ftp.informatik.uni-muenchen.de /pub/comp/platforms/next/Communication/programs/am.1.15.s.tar.gz --- so long - jolly ======================================================= Jolly alias Patrick Stein = jolly@cis.uni-muenchen.de Tel: ++49 +89 -211 06 70 (priv) +89 -950 57 34 " If any member of the family should die whilst in the shelter from contamination put them outside - but remember to take them first for identification purposes " - fgh =======================================================
Newsgroups: comp.sys.next.bugs From: alex@hal.rhein-main.de (Alexander Lehmann) Subject: Wrong time with DOS filesystem + fix (NS3.2 version) Content-Type: text/plain; charset=US-ASCII Message-ID: <CLC0tE.98K@hal.rhein-main.de> Content-Transfer-Encoding: 7bit Mime-Version: 1.0 Date: Wed, 16 Feb 1994 19:29:37 GMT I have posted the same article before, but since is in 3.3 also, here it is again. First of all, if you live on the westcoast and doen't care about DST, you can ignore this. I ran across a problem with the MSDOS filesystem in NS>=3.0. All files that are copied from a MSDOS disk onto the NeXT filesystems have a datestamp that is 9 hours newer that when read on a real (oh boy) MSDOS system. I checked into this problem and found out that the time are listed correctly, if PST is the current timezone. This could be considered a problem you have to live with rather than a bug, since the DOS driver is linked into the kernel and struct tm doesn't have a field for the timezone. On the other hand, it rather annoying to have the files suddenly beamed into the future. So I disassembled the dosfs_reloc program and found the functions that convert the MSDOS style times to UNIX times and vice versa. The problem is very simple to fix, since there only two addl # statements that add resp. subtract the seconds offset from GMT. I wrote a small perl program to do this automatically. (at the end of this post) If you want to apply this patch, first of all create a backup of /usr/filesystems/DOS.fs, since you might break something. (I don't think so, but who knows) You will have to change the $localtime variable at the start of the program to your timezone. e.g. GMT 0 PDT DST -7*3600 MET +1*3600 (that's where I live) then just execute 'patchmsd.pl' (this will not change anything, even if you are root). This will print the current time offset and complain, if it wants to change it. To actually change the offset, you have to become superuser and execute 'patchmsd.pl -p' If you have inserted a DOS disk since the last boot, you will have to reboot for the changes to become active (I think that dynamically loaded kernel programs are copied into memory and not paged readonly). After this procedure, DOS filetimes should be correct. If you are still using 3.0 or 3.1, you can apply the patch by doing 'patchmsd.pl -3[01]' resp. 'patchmsd.pl -3[01] -p', but this in not tested, since I just moved to 3.2. To make life easy, the program will change the what string in the file to identify the timezone the driver is patched, this will look something like this: PROGRAM:dosfs_reloc PROJECT:dos-32 DEVELOPER:root PATCHED: to GMT+2 Once again, I'd like to stress that it is important to create a backup copy of the filesystem driver and, of course, you should do this when nobody does any real work on the machine, since a kernel panic might occur, if something goes wrong. (It didn't happen with this fix, but you have to expect problems when dealing with the intestines of a system) And, of course, you're on your own, I feel not responsible when you screw up your box. O.K. here come the program: ------------------------------------------------------------------------------ #!/bin/sh # This is a shell archive (produced by shar 3.49) # To extract the files from this archive, save it to a file, remove # everything above the "!/bin/sh" line above, and type "sh file_name". # # made 02/16/1994 19:22 UTC by alex@hal # Source directory /alex/patchmsd # # existing files will NOT be overwritten unless -c is specified # # This shar contains: # length mode name # ------ ---------- ------------------------------------------ # 2525 -rwxr-xr-x patchmsd.pl # # ============= patchmsd.pl ============== if test -f 'patchmsd.pl' -a X"$1" != X"-c"; then echo 'x - skipping patchmsd.pl (File already exists)' else echo 'x - extracting patchmsd.pl (Text)' sed 's/^X//' << 'SHAR_EOF' > 'patchmsd.pl' && #! /usr/local/bin/perl # # patchmsd.pl: patch the time offset for MSDOS disks # # $Id: patchmsd.pl,v 1.10 1994/02/16 14:29:46 alex Exp $ # # written by Alexander Lehmann <alex@hal.rhein-main.de> # # use this at your own risk # X if( $ARGV[0] eq "-30" ) { # offsets for NeXTStep 3.0 X shift; X $offset1=0x1e00+988; X $offset2=0x1b58+988; X $whatstr=0x6f94; } elsif( $ARGV[0] eq "-31") { # offsets for NeXTStep 3.1 X $offset1=0x1e52+988; X $offset2=0x1bc0+988; X $whatstr=0x7114; } else { # offsets for NeXTStep 3.2 X $offset1=0x222e; X $offset2=0x1f9c; X $whatstr=0x7118; } X $localtime=+1*3600; # CET (change this to suit your needs) X $fsfile="/usr/filesystems/DOS.fs/dosfs_reloc"; #$fsfile="dosfs_reloc"; X open(FILE,$fsfile) || die "Can't read $fsfile\n"; X seek(FILE,$offset1,0) || die "Can't seek to filepos\n"; read(FILE,$inp,6)==6 || die "Can't read data\n"; ($c,$v1)=unpack("Sl",$inp); X $c==0x0680 || die "not an addl # command\n"; X seek(FILE,$offset2,0) || die "Can't seek to filepos\n"; read(FILE,$inp,6)==6 || die "Can't read data\n"; ($c,$v2)=unpack("Sl",$inp); close FILE; X $c==0x0680 || die "not an addl # command\n"; X $v1==-$v2 || die "positive and negative time offsets don't aggree\n"; X print "time offset is set to GMT"; if($v2<0) { X print $v2/3600; } elsif($v2>0) { X print "+",$v2/3600; } X print "\n"; X if($v2==$localtime) { X print "MSDOS timeoffset agrees with local time\n"; X exit 0; } X if( int(@ARGV)!=1 || $ARGV[0] ne "-p" ) { X print "Use patchmsd.pl -p to patch the time offset\n"; X exit 0; } X open(FILE,"+>>$fsfile") || die "Can't write $fsfile\n"; X seek(FILE,$offset1+2,0) || die "Can't seek to filepos\n"; syswrite(FILE,pack("l",-$localtime),4) || die "Can't write data\n"; X seek(FILE,$offset2+2,0) || die "Can't seek to filepos\n"; syswrite(FILE,pack("l",$localtime),4) || die "Can't write data\n"; X seek(FILE,$whatstr,0) || die "Can't seek to filepos\n"; read(FILE,$str,1024)>0 || die "Can't read data\n"; X if(substr($str,0,4) ne "@(#)") { X die "Can't change what string\n"; } X if(($null=index($str,"\0"))!=$[) { X $str=substr($str,0,$null); } X if(($start=index($str,"BUILT"))!=$[-1 || ($start=index($str,"PATCHED"))!=$[-1) { X $len=length($str)-$built; X $pat="PATCHED: to GMT"; X if($localtime<0) { X $pat .= sprintf ("%d", $localtime/3600); X } elsif($localtime>0) { X $pat .= sprintf ("+%d", $localtime/3600); X } X X substr($str,$start,$len)="$pat\0"; X X seek(FILE,$whatstr,0) || die "Can't seek to filepos\n"; X print FILE $str; } X close FILE; X print "file successfully patched\n"; X SHAR_EOF chmod 0755 patchmsd.pl || echo 'restore of patchmsd.pl failed' Wc_c="`wc -c < 'patchmsd.pl'`" test 2525 -eq "$Wc_c" || echo 'patchmsd.pl: original size 2525, current size' "$Wc_c" fi exit 0 ------------------------------------------------------------------------------ You are free to distribute this program as you wish, but please leave my name in. That's about it, bye ... Alexander Lehmann -- Alexander Lehmann, alex@hal.rhein-main.de | "On the Internet, or alexlehm@iti.informatik.th-darmstadt.de | nobody knows you're a dog."
Newsgroups: comp.sys.next.bugs From: swany@martha.utcc.utk.edu (Martin Swany) Subject: Adaptec 1542C and drives >1GB Message-ID: <1994Feb17.002809.18248@martha.utcc.utk.edu> Sender: usenet@martha.utcc.utk.edu (USENET News System) Organization: Univ. of Tenn. Computing Center, Knoxville Date: Thu, 17 Feb 1994 00:28:09 GMT Hi all! I'm having trouble with a 1.02 GB drive and an Adaptec 1542C. I know that DOS support for drives > 1 GB must be turned off for NS to run. But DOS won't see the drive if that switch is turned off. I'd like to have multiple partitions on this one drive. I was able to install NS with the switch turned on, but as soon as I run it I get a kernel panic. Anyone know of any workarounds? thanks in advance, m -- Martin Swany, swany@utkux.utcc.utk.edu
From: hickman@cse.unl.edu (Hubert B. Hickman) Newsgroups: comp.sys.next.bugs Subject: Sound problems on white hardware (TokenExpress, Intel GX, Compaq) Date: 18 Feb 1994 02:17:29 GMT Organization: University of Nebraska--Lincoln Distribution: world Message-ID: <2k18fp$sha@crcnis1.unl.edu> Keywords: TokenExpress, Intel GX/Pro, Compaq, NS3.2 We are having a hard time with sound recording on white hardware. Here is the situation: 1) We have several Intel GXs with a mixed bag of networking cards (none, EtherExpress, TokenExpress). We foudn that the 3.1 beta sound driver worked pretty well and were pleased. However, after upgrading to 3.2 we regularly have problems out of the GX audio when we record. Playing sounds is not a problem. Recording sounds *is* a problem. We can frequently get the GXs into a high-pitched feedback squeel that will rattle your brain. We have checked with NeXTAnswers and have obtainted the newest ECUs and BIOS upgrades from Intel, but this doesn't help the situation. Again, the 3.1 beta driver worked much better. 2) For the GXs with the TokenExpress cards, we have even worse problems. Under 3.1 and 3.2 we tested the GXs with ProAudio Spectrum cards and it works fine EXCEPT for the machines with the TokenExpress cards. The same behavoir has been observed with Compaq DeskPros with TokenExpress cards after upgrading from 3.1 to 3.2. Audio worked in 3.1, and does not record in 3.2. I think we may be looking at two problems here: 1) A very fragile GX audio driver in 3.2 for recording. 2) Problems with the PAS and the TokenExpress cards. Sooo - I would appreciate hearing from people using TokenExpress cards who have successfully recorded sound with the PAS cards, regradless of configuration. - If other people using GXs are having the same sound recording problem in 3.2, we would love to heard about it. Likewise, if you are using the GX audio heavily in 3.2 for recording the little problems, I would love to hear from you. Thanks, Hubert Hickman Hickman-Kenyon Systems, Inc. hhickman@andi.org (NeXTMail) (402) 697-1310 (402) 330-8613 (FAX)
Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs From: Greg_Anderson@afs.com (Gregory H. Anderson) Subject: BUG with Panel: canBecomeKeyWindow Message-ID: <1994Feb18.165748.971@afs.com> Sender: greg@afs.com Date: Fri, 18 Feb 1994 16:57:48 GMT To force a Panel to never become the key window (not even when you click its title bar), you must include this override in your Panel subclass: - (BOOL)canBecomeKeyWindow { return NO; } Here's the bug: If a later message from elsewhere in your program tells the Panel to setBecomeKeyOnlyIfNeeded:YES, it starts ignoring the return value from canBecomeKeyWindow and allows the window to become key anyway. You might argue that setBecomeKeyOnlyIfNeeded:YES is an intentional change in "when can the Panel become key?" state, but I don't buy it. A canBecomeKeyWindow override is only implemented when you intend an absolute restriction, and it should be irrevocable. Workaround: When you override canBecomeKeyWindow, also override doesBecomeKeyOnlyIfNeeded (return NO) and setBecomeKeyOnlyIfNeeded: (ignore the argument). This will fully protect your intent to keep the Panel from becoming key under any circumstances. This is a confirmed bug in NS 3.2. I have not tried any other version. It's not a major inconvenience, but it tripped me up for a while. -- Gregory H. Anderson | "History is the fiction we invent to Revisionist Autobiographer | persuade ourselves that events are Anderson Financial Systems | knowable and that life has order and greg@afs.com (NeXTmail OK) | direction." -- Calvin & Hobbes, 7/19/93
From: robert@steffi.demon.co.uk (Robert Nicholson) Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs Subject: Re: BUG with Panel: canBecomeKeyWindow Date: 18 Feb 1994 22:36:57 -0000 Organization: me organized? That's a joke! Message-ID: <2k3fu9$3c5@steffi.demon.co.uk> References: <1994Feb18.165748.971@afs.com> Greg_Anderson@afs.com wrote in comp.sys.next.programmer,comp.sys.next.bugs >To force a Panel to never become the key window (not even when you click >its title bar), you must include this override in your Panel subclass: > >- (BOOL)canBecomeKeyWindow >{ >return NO; >} > Do that without subclassing Mr Griffiths :-) -- "You know what's wrong with you?" (Audrey Hepburn, Cary Grant) "No, what?" "Nothing" (Charade, 1963) (ASCII for text only messages)
Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs From: Jayson Adams Subject: Re: BUG with Panel: canBecomeKeyWindow Message-ID: <1994Feb19.024717.2966@millennium.com> Sender: jayson@millennium.com Organization: Millennium Software Labs, Inc. Posting-Application: Newsgrazer Pro References: <1994Feb18.165748.971@afs.com> Date: Sat, 19 Feb 1994 02:47:17 GMT In article <1994Feb18.165748.971@afs.com>, Gregory H. Anderson writes: > Here's the bug: If a later message from elsewhere in your program tells > the Panel to setBecomeKeyOnlyIfNeeded:YES, it starts ignoring the return > value from canBecomeKeyWindow and allows the window to become key anyway. > > Workaround: When you override canBecomeKeyWindow, also override > doesBecomeKeyOnlyIfNeeded (return NO) and setBecomeKeyOnlyIfNeeded: > (ignore the argument). Here's a cleaner workaround: don't send setBecomeKeyOnlyIfNeeded:YES to the panel. __jayson
Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs From: dave@prim.demon.co.uk (Dave Griffiths) Subject: Re: BUG with Panel: canBecomeKeyWindow Message-ID: <1994Feb19.134436.484@prim.demon.co.uk> Organization: Primitive Software Ltd. References: <1994Feb18.165748.971@afs.com> <2k3fu9$3c5@steffi.demon.co.uk> Date: Sat, 19 Feb 1994 13:44:36 GMT In article <2k3fu9$3c5@steffi.demon.co.uk> robert@steffi.demon.co.uk (Robert Nicholson) writes: >Greg_Anderson@afs.com wrote in comp.sys.next.programmer,comp.sys.next.bugs >>To force a Panel to never become the key window (not even when you click >>its title bar), you must include this override in your Panel subclass: >> >>- (BOOL)canBecomeKeyWindow >>{ >>return NO; >>} >> > >Do that without subclassing Mr Griffiths :-) OK. Add this flag to Window: BOOL canBecomeKeyWindow; Modify the existing canBecomeKeyWindow method as follows: - (BOOL)canBecomeKeyWindow { return canBecomeKeyWindow; } and add this method: - setCanBecomeKeyWindow:(BOOL)canBecome { canBecomeKeyWindow = canBecome; return self; } As it stands, you're _forced_ to subclass just to configure it differently. Dave Griffiths PS: I agree with Greg that canBecomeKeyWindow should _always_ be obeyed.
Newsgroups: comp.sys.next.bugs From: boonlow@kits.sfu.ca (Boon Chong Benjamin Low) Subject: Failure to install NSFIP 3.2 Message-ID: <boonlow.761681194@sfu.ca> Sender: news@sfu.ca (seymour news) Organization: Simon Fraser University, Burnaby, B.C., Canada Date: Sat, 19 Feb 1994 18:06:34 GMT i***************************************************************i Hey guys, I have problem with installation of NSFIP 3.2 Plz help or offer me some advice if possible. I have bought NSFIP for 2 months !! already and still couldn't get it install due to one reason or another, sign ...:-( My System --------- AIR VESA motherboard 1 Western Digital IDE 340mb harddrive 1 Seagate SCSI ST31200 1 gig+ harddrive. NEC CDR 84-1 on id-5 scsi harddrive on id-0 and DPT PM 2021/90 SCSI controller on id-7 both the DPT and NEC are terminated at both ends and the harddrive is in the middle ATI Ultra Vesa w/2mb vram PAS studio sound card but I took it out while trying to install NSFIP My system has 32mb ram My Seagate SCSI harddrive is partition 500mb for DOS to try it out. The other 500mb is unused. I have Corel SCSI and my whole set up works well under DOS, Windows 3.1 and OS/2 2.1 Problems -------- Okay, I disable and took out the cable for my 51/4 b drive, also disabled in the bios. I boot the system up with installation diskette and CD in the NEC cdrom drive. reach (1) select english... (2) continue without loading additional drivers ... problem starts... :-(( The screen and this messages comes out --------------------------------------- resetting SCSI bus Eata controller at irq 15 registering irq15 ( it should be for my DPT controller ) sd0 : disk not ready type 1 to install NeXTSTEP (1) IDE disk #1 io-status=13, scsi-status=8, sense key=0 too many tries cd-rom disk could not be used installation cannot continue that's all, folks, sign... From then on, system will stop. I try to change interrupt, IO address e.t.c on my DPT card, and two days with this hell, I am still stuck with the same message. My WD IDE is on a AT controller, could that pose a problem ? Please help me guys, I try everything and nothing works, call tech-support and got no replies from now. Regards, Ben. (604) 420-3176
Newsgroups: comp.sys.next.bugs From: thf@zelator.de (Thomas Funke) Subject: Re: Compile problems w/ term112 on NSFIP Message-ID: <1994Feb14.121013.592@gamelan.uucp> Sender: thomas@gamelan.uucp (thomas) Organization: Disorganization References: <1994Feb13.174811.26787@beaver.cs.washington.edu> Date: Mon, 14 Feb 1994 12:10:13 GMT In article <1994Feb13.174811.26787@beaver.cs.washington.edu> bugni@cs.washington.edu (Paul F Bugni) writes: > p = (caddr_t)B->data + B->alloced - 1; > lib.c:174: illegal statement, missing `;' after `)' OK, let's have a short course in C-debugging: "Illegal statement": What does this message mean ? The statement is wrong. "Missing ; after )": The compiler thinks, the statement should be finished after the closing ')'. Let's have a look, where is a ')' ? Yeah: (caddr_t). Now comes the most difficult part: What does caddr_t mean? Obviously nothing for the compiler, therefore an error message. How can we solve that ? We must give caddr_t a meaning. How ? First, let's open HeaderViewer.app and look for 'caddr_t'. Great, HeaderViewer found it in sys/types.h. Now edit lib.c and insert #include <sys/types.h> at the beginning somewhere. Type 'make' again. Whow, it works. -- ------------------------------------------------------------------ Thomas Funke ** Unix-Consultant ** thf@zelator.de Brooks's Law: Adding manpower to a late software project makes it later ------------------------------------------------------------------
From: swiet@syzygy.uwaterloo.ca (Alexander Swietlicki) Newsgroups: comp.sys.next.bugs Subject: Re: Compile problems w/ term112 on NSFIP Date: 20 Feb 94 16:05:32 Organization: University of Waterloo Message-ID: <SWIET.94Feb20160532@syzygy.uwaterloo.ca> References: <1994Feb13.174811.26787@beaver.cs.washington.edu> <1994Feb14.121013.592@gamelan.uucp> In-reply-to: thf@zelator.de's message of Mon, 14 Feb 1994 12:10:13 GMT In article <1994Feb14.121013.592@gamelan.uucp> thf@zelator.de (Thomas Funke) writes: Yeah: (caddr_t). Now comes the most difficult part: What does caddr_t mean? That some former lisp programmer is getting even. :)
From: sanguish@digifix.com Newsgroups: comp.sys.next.announce,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.bugs,comp.sys.next.programmer,comp.sys.next.marketplace Subject: Quick Guide to NEXTSTEP information on the Internet Date: 20 Feb 1994 23:06:07 -0500 Organization: Next Announcements Message-ID: <2k9bvf$6po@digifix.digifix.com> This post is made weekly, to help 'point' users to more NEXTSTEP information Topics include: comp.sys.next newsgroups related newsgroups comp.sys.next newsgroups mailing list ftp sites NeXTanswers comp.sys.next.* newsgroups -------------------------- Comp.Sys.Next.Advocacy This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. Comp.Sys.Next.Announce Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) The NEXTSTEP FAQs are posted here monthly as well. This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Comp.Sys.Next.Bugs A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT- specific groups as well. Comp.Sys.Next.Hardware Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. Comp.Sys.Next.Marketplace NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. Comp.Sys.Next.Misc For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! Comp.Sys.Next.Programmer Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. The NEXTSTEP programmer FAQs are posted here. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. Comp.Sys.Next.Software This is a place to talk about [third party] software products that run on NEXTSTEP systems. Comp.Sys.Next.Sysadmin Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. related Newsgroups ------------------ Comp.Soft-Sys.Nextstep Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. Comp.Lang.Objective-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. Comp.Object Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original Comp.Sys.Next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News ------------------------------------------- Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep-d next-advocacy-d next-announce-d next-bugs-d next-hardware-d next-marketplace-d next-misc-d next-programmer-d next-software-d next-sysadmin-d (For a full description, send mail saying LISTS to <digestif@antigone.com>). The subscription syntax is essentially the same as LISTSERV's. To subscribe, send a message to <digestif@antigone.com> saying: SUB Listname YourName Example: SUB next-hardware-d John Doe The ftp sites ------------- cs.orst.edu: The main site for North American submissions nova.cc.purdue.edu: Lots of older stuff, but very short on disk space ftp.informatik.uni-muenchen.de: In Germany. terra.stack.urc.tue.nl (Dutch NEXTSTEP User Group) and cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp.next.com: See the below ftp.next.com and NextAnswers@next.com ------------------------------------- From the document 1000_Help from ftp.next.com Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, or you can transfer them by anonymous ftp. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to NeXTanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO <your-address> If you have any problem with the system or suggestions for improvement, please send mail to NeXTanswers-request@NeXT.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to NeXTanswers-request@NeXT.com. Written by: Eric P. Scott eps@toaster.SFSU.EDU and Scott Anguish sanguish@digifix.com Additions from: Greg Anderson (Greg_Anderson@afs.com) and Michael Pizolato (Michael_Pizolato@afs.com)
From: izumi@pinoko.berkeley.edu (Izumi Ohzawa) Newsgroups: comp.sys.next.bugs,comp.sys.next.software Subject: Stuck loginwindow screen saver Date: 21 Feb 1994 19:05:21 GMT Organization: University of California, Berkeley Distribution: world Message-ID: <2kb0lh$pmk@agate.berkeley.edu> On three of our machines, screen saver of loginwindow gets stuck in the save mode. Mouse clicks or keyboard events do not restore the screen. We can login if we type login and password correctly blindly (processes are running as confirmed via rlogin from another machine). However, the screen remains mostly black with NeXT's logo and the hostname. It is hard to duplicate this as it happens maybe once or twice a week, but definitely happens. This is on Black 040 with 3.2. Has anybody seen this happen? Any fix? -- Izumi Ohzawa [ $@Bg_78^=;(J ] USMail: University of California, 360 Minor Hall, Berkeley, CA 94720 Telephone: (510) 642-6440 Fax: (510) 642-3323 Internet: izumi@pinoko.berkeley.edu (NeXTMail OK)
Newsgroups: comp.sys.next.bugs From: nathan@laplace.csb.yale.edu (Nathan F. Janette) Subject: Re: Stuck loginwindow screen saver Message-ID: <1994Feb22.031157.21767@cs.yale.edu> Sender: news@cs.yale.edu (Usenet News) Organization: Yale University, Department of Computer Science, New Haven, CT References: <2kb0lh$pmk@agate.berkeley.edu> Date: Tue, 22 Feb 1994 03:11:57 GMT In article <2kb0lh$pmk@agate.berkeley.edu> izumi@pinoko.berkeley.edu (Izumi Ohzawa) writes: > On three of our machines, screen saver of loginwindow gets stuck > in the save mode. Mouse clicks or keyboard events do not restore > the screen. We can login if we type login and password correctly > blindly (processes are running as confirmed via rlogin from another > machine). However, the screen remains mostly black with NeXT's logo > and the hostname. > > It is hard to duplicate this as it happens maybe once or twice > a week, but definitely happens. This is on Black 040 with 3.2. > Has anybody seen this happen? Any fix? Yes, I've seen this on our DEC 486 running NEXTSTEP 3.2. I've also seen Gateway 2000 486 systems with ATI video cards running 3.2 exhibit a related (?) problem: the progress pie graph in the installer app sometimes stops updating for a while, and then suddenly jumps ahead. -- Nathan "USENET" Janette PPP link from hilbert.csb.yale.edu Please reply to: nathan@laplace.csb.yale.edu (NeXT)
Newsgroups: comp.sys.next.bugs From: "TWARECKI,PIOTR,MR" <B7NS000@MUSICB.MCGILL.CA> Subject: Disappearing (?) harddrive free space. Message-ID: <21FEB94.24744879.0240@VM1.MCGILL.CA> Sender: usenet@MUSICB.MCGILL.CA Organization: McGill University Date: Tue, 22 Feb 1994 03:54:42 GMT Hello, I apologize if this is a wrong group to post this. I have recently noticed that the amount of free harddrive space on my system was mysteriously shrinking from time to time. Since I have just purchased the machine, I'm not sure if this is a real problem or if there is some explanation to it. Here is what happends. After a fresh installation of NS3.2 on my 16/425 cube I was left with approximately 143M of free space. At some point I have noticed that this shrunk to 140M. Suspecting the swapping being responsible for it, I started almost all my applications (!) to get the machine to swap. I saw the space shrink to 115M. Then I restarted the machine from scratch to shrink the swap file, and I was left with 135M after the booting. I repeated the process, starting a few more applications, and following a restart I was left with 134M of free space. This implies that 6M has mysteriously disppeared from my drive. Is this an unusual behaviour, or is the system reserving some space on the disk other than the swap file itself? Any help help or hints are highly appreciated. If this is something really simple, I appologize for my ignorance. Thank you, Peter. Peter Twarecki <b7ns@MUSICB.McGill.CA>
Newsgroups: comp.sys.next.admin,comp.sys.next.bugs From: ses3@ellis.uchicago.edu (Steven Schwartz) Subject: Tape Drive Block Fix Message-ID: <1994Feb22.220005.12778@midway.uchicago.edu> Summary: Trying to do a restore Sender: news@uchinews.uchicago.edu (News System) Organization: University of Chicago -- Academic Information Technologies Date: Tue, 22 Feb 1994 22:00:05 GMT Suffered a major server crash here, and there's just one small problem: We have tape backups, but the backups won't run without a program my fellow sysadmin acquired named block.c, for handling block size inconsistencies with certain tape drives. If anyone still has the source for that program, I humbly beg, plead, and grovel --> Send it to me. For specifics: Trying to do a restore off a level 0 dump, both /usr/etc/restore rf /dev/rst0 yields: Tape/disk block size 512 not a multiple of dump block size (1024) /usr/etc/restore rfb /dev/rst0 (either 1024 or 512) yields: Tape/disk read error while trying to set up tape/disk (prompt to continue) Tape is not a dump tape. This is why we need block.c. Does anyone have the source code for this or know where we could find it? (I believe this program reset the default block size for dumps, permitting them to work.) Steven E. Schwartz, Academic Information Technologies 1-312-702-2580 email ses3@midway.uchicago.edu
From: erikkay@next.com (Erik Kay) Newsgroups: comp.sys.next.bugs Subject: Re: Stuck loginwindow screen saver Date: 23 Feb 1994 01:17:34 GMT Organization: NeXT, Inc. Message-ID: <2keare$7dr@rosie.next.com> References: <1994Feb22.031157.21767@cs.yale.edu> In article <1994Feb22.031157.21767@cs.yale.edu> nathan@laplace.csb.yale.edu (Nathan F. Janette) writes: > In article <2kb0lh$pmk@agate.berkeley.edu> izumi@pinoko.berkeley.edu (Izumi > Ohzawa) writes: > > On three of our machines, screen saver of loginwindow gets stuck > > in the save mode. Mouse clicks or keyboard events do not restore > > the screen. We can login if we type login and password correctly > > blindly (processes are running as confirmed via rlogin from another > > machine). However, the screen remains mostly black with NeXT's logo > > and the hostname. > > > > It is hard to duplicate this as it happens maybe once or twice > > a week, but definitely happens. This is on Black 040 with 3.2. > > Has anybody seen this happen? Any fix? Well, I can give you a good guess what's going on. There seems to be a race condition in loginwindow under 3.2 (and 3.1) that occassionally rears its ugly head. What happens is that to clear the screensaver, you click on a mouse button. This can lead to a race between the window packages in loginwindow where it will order the window out, and the packages order it back in right away. A workaround for 3.2 is to not use the mouse button to clear the loginwindow screensaver. if you always clear it by moving the mouse or typing at it, the screensaver should come back. If you forget at some point and hit the mouse button and the race happens (it's actually pretty rare... I really have a lot of trouble reproducing it), there's a workaround for you too. loginwindow thinks that the screensaver is off at this point, so if you wait for the screensaver to kick back in again (the cube will start moving again, and the mouse cursor won't be visible), you can clear it with a keystroke or mouse movement. If you have your loginwindow TimeToDim down around 10 or 15 seconds, this isn't such a painful thing (the default time is 5 minutes). hope that helps, Erik
Newsgroups: comp.sys.next.bugs From: "TWARECKI,PIOTR,MR" <B7NS000@MUSICB.MCGILL.CA> Subject: SOLVED: Disappearing (?) harddrive free space. Message-ID: <22FEB94.22601374.0086@VM1.MCGILL.CA> Sender: usenet@MUSICB.MCGILL.CA Organization: McGill University References: <21FEB94.24744879.0240@VM1.MCGILL.CA> Date: Wed, 23 Feb 1994 01:55:37 GMT Thank you very much to everyone who responded for all the help. I have received over 10 responses in just a few hours. With everyone's help, I was able to determine that it was indeed the swapfile that was responsible for the reduction in the free space. It seems that after the installation of NS3.2 the swapfile was way below the 16MB mark. When I started experimenting with swapping, the swapfile grew reducing the free space. After I restarted the computer, the swapfile was shrunk, but only to the low mark of 16MB or so, as it is supposed to. This seems to account for the 'missing' space. Since then I repeated my experiment and run all the applications I could, but the free space remains steady now, which confirms the above cause. Thanks again, Peter. Peter Twarecki <b7ns@MUSICB.McGill.CA>
Newsgroups: comp.sys.next.software,comp.sys.next.bugs,comp.sys.next.misc From: andrew@stone.com (Andrew Stone) Subject: Stone Design's Bug & Info email aliases now working... Message-ID: <1994Feb22.203502.1387@stone.com> Keywords: Stone Design Email Bounce Sender: andrew@stone.com Organization: Stone Design Corp Date: Tue, 22 Feb 1994 20:35:02 GMT If you had trouble sending email to info@stone.com or bugs@stone.com over the past 10 days (didn't get a response or got a bounced email), please retry - we've got it 'under control'. andrew -- ||<<->>||<<==>>||<<++>>||<<?>|<+>>||<<-->>||<<==>>||<<+>>|| !! Andrew Stone | (505) 345-4800 !! !! andrew@stone.com | Stone Design Corp !! ||<<->>||<<==>>||<<++>>||<<?>|<+>>||<<-->>||<<==>>||<<+>>||
From: mra@bink.mit.edu (Michael Raskin Andrews) Newsgroups: comp.sys.next.bugs Subject: /usr/etc/dump Date: 23 Feb 1994 17:43:14 GMT Organization: Massachvsetts Institvte of Technology Message-ID: <2kg4ji$glf@senator-bedfellow.MIT.EDU> I'm still having the problem with dumping: bink> /usr/etc/dump 0ufs /dev/rxt0 1200000 /dev/rsd0a DUMP: Date of this level 0 dump: Wed Feb 23 08:21:36 1994 DUMP: Date of last level 0 dump: the epoch DUMP: Dumping /dev/rsd0a (/) to /dev/rxt0 DUMP: mapping (Pass I) [regular files] DUMP: mapping (Pass II) [directories] swent_start: No such file or directory DUMP: estimated 650040 tape blocks on 0.03 tape(s). DUMP: dumping (Pass III) [directories] ... ... ... Note the "swent_start: No such file or directory" line. What the heck does this mean? Any response appreciated. Thanks. --- Michael Andrews
From: wolfgang@wi.WHU-Koblenz.de (Wolfgang Roeckelein) Newsgroups: comp.sys.next.bugs Subject: AdaptorOracle uses tables not included in the default V7 oracle catalog Date: 21 Feb 1994 17:11:32 GMT Organization: WHU Koblenz Message-ID: <2kaq04$j16@obelix.WHU-Koblenz.de> Hello everybody, perhaps this is of interest to some: NeXTstep 3.2 AdaptorOracle uses system tables which are not included in the default V7 oracle catalog (esp. accessible_tables etc.) Therefore DBModeler can't build a default model! As a temporary solution you can advice DB-Admins to execute catalog6.sql on their V7 server!! A better solution would be for NeXT to use the newer system tables! -- Dipl.-Wirtsch.-Inf. Voice: +49 261 6509 173 Wolfgang Roeckelein Fax: +49 261 6509 179 WHU Koblenz E-Mail: roeckelein@wi.whu-koblenz.de Burgplatz 2 (NeXTmail ok) D-56179 Vallendar Germany
From: griffid@gb.swissbank.com (Dave Griffiths) Newsgroups: comp.sys.next.bugs,comp.sys.next.programmer Subject: PSxyshow takes a second to draw 6000 characters Date: 23 Feb 1994 13:13:25 GMT Organization: Swiss Bank Corporation, High Timber St, London, UK Message-ID: <2kfkpl$g9k@gpo.gb.swissbank.com> Subject line says it all really. This is the main cause of a non-editable DBTableView being slow to draw. 100 lines with 60 chars/line took 1.2 seconds to draw with PSxyshow (this is under 3.2). An alternative method using PSmoveto/PSshow for each line took 0.3 seconds. Wrapping PSusertime() calls round the drawSelf proves that most of the second is spent in the Window Server. Sounds like we could do with some postscript benchmarks to test the Window Server against. Dave Griffiths PS: Hacker's tip: you can avoid the use of PSxyshow in your DBTextFormatter subclass by setting _xyShow to your own object in init. This object just needs to implement moveTo::, show:, reset:gray: and finalize.
Newsgroups: comp.sys.next.bugs From: boonlow@kits.sfu.ca (Boon Chong Benjamin Low) Subject: NSFIP can't recognize my DOS drive Message-ID: <boonlow.762043230@sfu.ca> Sender: news@sfu.ca (seymour news) Organization: Simon Fraser University, Burnaby, B.C., Canada Date: Wed, 23 Feb 1994 22:40:30 GMT " Hey guys, I planned to let NeXTSTEP to be bootable via OS/2 bootmanager. I have already installed NSFIP on my SCSI drive. During installation, I have to disable my IDE drive inorder for NSFIP to be booted up properly, noted from the advice and e-mail posted on how to install NSFIP to be bootable under OS/2 bootmanager, I even get a copy of nextanswers 1487. I went through everything carefully, I have DOS and OS/2 and OS/2 bootmanager on my IDE drive, a 500 mb partition for NSFIP on my SCSI and the rest for DOS. During installation, I disabled my IDE drive and only with my NeXTSTEP, installed it on my free 500mb on my SCSI drive. AFter everything, I type fdisk /mbr to disabled the NS bootmanager. Following the advice from the net, I went into NS, type localhost# su localhost# /usr/etc/fdisk /dev/rhd0h but I got this error messages : fdisk : unable to open /dev/rhd0h why !!! What I wanted to do was format my 7mb of free space inorder to transfer the master boot record over to the 7mb, but without access to my IDE drive, I can't get anything going, BTW, I did enabled my IDE drive first. Right now, inorder to enter NS, I have to boot from the NS diskette, and type sd()mach_kernel inorder to go into NS. I called NS tech-support, first they told me my IDE controller is not supported under NS, only intergrated IDE on the motherboard is supported, later , they said maybe not ?? secondly, they told me that NS DOESN't support DOS 6.0 because the fdisk for DOS 6 is different !!! they asked me to reinstall DOS 5 on my DOS drive before trying again, ?? before I commit myself to take out my PCDOS 6 and reinstall microsoft DOS 5.0 with alot of pain and agony, can someone confirm on the truth of this statement ?? Please, anyone, give me a hint on how I can get NS to recognize my IDE drive. Thanks alot for all your help/advice, Ben. (604) 420-3176
Newsgroups: comp.sys.next.bugs From: coconut@crash.cts.com (Brian Dear) Subject: Fax cover sheet forgetfulness Organization: CTS Network Services (CTSNET/crash), San Diego, CA Date: Thu, 24 Feb 1994 02:56:29 GMT Message-ID: <CLpK66.LF5@crash.cts.com> Sender: news@crash.cts.com (news subsystem) Is it a known bug that when you "Preview" a Fax cover sheet in NS3.2, that is, you Print something, then select Fax instead of printing it, and then create a cover sheet (click Yes and Edit) and then you go back to the Fax dialog and hit Preview, that once you're done previewing the cover sheet, it's lost and you gotta type it again? Seems to be several little quirks like this with the built-in fax stuff of NS3.2. Is this known already? (Acknowledged by NeXT as a bug? Will it ever get fixed?) - bd
Newsgroups: comp.sys.next.bugs From: fabien@free.fdn.org (Fabien Roy) Subject: Re: Tape Drive Block Fix Message-ID: <1994Feb23.202003.2365@free.fdn.org> Sender: news@free.fdn.org Organization: Fabien Roy Electronic Engineering (F.R.E.E.) - Paris, France. References: <1994Feb22.220005.12778@midway.uchicago.edu> Date: Wed, 23 Feb 1994 20:20:03 GMT In comp.sys.next.bugs article <1994Feb22.220005.12778@midway.uchicago.edu> you wrote: > Suffered a major server crash here, and there's just one small problem: > We have tape backups, but the backups won't run without a program my fellow > sysadmin acquired named block.c, for handling block size inconsistencies with > certain tape drives. > > If anyone still has the source for that program, I humbly beg, plead, and > grovel --> Send it to me. > > For specifics: Trying to do a restore off a level 0 dump, both > /usr/etc/restore rf /dev/rst0 > yields: > > Tape/disk block size 512 not a multiple of dump block size (1024) > > /usr/etc/restore rfb /dev/rst0 (either 1024 or 512) > > yields: > > Tape/disk read error while trying to set up tape/disk > (prompt to continue) > > Tape is not a dump tape. > > This is why we need block.c. Does anyone have the source code for > this or know where we could find it? > (I believe this program reset the default block size for dumps, > permitting them to work.) > > > Steven E. Schwartz, Academic Information Technologies 1-312-702-2580 > email ses3@midway.uchicago.edu to switch your tape to fixed length block: #include <sys/types.h> #include <nextdev/scsireg.h> #include <sys/mtio.h> main (argc, argv) char **argv; { int bs = 512; int fd = open("/dev/nrst0", 0); ioctl(fd, MTIOCFIXBLK, &bs); } to turn it back to variable: #include <sys/types.h> #include <nextdev/scsireg.h> #include <sys/mtio.h> main (argc, argv) char **argv; { int fd = open("/dev/nrst0", 0); ioctl(fd, MTIOCVARBLK); } Hope that helps. --fabien --------------------------------------------------------------------- Fabien_Roy@free.fdn.org (NextMail accepted) XO, 81 alle Danielle Casanova, 93320 Les Pavillons sous Bois France Tel: 33 1 4847 5198 Fax: 33 1 4847 5175
Newsgroups: comp.sys.next.bugs From: wave@media.mit.edu (Michael B. Johnson) Subject: 2 IB bugs (one crasher) in 3.2 Message-ID: <1994Feb25.011615.24298@news.media.mit.edu> Sender: news@news.media.mit.edu (USENET News System) Organization: MIT Media Laboratory Date: Fri, 25 Feb 1994 01:16:15 GMT I have sent this to bug_next, but thought folks might be interested... This consistenly crashes IB for me on black hardware, running a clean IB (no other palettes than the standard 4) under 3.2. - start up IB. - do cmd-n to open a new file. - drag a button onto the main window. - alt-drag out a matrix of them. - on the Matrix's inspector, press "inspect" to look at the prototype of the button cell for the matrix. - type cmd-3, while that ButtonCell's Attributes Inspector is up. The hope here was that I would see the size inspector of the button cell, so that I could tweak the cell's size exactly. Unfortunately, it brought up the size inspector on the Matrix. BUG #1. - now press "Close" on the ButtonCell's Inspector. IB crashes. BUG#2. --> Michael B. Johnson -- wave@media.mit.edu --> MIT Media Lab -- Computer Graphics & Animation Group --> 20 Ames St. E15-023G -- (617) 547-0563 (day office) --> Cambridge, MA 02139 -- (617) 253-0663 (night office) -- --> Michael B. Johnson -- wave@media.mit.edu --> MIT Media Lab -- Computer Graphics & Animation Group --> 20 Ames St. E15-023G -- (617) 547-0563 (day office) --> Cambridge, MA 02139 -- (617) 253-0663 (night office)
Newsgroups: comp.sys.next.bugs From: stimpy@toadstool.ann-arbor.mi.us (Gary L.) Subject: Re: 3.2 Workspace Manager bug? Message-ID: <CLr1v5.Cp7@toadstool.ann-arbor.mi.us> Organization: toadstool, Ann Arbor, MI References: <2ehtg0$bg5@disuns2.epfl.ch> <1993Dec13.191908.6245@newsgate.sps.mot.com> Date: Thu, 24 Feb 1994 22:16:14 GMT In article <1993Dec13.191908.6245@newsgate.sps.mot.com> mshaler@tdocad.sps.mot.com (Michael Shaler) writes: >Marco Scheurer writes > Hi, Is it me or is there an horrible bug in the Workspace Manager > with 3.2? It seems that sometime you cannot type a new filename > for a file, because it is automatically selected at more or less > random interval, while you type, so you are always destroying > what your are typing. > > [crunch] > >Yep, me too, but I have yet to establish repeatable conditions for this craziness in order to report it as a true, stomp it into the ground on sight BUGNeXT heteroptera... > It happens to me too, and it's driving me crazy !!! I HATE IT !!! I think it may be mild incompatablity with keyboards, when NeXTSTEP books it shows an AWK error with the keyboard. (Whatever that is, I dunno...) STiMPY -- ============================================== stimpy@toadstool.ann-arbor.mi.us stimpy@mmg.im.med.umich.edu
Newsgroups: comp.sys.next.sysadmin,comp.sys.next.bugs From: mark@ve6mgs.ampr.org (Mark G. Salyzyn) Subject: Re: Installing Serial Ports Organization: ADEC Systems Inc. Distribution: na Date: Thu, 24 Feb 94 18:38:34 GMT Message-ID: <1994Feb24.183834.25763@ve6mgs.ampr.org> Followup-To: comp.sys.next.sysadmin References: <139448@hydra.gatech.EDU> Keywords: mux driver, serial, ns/intel gt6963c@prism.gatech.EDU (John (kzin) Rudd) writes: >I picked up the MUX serial driver off of ftp.uni-stuttgart.de, >and installed it under NS/Intel. The Configure app worked just >fine, but I don't see any new devices in /dev.. Look in /usr/adm/messages and look at the log of the Mux driver's probe of the serial ports installed in the system. It should log the ports it found upon the reboot with the new configuration (and indicate that it found and identified the port as 8250 [not bloody likely, but it can tell the difference] 16450, 16550AFN, broken 16550 or 82150). The PostLoader uses the Kernel Server model to acquire the status of all the instances of the Mux driver to figure out which ports to make in /dev. If the Mux driver is not properly loaded, or does not find any probed ports, then the PostLoader will *not* make the devices. The Instance?.table for the Mux driver can override the names of the ports as well to `anything' you want ... The probe report is just before the Kernel claims the Mux driver instance is registered. >So... what do I need to do? Take a look at the /NextLibrary/Devices/SerialPorts.config/Instance?.table, /NextLibrary/Devices/Mux.config/Instance?.table and /NextLibrary/Devices/System.config/Instance?.table files and see it the values make sense for the driver. Send them to me if you don't feel they make sense ... >I'm trying to set up a way to use my internal modem for dialing >out.. Either using tip or some terminal program (is there something >avaible for free?).. and I'd eventually like to use it for SLIP. I use tip regularily at 38400 with the driver, and don't feel at a lack for capability. I think TipTop is having troubles with the locking mechanism in this driver (between /dev/cu?? and /dev/ttyd??) and have not had the opportunity to test and debug the problem (yet). It *should* have no problems if you do not set the port up as bidirectional (getty and TipTop). While I have your attention, the Mux 1.3 version of the driver *requires* the "Buffer Size" and RTS Water Mark specification in the Default.table files, I shipped it without these values in some of the example Condigurations, and some people might be tempted to use the 1.2 Instance?.table files without these configuration parameters. I have not accumulated enough bug fixes, improvements and workarounds to justify posting the 1.4 version of the driver ... Ciao, -- Mark Salyzyn
From: statman@stat.ufl.edu (Charles D. Kincaid) Newsgroups: comp.sys.next.bugs Subject: Re: 3.2 Workspace Manager bug? Date: 25 Feb 1994 14:30:46 GMT Organization: University of Florida Distribution: world Message-ID: <2kl22mINNem8@no-names.nerdc.ufl.edu> References: <CLr1v5.Cp7@toadstool.ann-arbor.mi.us> Hello, >Marco Scheurer writes > Hi, Is it me or is there an horrible bug in the Workspace Manager > with 3.2? It seems that sometime you cannot type a new filename > for a file, because it is automatically selected at more or less > random interval, while you type, so you are always destroying > what your are typing. If I understand you correctly, then this is happening to me also, but in NS3.1. It also seems to happen only on files in a directory that is mounted across the net. I assumed that the nfs connection was kind of funky. -- Sincerely, charles d. kincaid -------------------------------------------------------------------- Dept. of Statistics 'Damn fine coffee...and hot, too!' Univ. of Florida Pres: G-ville NeXT Users Group
Newsgroups: comp.sys.next.bugs From: swany@martha.utcc.utk.edu (Martin Swany) Subject: NS 3.2 on IBM ValuePoint's IDE Message-ID: <1994Feb25.165010.19704@martha.utcc.utk.edu> Sender: usenet@martha.utcc.utk.edu (USENET News System) Organization: Univ. of Tenn. Computing Center, Knoxville Date: Fri, 25 Feb 1994 16:50:10 GMT Hello All. I am having trouble installing NS 3.2 on an IBM ValuePoint's IDE drive. It's no problem to install on a SCSI drive in the same machine, but it will not acknowledge the IDE as a target disk. The ValuePoint is the 466DX2/D. thanks in advance. -- Martin Swany, swany@utkux.utcc.utk.edu
From: jhethcox@homey (John Hethcox) Newsgroups: comp.sys.next.bugs Subject: Project Builder Date: 25 Feb 1994 22:56:09 GMT Organization: Evans & Sutherland Computer Corporation Message-ID: <2klvm9$gn1@cnn.sim.es.com> Sometimes when I edit a file (using Edit), save it, and press the Run button in PB the app fails to rebuild (it thinks everything is up-to-date). I had this problem often with NS 3.0, less often with 3.1, and more often with NS 3.2 (Moto). Does anyone have any insight into this problem? Tanks in advance, John Hethcox Evans and Sutherland
Newsgroups: comp.sys.next.bugs From: alex@data.acs.calpoly.edu (Alex Raftis) Subject: Re: Project Builder Message-ID: <1994Feb26.002622.18323@rat.csc.calpoly.edu> Date: Sat, 26 Feb 94 00:26:22 GMT Organization: Cal Poly, SLO References: <2klvm9$gn1@cnn.sim.es.com> In article <2klvm9$gn1@cnn.sim.es.com> jhethcox@homey (John Hethcox) writes: > Sometimes when I edit a file (using Edit), save it, and press the Run > button in PB the app fails to rebuild (it thinks everything is > up-to-date). I had this problem often with NS 3.0, less often with 3.1, > and more often with NS 3.2 (Moto). Does anyone have any insight into this > problem? This can be a problem if you're running over an NFS network. It can happen when the clock on your machine and the machine your files are on get way out of sync. I've never experienced the problem, personally, on a stand alone machine. Alex -- ______________________________________________________ Internet: alex@data.acs.calpoly.edu (NeXT mail) araftis@galaxy.csc.calpoly.edu
From: jcs@alumni.caltech.edu (John C. Stevenson) Newsgroups: comp.sys.next.bugs Subject: White hardware won't tip Date: 26 Feb 1994 02:54:13 GMT Organization: California Institute of Technology, Pasadena Message-ID: <2kmdkl$f3c@gap.cco.caltech.edu> I have routinely "tip"ed my app from black to black hardware. I just installed on white hardware and have transfered (with tip) small files balck to white. When i try to transfer my tarred, compressed and uuencoded app (at 2.6MB) the transfer loses 40 bytes. Needles to say, that is a complete failure. In summary black to black tip of large files works black to white fails. any help out there?? I also tried white to white, but even though i had a tip cub9600 connection, the transfer rate was 2400 baud, to slow to test the large file transfer. any help out there? advTHANKSance jcs@alumni.caltech.edu
Newsgroups: comp.sys.next.bugs From: Robert La Ferla <Robert_La_Ferla@hot.com> Subject: Re: Project Builder Message-ID: <1994Feb26.140151.1266@hot.com> Sender: robertl@hot.com Organization: Hot Technologies References: <2klvm9$gn1@cnn.sim.es.com> Date: Sat, 26 Feb 1994 14:01:51 GMT Did you do a "make depend" to insure that your dependencies were built? That's the likely cause for your problem. PS. People won't be able to reply to your postings because your mail header is missing the domain. Robert La Ferla Hot Technologies NEXTSTEP ISV and Consulting In article <2klvm9$gn1@cnn.sim.es.com> jhethcox@homey (John Hethcox) writes: > Sometimes when I edit a file (using Edit), save it, and press the Run > button in PB the app fails to rebuild (it thinks everything is > up-to-date). I had this problem often with NS 3.0, less often with 3.1, > and more often with NS 3.2 (Moto). Does anyone have any insight into this > problem? > > Tanks in advance, > John Hethcox > Evans and Sutherland
From: gil@atlantic.mps.ohio-state.edu (Gil Rivlis) Newsgroups: comp.sys.next.bugs Subject: Re: Project Builder Date: 26 Feb 1994 16:35:54 GMT Organization: The Ohio State University Distribution: world Message-ID: <2kntpa$nlk@charm.magnus.acs.ohio-state.edu> References: <2klvm9$gn1@cnn.sim.es.com> John Hethcox writes > Sometimes when I edit a file (using Edit), save it, and press the Run > button in PB the app fails to rebuild (it thinks everything is > up-to-date). I had this problem often with NS 3.0, less often with 3.1, > and more often with NS 3.2 (Moto). Does anyone have any insight into this > problem? > I think PB does not check '.h' flies for updates. It checks '.m', '.c', '.psw' etc. Gil P.S. John, your 'From:' and 'Reply-To:' fields in your mailer are bad. --- Gil Rivlis, Physics Department, The Ohio State University gil@atlantic.mps.ohio-state.edu NeXTMail Capable
Newsgroups: comp.sys.next.bugs From: alex@babelfisch.in-berlin.de(Alexander Keiblinger) Subject: Help: WM exits due to excessive DPS errors Message-ID: <CLup7o.AE@babelfisch.in-berlin.de> Sender: root@babelfisch.in-berlin.de (Operator) Organization: me organized? That's a joke! Date: Sat, 26 Feb 1994 21:33:24 GMT Hy, please look at the following logfile and tell me what you think the reason could be. My NSI machine is nearly unusable, it crashs every few minutes. I had the problem before, but a installation from scratch fixed it. But now it is here again. I need to find the exact error now, or I will have to install every two weeks. All I can see is that the WM gets a SIGSEGV, but reinstalling a fresh binary did not fix it, so because of which events the WM gets the memory violation. Any hints what it is, or what I should check ? /usr/adm/messages: Feb 26 21:40:55 babelfisch reboot: Reboot complete Feb 26 21:41:11 babelfisch loginwindow[169]: loginwindow: Checking for DOS partitioned disk Feb 26 21:41:17 babelfisch loginwindow[183]: loginwindow: running /usr/lib/NextStep/Workspace.app/Workspace Feb 26 21:41:17 babelfisch Workspace[183]: logged in Feb 26 21:41:29 babelfisch mach: DOS File System: Initialized Feb 26 22:07:55 babelfisch WindowServer[171]: WindowServer: Exception caught: PS-108.1 on host `babelfisch', user `alex': Memory access exception on address 0x12ce36d8: invalid address (SIGSEGV), Feb 26 22:07:55 babelfisch WindowServer[171]: [ eip = 0x168c6 Feb 26 22:07:55 babelfisch WindowServer[171]: , from 0x15216 Feb 26 22:07:55 babelfisch WindowServer[171]: , from 0x978c Feb 26 22:07:55 babelfisch WindowServer[171]: , from 0x904a Feb 26 22:07:55 babelfisch WindowServer[171]: , from 0xae6c Feb 26 22:07:55 babelfisch WindowServer[171]: , from 0xab2e Feb 26 22:07:55 babelfisch WindowServer[171]: , from 0xa357 Feb 26 22:07:55 babelfisch WindowServer[171]: , from 0x11206b Feb 26 22:07:55 babelfisch WindowServer[171]: , from 0xa592 Feb 26 22:07:55 babelfisch WindowServer[171]: ] Feb 26 22:07:56 babelfisch WM[184]: DPS client library error: Error while writing to connection, DPSContext b6ac0, data -102 Feb 26 22:07:57 babelfisch NewsGrazer[191]: DPS client library error: Error while writing to connection, DPSContext 42c74, data -102 Feb 26 22:07:57 babelfisch NewsGrazer[191]: Exiting due to Window Server death Feb 26 22:07:57 babelfisch Preferences[185]: DPS client library error: Error while writing to connection, DPSContext 4ee30, data -102 Feb 26 22:07:57 babelfisch Preferences[185]: Exiting due to Window Server death Feb 26 22:07:56 babelfisch BackSpace[186]: DPS client library error: Error while writing to connection, DPSContext 40df0, data -102 Feb 26 22:07:58 babelfisch BackSpace[186]: Exiting due to Window Server death Feb 26 22:08:12 babelfisch loginwindow[169]: loginwindow: Workspace exited ts 0 cd 0 rc 0 sv 0 ss 0. Feb 26 22:08:13 babelfisch loginwindow[169]: dps_err_write Feb 26 22:08:13 babelfisch last message repeated 20 times Feb 26 22:08:13 babelfisch loginwindow[169]: exiting due to excessive DPS errors Feb 26 22:08:18 babelfisch loginwindow[293]: loginwindow: sound service already registered. Feb 26 22:08:18 babelfisch loginwindow[293]: loginwindow: _NXAudioIn service already registered. Feb 26 22:08:18 babelfisch loginwindow[293]: loginwindow: _NXAudioOut service already registered. Feb 26 22:08:22 babelfisch loginwindow[293]: loginwindow: Checking for DOS partitioned disk Feb 26 22:08:45 babelfisch loginwindow[298]: loginwindow: running /usr/lib/NextStep/Workspace.app/Workspace Feb 26 22:08:46 babelfisch Workspace[298]: logged in
From: klui@corp.hp.com (Ken Lui) Newsgroups: comp.sys.next.bugs Subject: Re: Help: WM exits due to excessive DPS errors Date: 27 Feb 1994 18:21:19 GMT Organization: Hewlett-Packard Company Message-ID: <2kqoav$n4n@hpscit.sc.hp.com> References: <CLup7o.AE@babelfisch.in-berlin.de> In article <CLup7o.AE@babelfisch.in-berlin.de>, Alexander Keiblinger <alex@babelfisch.in-berlin.de> wrote: >Hy, > >please look at the following logfile and tell me what you think the reason >could be. My NSI machine is nearly unusable, it crashs every few minutes. [...] >Feb 26 22:07:56 babelfisch WM[184]: DPS client library error: Error while >writing to connection, DPSContext b6ac0, data -102 I get this kind of error, too sometimes whenever I copy files from a DOS-formatted diskette to my home directory in the Workspace. It happens after the file has been copied and I have the Process window open--telling me how far along the process is. I've stopped using the Workspace to copy stuff and just use a shell. Ken -- Kenneth K.F. Lui, klui@corp.hp.com 3000 Hanover Street M/S 20BJ Corporate Financial Systems Palo Alto, CA 94304-1112 USA Core Application Technologies 1.415.857.3230 Fax 1.415.852.8026
Newsgroups: comp.sys.next.bugs From: Nitezki@NiDat.sub.org (Peter Nitezki) Subject: Re: AdaptorOracle uses tables not included in the default V7 oracle catalog Message-ID: <1994Feb27.144146.18675@nidat.sub.org> Sender: nitezki@nidat.sub.org Organization: private site of Peter Nitezki, Kraichtal, Germany References: <2kaq04$j16@obelix.WHU-Koblenz.de> Date: Sun, 27 Feb 1994 14:41:46 GMT In article <2kaq04$j16@obelix.WHU-Koblenz.de> wolfgang@wi.WHU-Koblenz.de (Wolfgang Roeckelein) writes: > Hello everybody, > > perhaps this is of interest to some: > > NeXTstep 3.2 AdaptorOracle uses system tables which are not included in > the default V7 oracle catalog (esp. accessible_tables etc.) > > Therefore DBModeler can't build a default model! > > As a temporary solution you can advice DB-Admins to execute catalog6.sql > on their V7 server!! > > A better solution would be for NeXT to use the newer system tables! > Just in case this not already known to the other Netters (except Wolfgang, of cause ;-) Neither NeXT nor Oracle ever promised the DBAdaptor-Kit to be compatible with Oracle 7. It is still bound to version 6! I need not mention that we all would like to see this changing soon. But on the other hand who would be in a position to include the concepts of triggers, stored procedures and declarative referencial integrity into DBKit? PL-SQL (a ADA derivative) is too far from Obj-C to be an easy couple (just my two cents). -- Peter Nitezki | Nitezki@NiDat.sub.org # Blessed art thou who knoweth Staarenbergstr. 44 | Tel.: +49 7251 62495 # not about the pleasure and D-76703 Kraichtal | Fax : +49 7251 69215 # delight of being hooked GERMANY | NeXTmail welcome !!! # up to the Net. Peter 1,3-5
Newsgroups: comp.sys.next.bugs From: Nitezki@NiDat.sub.org (Peter Nitezki) Subject: Re: Fax cover sheet forgetfulness Message-ID: <1994Feb27.144810.18734@nidat.sub.org> Sender: nitezki@nidat.sub.org Organization: private site of Peter Nitezki, Kraichtal, Germany References: <CLpK66.LF5@crash.cts.com> Date: Sun, 27 Feb 1994 14:48:10 GMT In article <CLpK66.LF5@crash.cts.com> coconut@crash.cts.com (Brian Dear) writes: > > Is it a known bug that when you "Preview" a Fax cover sheet in NS3.2, > that is, you Print something, then select Fax instead of printing it, > and then create a cover sheet (click Yes and Edit) and then you > go back to the Fax dialog and hit Preview, that once you're done > previewing the cover sheet, it's lost and you gotta type it again? > > Seems to be several little quirks like this with the built-in fax > stuff of NS3.2. Is this known already? (Acknowledged by NeXT as a bug? > Will it ever get fixed?) > I'm not ready to call this a bug! It is an oddity that comes from the implementation of FAXing as a sort of printing. The cover sheet is therefore no part of the document and can't be previewed like other documents. I shunn two other consequences of this behaviour: 1. you can't sen a cover sheet only, and 2. you can't append multiple documents without the use of special application software. -- Peter Nitezki | Nitezki@NiDat.sub.org # Blessed art thou who knoweth Staarenbergstr. 44 | Tel.: +49 7251 62495 # not about the pleasure and D-76703 Kraichtal | Fax : +49 7251 69215 # delight of being hooked GERMANY | NeXTmail welcome !!! # up to the Net. Peter 1,3-5
From: bihr@next.informatik.hu-berlin.de (Tom Bihr) Newsgroups: comp.sys.next.bugs Subject: Re: Project Builder Date: 28 Feb 1994 10:08:02 GMT Organization: Humboldt University Berlin, Department of Computer Science Distribution: world Message-ID: <2ksfq2$aoh@hahn.informatik.hu-berlin.de> References: <2kntpa$nlk@charm.magnus.acs.ohio-state.edu> In article <2kntpa$nlk@charm.magnus.acs.ohio-state.edu> gil@atlantic.mps.ohio-state.edu (Gil Rivlis) writes: > John Hethcox writes > > Sometimes when I edit a file (using Edit), save it, and press the Run > > button in PB the app fails to rebuild (it thinks everything is > > up-to-date). I had this problem often with NS 3.0, less often with 3.1, > > and more often with NS 3.2 (Moto). Does anyone have any insight into > this > > problem? > > > > I think PB does not check '.h' flies for updates. It checks '.m', > '.c', '.psw' etc. > I confirm: PB does not check .h-files for changes. Even if you try do "make depend", as Robert la Ferla in <1994Feb26.140151.1266@hot.com> recommends, PB will refuse to do the right thing. So if you want to rebuild after changing a h-file you have to touch your source files. Tom --- Tom Bihr Humboldt-Universitaet zu Berlin Fachbereich 16 - Informatik Unter den Linden 6 - 10099 Berlin Telefon: +49 30 20181306 E-Mail (NeXT-Mail willkommen): bihr@informatik.hu-berlin.de ___ _ | | ) Let the good times Roll... | |_)
Newsgroups: comp.sys.next.bugs From: dimitri@dutiws.twi.tudelft.nl (Dimitri Tischenko) Subject: make depend does work! (Was: Re: Project Builder ) Message-ID: <CLxrC1.Jn8@dutiws.twi.tudelft.nl> Organization: Delft University of Technology References: <2kntpa$nlk@charm.magnus.acs.ohio-state.edu> <2ksfq2$aoh@hahn.informatik.hu-berlin.de> Date: Mon, 28 Feb 1994 13:12:01 GMT In article <2ksfq2$aoh@hahn.informatik.hu-berlin.de>, Tom Bihr wrote: ... >I confirm: PB does not check .h-files for changes. Even if you try do "make >depend", as Robert la Ferla in <1994Feb26.140151.1266@hot.com> recommends, >PB will refuse to do the right thing. So if you want to rebuild after >changing a h-file you have to touch your source files. This is certainly not the case. First of all, PB doesn't have anything to do with dependency check. If you "build depend", a "make depend" command is issued by PB. The make process calls the C compiler with the -MM option and redirects the output to the file Makefile.dependencies. This file lists all dependencies between .h and .o files. This file is then included by the standard Makefile. So if you didn't change your Makefile, everything should work as stated. Dimitri > >Tom >--- > Tom Bihr Humboldt-Universitaer zu Berlin Fachbereich 16 - Informatik > Unter den Linden 6 - 10099 Berlin > Telefon: +49 30 20181306 > > E-Mail (NeXT-Mail willkommen): bihr@informatik.hu-berlin.de ___ _ > | | ) > Let the good times Roll... | |_) -- +----------------------------------------------------------------------------+ | Dimitri Tischenko | D.B.Tischenko@TWI.TUDelft.NL | NeXTmail preferred! | +---------------------------------------+------------------------------------+ | Delft University of Technology | Technische Universiteit Delft | | Fac Applied Math & Computer Science | Fac. Techn. Wiskunde & Informatica | | The Netherlands | Nederland | +---------------------------------------+------------------------------------+
Newsgroups: comp.sys.next.bugs From: mark@xexos.com (Mark Chamberlain) Subject: Re: Fax cover sheet forgetfulness Message-ID: <1994Feb28.143417.15079@xexos.com> Sender: news@xexos.com Organization: Xexos, Ltd (London) References: <1994Feb27.144810.18734@nidat.sub.org> Date: Mon, 28 Feb 1994 14:34:17 GMT In article <1994Feb27.144810.18734@nidat.sub.org> Nitezki@NiDat.sub.org (Peter Nitezki) writes: > I shunn two other consequences of this behaviour: > 1. you can't sen a cover sheet only, and > 2. you can't append multiple documents without the use of special > application software. I solved 1. I wrote a dead simple program that had a blank view, printed it, and then editted the PS output on the way to the print daemon to remove the second blank page - all fine and dandy, I now have a program that does one-page faxes using the cover sheet - except that it broke under 3.2, and I haven't had the time to find out why. For 2. I would really like to be able to select say 4 ps files from the Workspace and then have a Services entry that says "Fax to...", up pops the Fax panel and away I go. Do any of the commercial products do this? -- Mark Chamberlain +44 71 237 4535 Xexos Ltd fax +44 71 231 0844 London mark@xexos.com
From: sanguish@digifix.com Newsgroups: comp.sys.next.announce,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.bugs,comp.sys.next.programmer,comp.sys.next.marketplace Subject: Quick Guide to NEXTSTEP information on the Internet Date: 28 Feb 1994 01:34:54 -0500 Organization: Next Announcements Message-ID: <2ks3ae$71j@digifix.digifix.com> This post is made weekly, to help 'point' users to more NEXTSTEP information Topics include: comp.sys.next newsgroups related newsgroups comp.sys.next newsgroups mailing list ftp sites NeXTanswers comp.sys.next.* newsgroups -------------------------- Comp.Sys.Next.Advocacy This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. Comp.Sys.Next.Announce Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) The NEXTSTEP FAQs are posted here monthly as well. This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Comp.Sys.Next.Bugs A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT- specific groups as well. Comp.Sys.Next.Hardware Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. Comp.Sys.Next.Marketplace NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. Comp.Sys.Next.Misc For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! Comp.Sys.Next.Programmer Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. The NEXTSTEP programmer FAQs are posted here. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. Comp.Sys.Next.Software This is a place to talk about [third party] software products that run on NEXTSTEP systems. Comp.Sys.Next.Sysadmin Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. related Newsgroups ------------------ Comp.Soft-Sys.Nextstep Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. Comp.Lang.Objective-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. Comp.Object Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original Comp.Sys.Next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News ------------------------------------------- Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep-d next-advocacy-d next-announce-d next-bugs-d next-hardware-d next-marketplace-d next-misc-d next-programmer-d next-software-d next-sysadmin-d (For a full description, send mail saying LISTS to <digestif@antigone.com>). The subscription syntax is essentially the same as LISTSERV's. To subscribe, send a message to <digestif@antigone.com> saying: SUB Listname YourName Example: SUB next-hardware-d John Doe The ftp sites ------------- cs.orst.edu: The main site for North American submissions nova.cc.purdue.edu: Lots of older stuff, but very short on disk space ftp.informatik.uni-muenchen.de: In Germany. terra.stack.urc.tue.nl (Dutch NEXTSTEP User Group) and cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp.next.com: See the below ftp.next.com and NextAnswers@next.com ------------------------------------- From the document 1000_Help from ftp.next.com Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, or you can transfer them by anonymous ftp. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to NeXTanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO <your-address> If you have any problem with the system or suggestions for improvement, please send mail to NeXTanswers-request@NeXT.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to NeXTanswers-request@NeXT.com. Written by: Eric P. Scott eps@toaster.SFSU.EDU and Scott Anguish sanguish@digifix.com Additions from: Greg Anderson (Greg_Anderson@afs.com) and Michael Pizolato (Michael_Pizolato@afs.com)
From: Hal.Varian@umich.edu Newsgroups: comp.sys.next.bugs Subject: Re: Fax cover sheet forgetfulness Date: 28 Feb 1994 22:53:08 GMT Organization: University of Michigan - College of Literature, Science, and TheArts Distribution: world Message-ID: <2ktskl$jp3@controversy.math.lsa.umich.edu> References: <1994Feb28.143417.15079@xexos.com> In article <1994Feb28.143417.15079@xexos.com> mark@xexos.com (Mark Chamberlain) writes: > For 2. I would really like to be able to select say 4 ps files from the > Workspace and then have a Services entry that says "Fax to...", up > pops the Fax > panel and away I go. Do any of the commercial products do this? This isn't quite what you want, but maybe it will help: /* program for putting PS files into one PS file * * Copyright National Science Foundation 1988 * * Written June 1988 by Chao-Ping Yang */ /* Usage: collate file1.ps file2.ps ... > result.ps or collate -o result.ps file1.ps file2.ps (Note the -o switch is stronger so that collate -o result.ps file.... >another.ps will have the result in result.ps and another.ps will be empty. ) there is also a -R switch to stop reversal of the files . */ # include <stdio.h> # include "common.h" #ifdef Macintosh #undef stdout #endif # define _header "%!\n% Proposal for the National Science Foundation" # define PS_BEGIN "/NSF-proposal-NSF save def /NSF-count-NSF countdictstack def" # define PS_END "clear NSF-count-NSF 1 countdictstack 1 sub { pop end } for NSF-proposal-NSF restore" COLLATEPS(argc,argv) /* this is the main program on non Mac computers */ int argc; char *argv[]; { FILE *output,*input,*fopen(); char *s; int c, reverse = 1; /* default putting cover page at the end * of the command line */ #ifdef Macintosh extern FILE *stdout; #endif output = stdout; /* need redirection */ while (--argc > 0 && (*++argv)[0] =='-') for ( s = argv[0]+1; *s != '\0'; s++) switch (*s) { case 'R':case 'r': reverse = 0; break; case 'O': case 'o': if( ( output = fopen(*++argv,"w") ) == NULL ) { print_stderr("\n## Can not create %s.\n",*argv); finish(0); } argc--; break; default: print_stderr("Illegal option: %c\n",*s); argc = 0; break; } if( argc < 1 ) { print_stderr( "Usage: collate [-R] [-o result.ps] file1.ps file2.ps ... [> result.ps]\n",0); print_stderr( "where the R switch will stop putting files in reverse.\n",0); print_stderr( "and the o switch set the file that followed as the output file.\n",0); finish(0); } fprintf(output,"%s\n",_header); /* PS header for the collated file */ /************* Start copying PS files ************************/ if(reverse) argv += argc - 1; /* move to the end of the file names */ while ( --argc >= 0) { #ifdef Macintosh /* So it won't cause problem under VMS */ print_stderr("%s ",*argv); #endif if( (input=fopen(*argv,"r") ) == NULL) { #ifdef Macintosh fclose(output); #endif print_stderr("\n## %s: File not found.\n",*argv); finish(1); } fprintf(output,"\n%s\n",PS_BEGIN); while ( ( c = getc(input) ) != EOF ) putc(c,output); fprintf(output,"\n%s\n",PS_END); if(reverse) argv --; else argv ++; fclose(input); } #ifdef Macintosh print_stderr("\nAll done!\n",0); fclose(output); #endif finish(0); } -- Hal.Varian@umich.edu Hal Varian voice: 313-764-2364 Dept of Economics fax: 313-764-2364 Univ of Michigan Ann Arbor, MI 48109-1220
Newsgroups: comp.sys.next.bugs From: bbum@friday.com Subject: Re: Fax cover sheet forgetfulness Message-ID: <CLxtMp.G9@friday.com> Sender: bbum@friday.com (Bill Bumgarner) Organization: Friday Software & Consulting; a virtual company References: <1994Feb27.144810.18734@nidat.sub.org> Date: Mon, 28 Feb 1994 14:01:36 GMT In a related note, it seems that my machine(s) have lost the ability to produce cover sheets at all... anyone have a clue as to why this would happen? b.bum Peter Nitezki writes > In article <CLpK66.LF5@crash.cts.com> coconut@crash.cts.com (Brian Dear) > writes: > > > > Is it a known bug that when you "Preview" a Fax cover sheet in NS3.2, > > that is, you Print something, then select Fax instead of printing it, > > and then create a cover sheet (click Yes and Edit) and then you > > go back to the Fax dialog and hit Preview, that once you're done > > previewing the cover sheet, it's lost and you gotta type it again? > > > > Seems to be several little quirks like this with the built-in fax > > stuff of NS3.2. Is this known already? (Acknowledged by NeXT as a bug? > > Will it ever get fixed?) > > > I'm not ready to call this a bug! It is an oddity that comes from the > implementation of FAXing as a sort of printing. The cover sheet is > therefore no part of the document and can't be previewed like other > documents. > > I shunn two other consequences of this behaviour: > 1. you can't sen a cover sheet only, and > 2. you can't append multiple documents without the use of special > application software. > -- > Peter Nitezki | Nitezki@NiDat.sub.org # Blessed art thou who knoweth > Staarenbergstr. 44 | Tel.: +49 7251 62495 # not about the pleasure and > D-76703 Kraichtal | Fax : +49 7251 69215 # delight of being hooked > GERMANY | NeXTmail welcome !!! # up to the Net. Peter 1,3-5 -- b.bum <bbum@friday.com> | "blah blah blah" WhiteLight Systems, INC. | - someone somewhere
Newsgroups: comp.sys.next.bugs From: francisr@stupid.ucs.indiana.edu (Rob Francis) Subject: Re: Help: WM exits due to excessive DPS errors Message-ID: <CLyvCp.AEq@usenet.ucs.indiana.edu> Sender: news@usenet.ucs.indiana.edu (USENET News System) Organization: Indiana University References: <CLup7o.AE@babelfisch.in-berlin.de> <2kqoav$n4n@hpscit.sc.hp.com> Date: Tue, 1 Mar 1994 03:36:24 GMT In article <2kqoav$n4n@hpscit.sc.hp.com>, Ken Lui <klui@corp.hp.com> wrote: >In article <CLup7o.AE@babelfisch.in-berlin.de>, >Alexander Keiblinger <alex@babelfisch.in-berlin.de> wrote: >>Hy, >> >>please look at the following logfile and tell me what you think the reason >>could be. My NSI machine is nearly unusable, it crashs every few minutes. >[...] >>Feb 26 22:07:56 babelfisch WM[184]: DPS client library error: Error while >>writing to connection, DPSContext b6ac0, data -102 > >I get this kind of error, too sometimes whenever I copy files >from a DOS-formatted diskette to my home directory in the >Workspace. It happens after the file has been copied and I have >the Process window open--telling me how far along the process is. >I've stopped using the Workspace to copy stuff and just use a >shell. Uh uh. I think we've got a nasty bug. Several of my NetInfo clients wouldn't boot today. They'd get as far as trying to start the window server, then I would get the same error message you folks describe. I looked on their netinfo master, and it had a couple of problems, all of them about not being able to bind to ports. Here is one of them: Feb 28 18:05:49 foo syslog: NXMiniApp: Cannot get bootstrap port :1102 I rebooted the server, and everything seems okay for the time being. I'm going to write this up and send it to bug_next. -rob --- Rob Francis Unix Systems Specialist, University Computing Services Indiana University, francisr@indiana.edu (812)855-0327
From: Charlesa@learned.co.uk Newsgroups: comp.sys.next.software,comp.sys.next.programmer,comp.sys.next.misc,comp.sys.next.bugs Subject: Backspace bug? Date: Tue, 01 Mar 94 17:35:08 BST Organization: EUnet GB Message-ID: <2kvue7$pgg@marble.Britain.EU.net> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII I can't get the BackSpace modules to complie correctly with 3.2. The make install keeps looking for a file which doesn't exist. Anybody else run into this? Charles
From: paterson@gdss.commerce.ubc.ca (Toby Paterson) Newsgroups: comp.sys.next.bugs Subject: Re: Project Builder Date: 1 Mar 1994 20:51:02 GMT Organization: The University of British Columbia Message-ID: <2l09rm$ha0@nntp.ucs.ubc.ca> References: <2klvm9$gn1@cnn.sim.es.com> In article <2klvm9$gn1@cnn.sim.es.com> jhethcox@homey (John Hethcox) writes: > Sometimes when I edit a file (using Edit), save it, and press the Run > button in PB the app fails to rebuild (it thinks everything is > up-to-date). I had this problem often with NS 3.0, less often with 3.1, > and more often with NS 3.2 (Moto). Does anyone have any insight into this > problem? > We occasionally get this problem also. We figured out that it has something to do with the file server's system clock being out of sync (by sometimes as much as a couple of minutes) with my local machine's system clock. It was a long time ago, and I can't remember the details. X -- \\ / Who: Toby Paterson \\/ How: paterson@gdss.commerce.ubc.ca //\ What: Grunt and NeXT hacker; GDSS Fellowship // \ Where: University of British Columbia
From: finton@homer.cs.wisc.edu (David Finton) Newsgroups: comp.sys.next.bugs Subject: Possible bug in RTF object of Simson & Mahoney? Date: 1 Mar 1994 22:54:02 GMT Organization: U of Wisconsin CS Dept Message-ID: <2l0h2a$h1a@spool.cs.wisc.edu> I'm using the RTF object from Simson & Mahoney's book. I've added a method which is based on their `bold' method, and it works fine in NS 2.0, but not in NS 3.x. So I tested their `bold' method, and it also fails in the same way for NS3.x. I wonder if this is a known bug, if there's a fix, or if I'm just doing something goofy. Here's how I use RTF: [textDoc setSel:length :length]; rtf = [ [RTF allocFromZone:[self zone]] init ]; [rtf bold:YES]; [rtf append:buf]; [rtf bold:NO]; [textDoc replaceSelWithRichText:[rtf stream] ]; [rtf free]; This allows me to add a single character with a different font (say, bold) to the text object, and then proceed normally. But in NS3.x the font never changes back to normal. That is, after I've added a single bold character to the text object with this code, *all* the remaining characters added are bold! But everything works perfectly in NS2.0. The method I'm adding uses a similar scheme to do half-shading instead of bold on selected characters; it's crucial for showing which data is included. So we can't upgrade unless we can solve this problem. By the way, all my testing has been on black hardware. Any ideas? --David Finton
#################################################################### Newsgroups: comp.sys.next.bugs From: bbum@friday.com Subject: Re: make depend does work! (Was: Re: Project Builder ) Message-ID: <CM046M.vw@friday.com> Sender: bbum@friday.com (Bill Bumgarner) Organization: Friday Software & Consulting; a virtual company References: <CLxrC1.Jn8@dutiws.twi.tudelft.nl> Date: Tue, 1 Mar 1994 19:44:45 GMT Dimitri Tischenko writes > This is certainly not the case. First of all, PB doesn't have anything > to do with dependency check. If you "build depend", a "make depend" > command is issued by PB. The make process calls the C compiler with > the -MM option and redirects the output to the file Makefile.dependencies. > This file lists all dependencies between .h and .o files. This file is > then included by the standard Makefile. So if you didn't change your > Makefile, everything should work as stated. Unfortunately, this has the nasty side-effect of recompiling just about everything in the project if you make a change to a header file AND are using precompiled (or globbed) headers. -- b.bum <bbum@friday.com> | "blah blah blah" WhiteLight Systems, INC. | - someone somewhere
From: wolfgang@wi.WHU-Koblenz.de (Wolfgang Roeckelein) Newsgroups: comp.sys.next.bugs Subject: Re: AdaptorOracle uses tables not included in the default V7 oracle catalog Date: 2 Mar 1994 18:54:46 GMT Organization: WHU Koblenz Message-ID: <2l2ndm$c6c@obelix.WHU-Koblenz.de> References: <1994Feb27.144146.18675@nidat.sub.org> In article <1994Feb27.144146.18675@nidat.sub.org> Nitezki@NiDat.sub.org (Peter Nitezki) writes: > In article <2kaq04$j16@obelix.WHU-Koblenz.de> wolfgang@wi.WHU-Koblenz.de > (Wolfgang Roeckelein) writes: > Neither NeXT nor Oracle ever promised the DBAdaptor-Kit to be compatible > with Oracle 7. It is still bound to version 6! this is absolutly incorrect. Look at the document 1359_DBKit_Adaptor_Availability_Guide.ps from NeXTanswers. And the Adpator from 3.2 _runs_ with V7 except for the mentioned trouble. So please watch what you are saying!!!!!!!!!!!!!! Greetings, Wolfgang Roeckelein -- Dipl.-Wirtsch.-Inf. Voice: +49 261 6509 173 Wolfgang Roeckelein Fax: +49 261 6509 179 WHU Koblenz E-Mail: roeckelein@wi.whu-koblenz.de Burgplatz 2 (NeXTmail ok) D-56179 Vallendar Germany
Newsgroups: comp.sys.next.bugs From: tmoore@nightingale.con.utk.edu (Terry Moore) Subject: re: NS 3.2 on IBM ValuePoint's IDE Message-ID: <1994Mar2.181809.22950@martha.utcc.utk.edu> Sender: usenet@martha.utcc.utk.edu (USENET News System) Organization: University of Tennessee Computing Center Date: Wed, 2 Mar 1994 18:18:09 GMT Apparently there is a known bug which keeps NS from running with the IDE controllers that IBM uses on its Value Points. My question is: Is there a fix for this in the works, and if so when? The answer to this question will affect my ability to convince my organization to make the jump to NS, and a decision on will be made on this very soon. It's frustrating as hell to have the path to a clearly superior OS platform blocked by such a problem. Terry Moore tmoore@nightingale.con.utk.edu
Newsgroups: comp.sys.next.bugs From: bruce@TotSysSoft.com (Bruce Gingery) Subject: Re: Project Builder Message-ID: <1994Mar3.064754.1264@ToTSySSoft.com> Sender: news@ToTSySSoft.com Organization: Total System Software References: <2ksfq2$aoh@hahn.informatik.hu-berlin.de> Date: Thu, 3 Mar 1994 06:47:54 GMT In article <2ksfq2$aoh@hahn.informatik.hu-berlin.de> bihr@next.informatik.hu-berlin.de (Tom Bihr) writes: -> In article <2kntpa$nlk@charm.magnus.acs.ohio-state.edu> -> gil@atlantic.mps.ohio-state.edu (Gil Rivlis) writes: -> > John Hethcox writes -> > > Sometimes when I edit a file (using Edit), save it, -> > > and press the Run button in PB the app fails to -> > > rebuild (it thinks everything is up-to-date). I had -> > > this problem often with NS 3.0, less often with 3.1, -> > > and more often with NS 3.2 (Moto). Does anyone have -> > > any insight into this problem? -> > > -> > -> > I think PB does not check '.h' flies for updates. It -> > checks '.m', '.c', '.psw' etc. -> > -> I confirm: PB does not check .h-files for changes. Even -> if you try do "make depend", as Robert la Ferla in -> <1994Feb26.140151.1266@hot.com> recommends, PB -> will refuse to do the right thing. So if you want -> to rebuild after changing a h-file you have to -> touch your source files. -> .. OR do a Make-Clean from PB. I've also encountered this with source files modified (corrected) during the Make. The source stamp date ends up later than the .o files, but not later than the executable. Hence my in-makefile revision annotations and "what" lines in the sources which are generated into executables from them. Bruce Gingery bruce@TotSysSoft.com
From: robert@steffi.demon.co.uk (Robert Nicholson) Newsgroups: comp.sys.next.misc,comp.sys.next.bugs Subject: Fax Software question. Date: 3 Mar 1994 13:57:19 -0000 Organization: me organized? That's a joke! Message-ID: <2l4qbv$1n2@steffi.demon.co.uk> I run NXFax with a 1496E but this is probably a NeXT FAX software thing. Anyway, in the options for the the NXFAX driver you can give it a phone number. Well I assume that's for a reason? I sent a fax earlier today and it didn't send that number to their fax modem at the start of transmission. What's that number for? IMHO: If the software _does_ show that number then Preview should also show that little header at the top of the FAX too. -- "You know what's wrong with you?" (Audrey Hepburn, Cary Grant) "No, what?" "Nothing" (Charade, 1963) (ASCII for text only messages)
From: silvio@jabberwock.swarthmore.edu (Silvio P Eberhardt) Newsgroups: comp.sys.next.bugs Subject: NXStrHash problem? Date: 3 Mar 1994 22:29:55 GMT Organization: Swarthmore College, Swarthmore, PA, USA Message-ID: <2l5od3$bt3@larch.cc.swarthmore.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Keywords: NXStrHash An App that worked just fine under 3.0 won't even get off the ground when recompined under (newly installed) 3.2. Has anyone experienced the following phenotype? ----------- When trying to run the program, I get the following message using gdb: Program generated(1): Memory access exception on address 0x31 (protection failure). 0x5009c6a in NXStrHash () Here are some of the surrounding addresses: #0 0x5009c6a in NXStrHash () #1 0x50091ba in NXHashGet () #2 0x500c9ec in NXUniqueString () #3 0x603b6f0 in _NXMakeAtomList () #4 0x603e366 in -[View(Drag) registerForDraggedTypes:count:] () Reading in symbols for DragReceiveView.m...done. #5 0x4100 in -[DragReceiveView registerTypes] (self=0x10b93c, _cmd=0xb6fc) at DragReceiveView.m:15 #6 0x4144 in -[DragReceiveView initFrame:] (self=0x10b93c, _cmd=0x618ad05, frameRect=0x1079cc) at DragReceiveView.m:23 #7 0x604be16 in -[CustomView nibInstantiate] () Any help would be greatly appreciated.
Newsgroups: comp.sys.next.bugs From: paul@seer.demon.co.uk (Paul Lynch) Subject: Re: Fax Software question. Message-ID: <1994Mar3.221238.4140@seer.demon.co.uk> Sender: paul@seer.demon.co.uk Organization: P & L Systems References: <2l4qbv$1n2@steffi.demon.co.uk> Date: Thu, 3 Mar 1994 22:12:38 GMT In article <2l4qbv$1n2@steffi.demon.co.uk> robert@steffi.demon.co.uk (Robert Nicholson) writes: > I run NXFax with a 1496E but this is probably a NeXT FAX software thing. > > Anyway, in the options for the the NXFAX driver you can give it a > phone number. Well I assume that's for a reason? Yes, it's the little number that appears at the top of a fax transmission. > I sent a fax earlier today and it didn't send that number to their fax > modem at the start of transmission. What's that number for? It does normally. > IMHO: If the software _does_ show that number then Preview should also > show that little header at the top of the FAX too. That, I agree with. Paul -- Paul Lynch P & L Systems (NeXTmail) paul@seer.demon.co.uk Tel: (0494)671501 9 Stable Lane, Seer Green, Fax: (0494)680228 Bucks, HP9 2YT, UK
Newsgroups: comp.sys.next.bugs Subject: Re: Fax Software question. Message-ID: <1994Mar4.075543.29640@urz.unibas.ch> From: frank@ifi.unibas.ch (Robert Frank) Date: Fri, 4 Mar 1994 07:55:43 GMT Sender: news@urz.unibas.ch (USENET News System) References: <2l4qbv$1n2@steffi.demon.co.uk> Organization: Institut fuer Informatik In article <2l4qbv$1n2@steffi.demon.co.uk> writes: > I run NXFax with a 1496E but this is probably a NeXT FAX software thing. > > Anyway, in the options for the the NXFAX driver you can give it a > phone number. Well I assume that's for a reason? So that the receiver can identifiy you. It would be even nicer if you could also supply a short string as is usual on 'real' fax devices. > I sent a fax earlier today and it didn't send that number to their fax > modem at the start of transmission. What's that number for? This has happend to me too. I'm not sure what caused this to happen and I haven't tested it with other devices yet. > IMHO: If the software _does_ show that number then Preview should also > show that little header at the top of the FAX too. No, it shouldn't. The little header is generated by the receiving device, so the sending software can't say what will happen. It could be that the receiving device adds a lot of other things to the header, or that it leaves it out altogether. -- Institut fuer Informatik tel +41 (0)61 321 99 67 Universitaet Basel fax. +41 (0)61 321 99 15 Robert Frank Mittlere Strasse 142 rfc822: frank@ifi.unibas.ch (NeXT mail ok) CH-4056 Basel X400: S=frank;OU=ifi;O=unibas;P=switch;A=arcom;C=ch Switzerland
From: mow@marsu.tynet.sub.org (Markus Wenzel) Newsgroups: comp.sys.next.sysadmin,comp.sys.next.hardware,comp.sys.next.bugs Subject: EATA timeout report #1 (summary) Date: 4 Mar 1994 13:08:06 GMT Organization: Palumbian Research Labs Message-ID: <2l7brm$1qo@marsu.tynet.sub.org> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Keywords: DPT 2022, SCSI timout, crashes EATA timeout report =================== Description of the problem: The DPT EISA controllers 2022/9x and 2012/9x cause disastrous crahes under NeXTSTEP Intel 3.x. These crashes are initiated by an "EATA timeout" error message in /usr/adm/messages, and a few seconds later any disk access on the SCSI bus is blocked. If this happens when file system relevant information is written to disk, the resulting damage to the filesystem is often so big that a complete restore from tape is the only way left. A typical /usr/adm/messages entry looks like: Feb 14 04:06:57 kato mach: EATA timeout Feb 14 04:06:57 kato mach: Resetting SCSI Bus... Feb 14 04:06:57 kato mach: sd1: I/O Timeout; Retrying. Feb 14 04:06:57 kato mach: target:2 lun:0 op:Read block:1458928 blockCount:16 Feb 14 04:06:57 kato mach: EATA timeout Feb 14 04:06:57 kato mach: Resetting SCSI Bus... Feb 14 04:06:57 kato mach: sd0: I/O Timeout; Retrying. Feb 14 04:06:57 kato mach: target:0 lun:0 op:Read block:303648 blockCount:6 Feb 14 04:06:57 kato mach: sd1: I/O Timeout; Retrying. Feb 14 04:06:57 kato mach: target:2 lun:0 op:Read block:1898816 blockCount:16 Feb 14 04:07:27 kato mach: EATA timeout Feb 14 04:07:27 kato mach: Resetting SCSI Bus... Feb 14 04:07:27 kato mach: sd1: I/O Timeout; Retrying. Feb 14 04:07:27 kato mach: target:2 lun:0 op:Write block:1458928 blockCount:1 6 Equivalent crashes were reported by: Rene Kulschewski <rene@rkt.in-berlin.de> martin@cube.han.de (Martin Ortlepp) Rainer Staringer <rainer@fml.tuwien.ac.at> John B. Hampshire II <hamps@richibucto.jpl.nasa.gov> ramesh@bodhi.esys.cwru.edu (Ramesh Dodamani) Varun Mitroo <mitroo@magnus.acs.ohio-state.edu> Milo Velimirovic <milov@fingers.acs.uwlax.edu> Markus Wenzel <mow@marsu.tynet.sub.org> Additional notes: The crash frequency varies from twice a day up to once a month. It is reported that the crashes occur with SmartROM rev 2B and 2C / BIOS rev 5B and 5C. The only characteristic which is common to all listed systems might be that in all of them Fast-SCSI-II harddisks are installed, e.g. DEC DSP3107, DEC DSP 3053, HP C2247, IBM 0662S12 Weird effects I noticed on my system: - No timeouts occur any longer if I replace the 2022 with a ISA 2021 adapter. - The same 2022 in a friend's system with an old and rather slow Fujitsu drive does not cause any trouble there. Effects noticed on more than one system: - Decreasing the transfer rate from 10 MB/s to 5 MB/s makes the timeouts less frequent, but they still occur. Performance goes down with it seriously, so this cannot be considered as a "solution". My ECU configuration is exactly as described in NeXTanswers #1457 and I have no cache installed on my DPT. (I cannot speak for the settings of the other systems.) Bad cabling and termination cannot be the reason for all eight systems listed above. Personal experiments with termination (switching disk order in the chain, leaving other devices out) did not cure the problem. System configurations in detail: my own: EISA/VL 486DX2 66, SiS Chipset, AMI BIOS, 32 MB DPT 2022/95, BIOS Rev. 2C (purchased in January 94) Toshiba CD-ROM 3201, Sankyo CP 525 QIC, DEC DSP 3053L, IBM 0662S12 (with active termination on). MTBC: about 2 days. Milo Velimirovic <milov@fingers.acs.uwlax.edu> Dell 4066/XE (i486 66MHz 24Mb) S3 805 w1MB Video RAM NeXTSTEP 3.2 DPT 2012/90, BIOS Rev. 5B, purchased summer 1993 HP 4mm DAT SCSI ID 6. DEC DSP3105 1GB disk on / SCSI ID 0. 3 HP C2247 1GB disks mounted for users' files, SCSI ID's 1,2, and 4. Dell 4066/XE (i486 66MHz 32Mb) S3 805 w1MB Video RAM NeXTSTEP 3.1 DPT 2012/90, BIOS Rev. 5B, purchased late summer 1993 3 HP C2247 1GB disks mounted for users' files, SCSI ID's 0,2 and 4. Dell 4066/XE (i486 66MHz 32Mb) S3 805 w1MB Video RAM NeXTSTEP 3.2 DPT 2012/90, BIOS Rev. 5B, purchased late summer 1993 2 HP C2247 1GB disks mounted for users' files, SCSI ID's 0 and 1. Typically, the first machine will run for several days, sometimes as long as two and a half weeks and then crash. The second machine hasn't had a crash for many weeks but stuff like this shows up in its /usr/adm/messages file: [...] ramesh@bodhi.esys.cwru.edu (Ramesh Dodamani) EISA/VL 486DX2 66, SiS Chipset, AMI BIOS, 32 MB DPT 2022/95, BIOS Rev. 2B Archive Viper 525MB Tape Drive Seagate 1GB Drive. Varun Mitroo <mitroo@magnus.acs.ohio-state.edu> I get crashes very rarely, about once a month. My hardware: AMI Enterprise III 486DX2 66, 48 MB RAM DPT 2022/95, ROM upgraded to 2C Toshiba 3401 CD-ROM Pinnacle 650 MB Magneto=optical drive NeXT color printer Maxtor 1240 HDD The optical drive and printer are external, and the printer is terminated. Everything usually works very well, and I sometimes get occasional time outs which usually recover. It's annoying, but not that frequent, about once a week. -- Markus Wenzel System administration, Consulting, Networking mow@marsu.tynet.sub.org on... NeXTSTEP / Unix / Novell / Windows NT IRC: Marsu Veni, vidi, NeXTSTEPi.
Newsgroups: comp.sys.next.hardware,comp.sys.next.bugs From: yong@iitmax.iit.edu (Yong Yoo) Subject: Help:: Adaptec 1542B with new bios. Message-ID: <1994Mar6.001322.15651@iitmax.iit.edu> Organization: Illinois Institute of Technology, Chicago Date: Sun, 6 Mar 94 00:13:22 GMT I've just upgrade the new bios on my Adaptec 1542B controller card so it can recognize drive bigger than 1 gig. I think it was extended translation schemes or something like that. Anyway, After installing the new bios, I've installed MSDOS on first to see if the drive was ok. After that I took out the dos partition and ran the NeXTStep 3.2 installation. It was fine until after loading the controller drivers, (parts where system loads whole bunch of drivers like, adaptec 1542B drivers DPT drivers and ide drivers) I pressed 2 to continue without adding new drivers(I assumed even with new bios, old drivers will be compatiable) that is when I get a weird messages and keep printing ": thread: waitForInterrupt: returns -735" on the screens. I've had NeXTStep 3.2 for a while on same scsi drive( Seagate ST41650N, 1.35 gig formated) just that I could only used upto 1 gig. I though it might be the new bios and wanted to see if what happen if I disable the extended translation schemes. Interesting enough I had no problem installing the NeXTStep. So I do know that the bios is fine just that when I enable the extended translation schemes, it crash and keep printing out same error messges on the screen. So I am assuming I have to get a new drivers before I can installing NeXTStep. But only new drivers I can find for Adaptec is one I can upgrade AFTER I've installed the NeXTStep on the drive already. Are there other people who had same problem and/or knows how to fix it? I would be very appreciated if you to send me an email or post to the news net. My email address is "yong@iitmax.acc.iit.edu" thanks in advance. Yong Yoo
Newsgroups: comp.sys.next.bugs From: Nitezki@NiDat.sub.org (Peter Nitezki) Subject: Re: AdaptorOracle uses tables not included in the default V7 oracle catalog Message-ID: <1994Mar5.092332.24884@nidat.sub.org> Sender: nitezki@nidat.sub.org Organization: private site of Peter Nitezki, Kraichtal, Germany References: <2l2ndm$c6c@obelix.WHU-Koblenz.de> Date: Sat, 5 Mar 1994 09:23:32 GMT In article <2l2ndm$c6c@obelix.WHU-Koblenz.de> wolfgang@wi.WHU-Koblenz.de (Wolfgang Roeckelein) writes: > In article <1994Feb27.144146.18675@nidat.sub.org> > Nitezki@NiDat.sub.org (Peter Nitezki) writes: > > In article <2kaq04$j16@obelix.WHU-Koblenz.de> > > wolfgang@wi.WHU-Koblenz.de (Wolfgang Roeckelein) writes: > > Neither NeXT nor Oracle ever promised the DBAdaptor-Kit to be > > compatible with Oracle 7. It is still bound to version 6! > this is absolutly incorrect. Look at the document > 1359_DBKit_Adaptor_Availability_Guide.ps from NeXTanswers. > And the Adpator from 3.2 _runs_ with V7 except for > the mentioned trouble. > So please watch what you are saying!!!!!!!!!!!!!! > If you still need to load the version 6 bindings to run DBKit on Oracle 7. Now what would you say DBKit is bound to!? I still see no argument against the core of my previous posting. But I must admit that I did t read all relevant publications befor making that statement. Again, I'm stubborn enough to be not convinced of my error! -- Peter Nitezki | Nitezki@NiDat.sub.org # Blessed art thou who knoweth Staarenbergstr. 44 | Tel.: +49 7251 62495 # not about the pleasure and D-76703 Kraichtal | Fax : +49 7251 69215 # delight of being hooked GERMANY | NeXTmail welcome !!! # up to the Net. Peter 1,3-5
Newsgroups: comp.sys.next.bugs From: Greg_Anderson@afs.com (Gregory H. Anderson) Subject: Re: File locking bug in 3.1 Message-ID: <1994Mar6.000436.5579@afs.com> Sender: greg@afs.com References: <762689800snz@amber.demon.co.uk> Date: Sun, 6 Mar 1994 00:04:36 GMT In article <762689800snz@amber.demon.co.uk> richard@amber.demon.co.uk (Richard Frith) writes: > The advisory locking of regions in files does not seem to work in 3.1 > Intel. Can anyone tell me if this major bug is fixed in 3.2? > I really want to use lockf() or the locking functions of fcntl() as I > have unix software to port which needs locking of regions rather than > entire files (locking of an entire file works). To the best of my knowledged, a working version of lockf() has never been implemented in any version of NEXTSTEP. -- Gregory H. Anderson | "History is the fiction we invent to Revisionist Autobiographer | persuade ourselves that events are Anderson Financial Systems | knowable and that life has order and greg@afs.com (NeXTmail OK) | direction." -- Calvin & Hobbes, 7/19/93
From: sanguish@digifix.com Newsgroups: comp.sys.next.announce,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.bugs,comp.sys.next.programmer,comp.sys.next.marketplace Subject: Quick Guide to NEXTSTEP information on the Internet Date: 6 Mar 1994 23:09:00 -0500 Organization: Next Announcements Message-ID: <2le9cs$i6r@digifix.digifix.com> This post is made weekly, to help 'point' users to more NEXTSTEP information Topics include: comp.sys.next newsgroups related newsgroups comp.sys.next newsgroups mailing list ftp sites NeXTanswers comp.sys.next.* newsgroups -------------------------- Comp.Sys.Next.Advocacy This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. Comp.Sys.Next.Announce Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) The NEXTSTEP FAQs are posted here monthly as well. This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Comp.Sys.Next.Bugs A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT- specific groups as well. Comp.Sys.Next.Hardware Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. Comp.Sys.Next.Marketplace NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. Comp.Sys.Next.Misc For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! Comp.Sys.Next.Programmer Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. The NEXTSTEP programmer FAQs are posted here. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. Comp.Sys.Next.Software This is a place to talk about [third party] software products that run on NEXTSTEP systems. Comp.Sys.Next.Sysadmin Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. related Newsgroups ------------------ Comp.Soft-Sys.Nextstep Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. Comp.Lang.Objective-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. Comp.Object Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original Comp.Sys.Next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News ------------------------------------------- Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep-d next-advocacy-d next-announce-d next-bugs-d next-hardware-d next-marketplace-d next-misc-d next-programmer-d next-software-d next-sysadmin-d (For a full description, send mail saying LISTS to <digestif@antigone.com>). The subscription syntax is essentially the same as LISTSERV's. To subscribe, send a message to <digestif@antigone.com> saying: SUB Listname YourName Example: SUB next-hardware-d John Doe The ftp sites ------------- cs.orst.edu: The main site for North American submissions nova.cc.purdue.edu: Lots of older stuff, but very short on disk space ftp.informatik.uni-muenchen.de: In Germany. terra.stack.urc.tue.nl (Dutch NEXTSTEP User Group) and cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp.next.com: See the below ftp.next.com and NextAnswers@next.com ------------------------------------- From the document 1000_Help from ftp.next.com Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, or you can transfer them by anonymous ftp. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to NeXTanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO <your-address> If you have any problem with the system or suggestions for improvement, please send mail to NeXTanswers-request@NeXT.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to NeXTanswers-request@NeXT.com. Written by: Eric P. Scott eps@toaster.SFSU.EDU and Scott Anguish sanguish@digifix.com Additions from: Greg Anderson (Greg_Anderson@afs.com) and Michael Pizolato (Michael_Pizolato@afs.com)
Newsgroups: comp.sys.next.bugs From: gerti@tms-gmbh.de (Gerd Knops) Subject: SCSI fails with SR_IOST_INVALID Message-ID: <CM5nEF.138@tms-gmbh.de> Sender: usenet@tms-gmbh.de Organization: tms GmbH, Regensburg, Germany Date: Fri, 4 Mar 1994 19:27:50 GMT Whenever I try to use a SCSI-Command in the range of 0xcx, I get the SR_IOST_INVALID error (scsireg.h says: /* should never be seen */). This is on white hardware, with an Adaptec 1542B. Any ideas? Gerd
Newsgroups: comp.sys.next.bugs From: fozztexx@nvc.cc.ca.us (Chris Osborn) Subject: Re: Fax Software question. Message-ID: <CM9A55.94M@nvc.cc.ca.us> Sender: news@nvc.cc.ca.us Organization: Napa Valley College References: <2l4qbv$1n2@steffi.demon.co.uk> <1994Mar4.075543.29640@urz.unibas.ch> Date: Sun, 6 Mar 1994 18:31:52 GMT In article <1994Mar4.075543.29640@urz.unibas.ch>, Robert Frank <frank@ifi.unibas.ch> wrote: >No, it shouldn't. The little header is generated by the receiving device, >so the sending software can't say what will happen. It could be that the >receiving device adds a lot of other things to the header, or that it >leaves it out altogether. I tend to doubt it's the receiving device that makes the header, since I get lots of faxes on my NeXT that have headers in different fonts, and also sometimes with goofy characters (like the picture of a telephone). I think it's the sender that puts the little header up there. I too have never seen a header line on the faxes that I have sent from my NeXT. -- Chris Osborn, Network Administrator Voice: 707 253 3130 Napa Valley College Fax: 707 253 3063 fozztexx@nvc.cc.ca.us fozztexx@groucho.sonoma.edu MIME/NeXTMail ok at both
From: dtw07138@menudo.uh.edu (David Wang) Newsgroups: comp.sys.next.hardware,comp.sys.next.bugs Subject: please help me with a 1542B/Sony 541 CD-ROM Date: 7 Mar 1994 09:58:07 -0600 Organization: University of Houston Distribution: usa Message-ID: <2lfiuf$lsf@menudo.uh.edu> Hi, I just got a Adaptec 1542B SCSI controller and a sony 541 single speed CD-ROM. I was trying to install NeXT Step with it, and so far, no luck. The strange symptom is that the SCSI Bios of the SCSI card refuses to load upon bootup, (it works in my friend's machine, so may be my MB just doesn't like the SCSI controller? I have a 486 VLB board with Sis chipset) However, if I go under DOS and install the EZ SCSI software the CD-ROM seems to work just fine. However, NS doesn't like this one bit. Teh installation gets hung on the part where NS tries to Reset the SCSI bus, I've tried it about 10 to 15 times now, and at times, NS can get as far as recognizing the CD-ROM, and then says that it has problems reading off of the CD-ROM or something. However it always ends up getting hung up, maybe after one or several AHA timeouts (what are these anyways?) The CD-ROm and the controller card seems to be good because they work under DOS, and if I remember correctly, these are the same hardware that NeXT uses. Do I need a double speed CD-ROM just to install NS? The card is in its defaul configuration, and I've changed the interrupt and DMA channels and SCSI ID's to all sorts of combinations, all without any luck. right now, the setting is that the Adaptec 1542B is running with SCSI ID 7, the CD-ROM has SCSI ID 0. and the card is using IRQ 11, and DMA channel 5. any one has any ideas? Thanks in advance David
From: annard@theborg.stack.urc.tue.nl (Annard Brouwer) Newsgroups: comp.sys.next.hardware,comp.sys.next.sysadmin,comp.sys.next.bugs Subject: Can't ftp large files, only small ones Followup-To: comp.sys.next.sysadmin Date: 7 Mar 1994 19:44:41 GMT Organization: the Borg Distribution: world Message-ID: <2lg079$bi@theborg.stack.urc.tue.nl> Hello, Perhaps someone can help me shed some light on this. I have at my disposal a brand X 486 machine running NEXTSTEP 3.2. Within it is a SMC EtherCard Plus Elite 16 Ethernet Adapter. Now the problem: 1) I upgraded from 3.1 to 3.2 using the Upgrader.app and I couldn't connect to machines outside my domain. However within my domain things went OK (though not too fast). 2) I noticed the same problem at home, where I had used the same method to upgrade my NeXTstation. So I did the upgrade from scratch. Now I can get out of my domain but if I want to ftp a file which is larger than say 26K, the process is stuck and the only thing left to me is to abort the ftp (which works). So it looks like large packets cannot arrive intact for some reason. With a ethernet sniffer I found out that the symptom of problem 1 was due to the fact that the next acknowledged packets which would be sent in the future while the other side ignored them and tried to retransmit. This might be considered a bug because the problem disappeared after an upgrade from scratch. I haven't "sniffed" at problem 2 yet. What is happening here? Is it due to a software problem, or is something wrong with the ethernet card? Annard -- Give me enough bandwidth and I can transmit the earth. - Annard Brouwer annard@stack.urc.tue.nl (NeXTmail appreciated)
From: doyle@prufrock.rockefeller.edu (Mark D. Doyle) Newsgroups: comp.sys.next.bugs Subject: Workspace Manager, root, and NFS Message-ID: <1994Mar7.165005.15205@rockyd.rockefeller.edu> Organization: Rockefeller University Date: Mon, 7 Mar 94 16:50:05 EST Hi all, I was helping a friend of mine at another site (he posted last week) with a problem that I thought resulted from a botched installation of tcp-wrapper stuff, but then I noticed that the same thing happens on our small network of NeXT's here. The network has one NetInfo server that also exports directories from an external drive to two clients. The problem is that logging in as root on a client machine and selecting a particular directory or file causes the Workspace to crash. The directories that caused this behavior was one user's home directory (out of seven) and 2 applications in /LocalApps. A little investigation showed that the two apps that caused the crashes (with NFS read error 13, by the way), were two old 2.0 apps that were installed with permission 750. Changing them to 755 fixed the problem. I had exported all the directories without root access for the client machines. So I went and granted them root access and now everything (in particular that one user's directory) works perfectly. The only thing I don't understand is what in that particular user's directory would cause such behavior. (The overall permissions on his directory are the same as everyone else's.) I should add that the Workspace only crashed for root on the client machines. Any other user could log in and select the directory and the apps without crashing, as could root on the server. I would be grateful if anyone has any ideas as to why this had happened for this particular user only. I have posted this to c.s.n.bugs because I consider it a bug for root's WorkSpace to crash just because you select a file/directory that has the wrong permissions. I guess it some conflict with root's idea that it should be able to read anything and the fact that it has been denied root access to the NFS server. Thanks for any info, Mark Doyle doyle@theory.rockefeller.edu
Newsgroups: comp.sys.next.bugs,comp.sys.next.hardware From: leigh@cs.uwa.oz.au (Leigh SMITH) Subject: patch for NS/FIP ET4000 video bug Message-ID: <leigh.762771948@wambenger> Summary: If you can't get your ET4000 at 1024x768, try this Keywords: ET4000,NS/FIP Sender: usenet@bilby.cs.uwa.edu.au Organization: Dept. Computer Science, University of Western Australia. Date: Fri, 4 Mar 1994 09:05:48 GMT I have produced a patch to the TsengLabs4000 driver to solve a bug setting the dot-clock. For some reason, the 3.2 driver does not program the Tseng Labs clock bit which allows access to the extended clocks, such as 65Mhz for 1024x768. I presume the original hardware tested must power up with the bit set whereas my ET4000 card (Total Peripherals clone) doesn't. This patch may not solve your ET4000 problems but if you have the developer version and can't get your card running, it could be worth the effort. The patch is pretty simple: 1. Copy the directory /NextDeveloper/Examples/DriverKit/TsengLabsET4000 2. Patch the file TsengLabsET4000/TsengLabsET4000_reloc.tproj/TsengLabsET4000.m (relative to the directory you copied to), in the TsengLabsET4000_1024x768x2x60hz_crtControllerRegisters[] array: Old - {0x14,0x00}, {0x15,0x04}, {0x16,0x22}, {0x17,0xC3}, {0x18,0xFF}, {0x33,0x00}, {0x35,0x00}, {-1, 0} New - {0x14,0x00}, {0x15,0x04}, {0x16,0x22}, {0x17,0xC3}, {0x18,0xFF}, {0x33,0x00}, {0x34,0x02}, {0x35,0x00}, {-1, 0} (Should be line 68) and in TsengLabsET4000_640x480_VGA_crtControllerRegisters[] array: Old - {0x33,0x00}, {-1, 0} New - {0x33,0x00}, {0x34,0x00}, // Disable the 65Mhz clock {-1, 0} (Should be line 237) 3. Edit TsengLabsET4000/English.lproj/Localizable.strings (my apologies to our foreign language counterparts, your kilometres may vary :-) and change the names to be able to distinguish them from the original driver, say: "TsengLabsET4000" = "My Tseng Labs ET4000"; "Long Name" = "My Tseng Labs ET4000-Based SuperVGA Video Adapter"; 4. Double click TsengLabsET4000/PB.project to bring up Project Builder, click Builder and then click Build with the target set to bundle. This should recompile the driver and produce TsengLabsET4000.config. 5. Log in as root. 6. Rename /NextLibrary/Devices/TsengLabsET4000.config to OriginalTsengLabsET4000.config. 7. Copy your new TsengLabsET4000.config into /NextLibrary/Devices 8. Run Config.app, select Display and remove your Default VGA driver (blech) then add My Tseng Labs ET4000 from the list of drivers (the original is still called Tseng Labs ET4000 - be careful. 9. Save the new configuration, logout and restart the machine. If everything worked, you should see a MUCH better image. If not, you'll need to restart with config=Default and BugNeXT about fixing their driver (not me!). 10. Caveat: When closing down the driver to 640x480 mode to tell you its safe to turn off the computer, the VGA programming seems to be done by code other than the ET4000 driver. Consequently, the extended dot-clock stays activated and this final message is smeared across the screen. I don't know if this is correct, but SoftPC seems to use the 640x480 code in the driver correctly, so I presume the solution for this problem is an update to the shutdown code. It's a small price to pay for 1024x768, IMHO :-). 11. Email me if these instructions make no sense, and I'll see if I can help. -- Leigh Smith, Dept of Computer Science: leigh@cs.uwa.edu.au, +61-9-380-1947 University of Western Australia Fax: +61-9-380-1089 Home NeXTMail: leigh@psychokiller.dialix.oz.au, +61-9-382-3071 *---=----=----=----=----=----=----=----=----=----=----=----=----=----=----=---*
Newsgroups: comp.sys.next.bugs From: boonlow@kits.sfu.ca (Boon Chong Benjamin Low) Subject: Terminal problems while logging in ... Message-ID: <boonlow.763111263@sfu.ca> Sender: news@sfu.ca (seymour news) Organization: Simon Fraser University, Burnaby, B.C., Canada Date: Tue, 8 Mar 1994 07:21:03 GMT Hey guys, I am running NeXTStep (for Intel Processors) and I am getting some problems with terminal emulations. Whenever I login from home ( through serial-line / modem ), my screen would be all juggle/mess up. Which ever command I type, ls or whatever ... the files are all over the place on the screen. Also, if I type resize, the modem will hang up. I believe it is a local problem, maybe..., as I uses the vt-100 emualtion both sides ( home and in school ), or even if I loginto a NeXT machine ( pepper, orange e.t.c ) I will get the same results. It seems to be a pretty simple problem on my side, but sign... I just can't figure it ;-( If anyone can offer me some advice/solution, I would be really gratefull. Thanks for any help provided. Sincerely, Ben.
From: tgschwin@PROBLEM_WITH_INEWS_GATEWAY_FILE (Thomas Gschwind) Newsgroups: comp.sys.next.bugs Subject: keyboard.app v3 Date: 8 Mar 1994 12:55:15 GMT Organization: Technical University Vienna, Austria Message-ID: <2lhsjj$fj3@email.tuwien.ac.at> is there a newer version than keyboard.app v3??? (Version 3 is quite buggy (using a 101 key pc keyboard)!). Or is there some other way to redefine the function keys? (In our version [F2] maps to !, [F3] to "...). Not, what you usually want. :-( Thomas -- \|/ (o o) O------------oOo--(_)--oOo------------O------------------------------------O | email: | | | tom@siegfried.smc.univie.ac.at | DOS is too DOSASTROUS for you??? | | tgschwin@cslab.tuwien.ac.at | Try UN*X!!! | | | | O-------------------------------------O------------------------------------O
Newsgroups: comp.sys.next.hardware,comp.sys.next.bugs From: Ralph_Jung@radical.com (Ralph Jung) Subject: Re: patch for NS/FIP ET4000 video bug Message-ID: <1994Mar8.180416.11745@radical2.radical.com> Sender: ralph@radical2.radical.com Organization: Radical System Solutions, Inc. References: <leigh.762771948@wambenger> Distribution: usa Date: Tue, 8 Mar 1994 18:04:16 GMT Leigh SMITH writes > > I have produced a patch to the TsengLabs4000 driver to solve a bug setting > the dot-clock. For some reason, the 3.2 driver does not program the Tseng Labs > clock bit which allows access to the extended clocks, such as 65Mhz for > 1024x768. I presume the original hardware tested must power up with the bit set > whereas my ET4000 card (Total Peripherals clone) doesn't. This patch may not > solve your ET4000 problems but if you have the developer version and can't get > your card running, it could be worth the effort. > I tried the patch with my old Diamond SpeedSTAR PLUS Version 4.23 and it works GREAT!!!! My out-of-sync video problem went away. 1024 X 768 on a 14" monitor isn't the best, but it sure beats the heck out of 640 X 480!!!! A job well done!!!! Thanks!!! -- Ralph Jung ( Ralph_Jung@Radical.Com ) Radical System Solutions, Inc. NeXTmail accepted rad~i~cal \'rad-i-kel\ adj. - marked by a considerable departure from the usual or traditional: EXTREME
From: wolfgang@wi.WHU-Koblenz.de (Wolfgang Roeckelein) Newsgroups: comp.sys.next.bugs Subject: Re: AdaptorOracle uses tables not included in the default V7 oracle catalog Date: 9 Mar 1994 11:24:37 GMT Organization: WHU Koblenz Message-ID: <2lkbll$rjc@obelix.WHU-Koblenz.de> References: <1994Mar5.092332.24884@nidat.sub.org> In article <1994Mar5.092332.24884@nidat.sub.org> Nitezki@NiDat.sub.org (Peter Nitezki) writes: > If you still need to load the version 6 bindings to run DBKit on Oracle 7. > Now what would you say DBKit is bound to!? > > I still see no argument against the core of my previous posting. But I > must admit that I did t read all relevant publications befor making that > statement. Again, I'm stubborn enough to be not convinced of my error! Well, in 3.2 there is an OracleAdaptor and an Oracle6Adaptor. Also the problem are not bindings, only the tables which contain the accessible tables, etc. for the user. Wolfgang Roeckelein -- Dipl.-Wirtsch.-Inf. Voice: +49 261 6509 173 Wolfgang Roeckelein Fax: +49 261 6509 179 WHU Koblenz E-Mail: roeckelein@wi.whu-koblenz.de Burgplatz 2 (NeXTmail ok) D-56179 Vallendar Germany
From: robert@steffi.demon.co.uk (Robert Nicholson) Newsgroups: comp.sys.next.bugs Subject: Another FAX software question. Date: 9 Mar 1994 13:13:40 -0000 Organization: me organised, that's a joke. Message-ID: <2lki24$b9@steffi.demon.co.uk> When is the Address information physically saved in the FAX panel? It appears to me that a reboot/logout having made the changes to the database loses all changes. Is this like the FileViewer in that you have to logout before changes are saved? IMHO: In this particular case it should be either as soon as I hit the button Fax Now or even sooner with Add/Delete/Modify. Or at the very least through the Panels windowWillClose method. Another problem which most of you area ware of is that repeated FAXing of the same document (not using redialing) automatically removes the previous faxes cover page information. Are these problems being addressed? -- "You know what's wrong with you?" (Audrey Hepburn, Cary Grant) "No, what?" "Nothing" (Charade, 1963) (ASCII for text only messages)
Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs From: barry@ictv.com (Barry Lustig) Subject: WorkSpace Manager failure on large NFS mounted filesystems Message-ID: <1994Mar9.181536.18065@ictv.com> Keywords: NFS Fileservers mounts Workspace Organization: Nicholas-Applegate Capital Management, San Diego, CA, +1 619 687-8000 Date: Wed, 9 Mar 1994 18:15:36 GMT Both WM.app and FileMover.tool incorrectly handle disk space calculations on NFS mounted filesystems that are larger than 2GB. This might not be more than a cosmetic problem except that FileMover.tool will not copy or move files to a filesystem that it determines is full. Included below is a snippet of code from FileMover.tool. A similar snippet can be found in WM. fp@(-0x432) corresponds to f_bfree in the statfs structure and fp@(-0x43e) corresponds to f_bsize. Notice that the mulsi of f_bfree and f_bsize will force d2 to go negative when the result is between 2 and 4GB; and will overflow d2 when the result is greater than 4GB. Since the GNU C compiler supports a long long datatype, the easiest workaround would be to do the calculations into a long long instead of a long. As a temporary workaround, I changed the sense of the branch at 53ce from bne to beq. When the statfs fails, FileMover.tool defaults to doing the copy. I haven't checked to see what the behavior will be when the filesystem really is full. I assume that it will be the same as any command line copy that fills up a filesystem. 53c6 bsr _statfs 53cc tstl d0 53ce bne L53ec 53d0 movel fp@(-0x432),d2 53d4 mulsl fp@(-0x43e),d2 53da movel d2,d1 53dc movel d2,d0 To workaround: cp -p FileMover.tool FileMover.tool-NFSBUG objpatch -b 0x67 FileMover.tool 0x53ce struct statfs { long f_type; /* type of info, zero for now */ long f_bsize; /* fundamental file system block size */ long f_blocks; /* total blocks in file system */ long f_bfree; /* free blocks */ long f_bavail; /* free blocks available to ruser */ long f_files; /* total file nodes in file system */ long f_ffree; /* free file nodes in fs */ fsid_t f_fsid; /* file system id */ long f_spare[7]; /* spare for later */ }; --
Newsgroups: comp.sys.next.bugs From: rsilver@mtkgc.com (Russell Silverman) Subject: Getting an RDP trap using kermit on 3.1 Message-ID: <CMELJF.67x@mtkgc.com> Keywords: kermit, kernel trap, RDP Sender: news@mtkgc.com Organization: Sakura Global Capital, NYC Date: Wed, 9 Mar 1994 15:26:03 GMT When I use kermit ( I forget the revision, but it is the most recent ). It works fine the first time. Then, if the machine has been running for a time, ( say 8 hours ), I'll try dialing a number, and the following window pop up. Kernel Trap (<--This is in the title bar) RDP error..... type 'c' to continue When I type 'c' the window recurs, and my only course is to reboot. With my Gateway 66V, the fsck takes ~10 minutes, and I am prone to data corruption with this (non-)method. I compiled the kermit on my box. It was a clean compile, and everything works fine, downloads, etc. Once or twice I have left the machine running through some large downloads, and I have come back to rebooted machine. I'm not sure whether this is part of the same phenomena, or not. I mention it as an aside. Personally, I think the real bug here is that this glorious operating system, with so many amazing programs, including fax software, comes without any communications software, whacky. When I get a solution to this problem, and get some time, I will write a NeXT GUI for kermit and post it. Has anyone else experienced this with 3.1 or 3.2 ? I am going to be upgrading, so the fix may be in the mail. Any help would be appreciated.
From: jr@sade.ka.sub.org (Jochen Richter) Newsgroups: comp.sys.next.bugs Subject: Console messages netmsgserver Date: 10 Mar 1994 06:04:01 GMT Organization: The Home Of The Pilhuhn Message-ID: <2lmd8h$nr7@pilhuhn.pilhuhn.sub.org> Whenever I start my network I get this console messages: Mar 10 07:02:31 sade netmsgserver[21]: netname_main.msg_send fails, kr = -102. Mar 10 07:02:31 sade netmsgserver[21]: netname_main.port_type fails, kr = 4. What could be the reason for that? -- Jochen Richter Phone: +49-721-696922 Zahringerstrasse 57 Fax: +49-721-696988 D-76133 Karlsruhe e-mail: jr@sade.ka.sub.org Germany jr@resy2.kfk.de (>50K)
From: rbp@investor.pgh.pa.us (Bob Peirce #305) Newsgroups: comp.sys.next.bugs,comp.sys.next.software,comp.apps.spreadsheets Subject: Improv - Saving graphs - bug Message-ID: <1994Mar9.121903.17722@investor.pgh.pa.us> Date: 9 Mar 94 12:19:03 GMT Organization: Cookson, Peirce & Co., Pittsburgh, PA I am temporarily news impaired. I can post but my feed is out. Since I do not know when news will begin to flow again, I would appreciate it if you would mail any solutions to venetia@investor.pgh.pa.us. Thank you. Now for the problem. I do a lot of graphing in Improv (1.0 on a NeXT). I notice that much too frequently saving the spreadsheet loses the graphs. I have watched this at the command line level and what seems to happen is changes are put in a new spreadsheet named xxx~.imp. When you save the sheet, xxx~.imp is copied to xxx.imp. Unfortunately, the graphs don't always make it. I have had to resort to saving the sheet and copying the graphs to a safe place before exiting. This is not a very satisfactory solution. Does anybody know of a guaranteed safe series of steps to save and exit an Improv spreadsheet without losing the graphs? -- Bob Peirce, Pittsburgh, PA rbp@investor.pgh.pa.us 412-471-5320 venetia@investor.pgh.pa.us [NeXT Mail] !uunet.uu.net!pitt!investor!rbp [UUCP]
Newsgroups: comp.sys.next.bugs From: fozztexx@nvc.cc.ca.us (Chris Osborn) Subject: Loginwindow brightness restoration strangeness Message-ID: <CMEz23.3wt@nvc.cc.ca.us> Sender: news@nvc.cc.ca.us Organization: Napa Valley College Date: Wed, 9 Mar 1994 20:18:02 GMT Does anyone know why loginwindow subtracts 3 from the brightness when it restores it? I hacked a program to save the current settings on logout, and the screen will keep getting dimmer and dimmer each time you login. Also, why is 0 the maximum volume? And why is the value clipped at 29? -- Chris Osborn, Network Administrator Voice: 707 253 3130 Napa Valley College Fax: 707 253 3063 fozztexx@nvc.cc.ca.us fozztexx@groucho.sonoma.edu MIME/NeXTMail ok at both
Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs From: barry@ictv.com (Barry Lustig) Subject: Re: WorkSpace Manager failure on large NFS mounted filesystems Message-ID: <1994Mar10.213107.1149@ictv.com> Keywords: NFS Fileservers mounts Workspace Sender: Barry Lustig <barry@nacm.com> Organization: Nicholas-Applegate Capital Mgmt., San Diego, CA, 619 687-8000 References: <1994Mar9.181536.18065@ictv.com> Date: Thu, 10 Mar 1994 21:31:07 GMT Please note that the patch I posted is against the 3.2 version of FileMover.tool. It is very unlikely that it will work again any other version. Barry Lustig P.S. objpatch was written by Louie Mamakos. Here is a copy. CUT HERE --------------------------------------------- #!/usr/bin/perl $objfile = shift(@ARGV); @PATCHES = (); while ($objfile =~ /^-/) { if ($objfile =~ s/^-(B|b)//) { $objfile = shift(@ARGV) unless $objfile ne ''; push(@PATCHES, 1, $objfile); } elsif ($objfile =~ s/^-(w|W)//) { $objfile = shift(@ARGV) unless $objfile ne ''; push(@PATCHES, 2, $objfile); } elsif ($objfile =~ s/^-(L|l)//) { $objfile = shift(@ARGV) unless $objfile ne ''; push(@PATCHES, 4, $objfile); } elsif ($objfile =~ /^-n/) { $nflag = 1; } elsif ($objfile =~ /-v/) { $vflag = 1; } $objfile = shift(@ARGV); } open(OTOOL, "/bin/otool -l $objfile|") || die "otool: $!"; while(<OTOOL>) { chop; next unless /^Section$/; $_ = <OTOOL>; chop; die "1" unless /^ sectname (.*)$/; $sectname = $1; $_ = <OTOOL>; chop; die "2" unless /^\s*segname (.*)$/; $segname = $1; $_ = <OTOOL>; chop; die "3" unless /^\s*addr 0x(.*)$/; $addr = hex($1); $_ = <OTOOL>; chop; die "4" unless /^\s*size 0x(.*)$/; $size = hex($1); $_ = <OTOOL>; chop; die "5" unless /^\s*offset (.*)$/; $offset = $1; next unless $size > 0 && $offset > 0; $info = sprintf("%08x %08d %08d $segname $sectname\n", $addr, $size, $offset); push(@segs, $info); } close(OTOOL); @segs = sort @segs; print "Address Length File Offset\n" if $vflag; print @segs if $vflag; open(FILE, "+<$objfile") || die "open '$objfile': $!"; while ($_ = shift) { $addr = $_; if ($addr =~ s/^0(x)*//) { $addr = hex($addr); } $fileoff = &getoffset($addr); @P = @PATCHES; while ($patlen = shift(@P)) { $patvalue = shift(@P); ($patvalue =~ m/^no+p$/) && ($patvalue = "0x4e71"); ($patvalue =~ m/^moveq_#0,d0$/) && ($patvalue = "0x7000"); if ($patvalue =~ s/^0(x)*//) { $patvalue = hex($patvalue); } $old = &getnbytes($fileoff, $patlen); printf "Writing $patlen bytes of 0x%x (was 0x%x) to 0x%x (file offset %d)\n", $patvalue, $old, $addr, $fileoff; $nflag || &putnbytes($fileoff, $patvalue, $patlen); $fileoff += $patlen; } } sub getoffset { local($value) = @_[0]; local($realvalue) = $value; local($addr, $size, $offset, $segname, $sectname, $off, $fileoff); if ($value =~ s/^0(x)*//) { $value = hex($value); } foreach $seg (@segs) { ($addr, $size, $offset, $segname, $sectname)=split(/\s+/,$seg); $addr = hex($addr); if ($value >= $addr && $value < ($addr + $size)) { $off = $value - $addr; $fileoff = $off + $offset; #print "Address is at offset $off, at file address $fileoff\n"; return ($fileoff); } } die "Address $realvalue is not in any segment of file\n"; } sub putnbytes { local($txtaddr) = @_[0]; local($value) = @_[1]; local($length) = @_[2]; local($offset); $word = 0; vec($word, 0, 8*$length) = $value; seek(FILE, $txtaddr, 0) || die "seek $!"; syswrite(FILE, $word, $length) || die "read $!"; return $value; } sub getnbytes { local($txtaddr) = @_[0]; local($word) = 0; local($length) = @_[1]; seek(FILE, $txtaddr, 0) || die "seek $!"; read(FILE, $word, $length) || die "read $!"; $value = vec($word, 0, 8*$length); return $value; } --
Newsgroups: comp.sys.next.bugs From: ebarker@world.std.com (michael j barnes) Subject: Panic with SoftPC and Novell Message-ID: <CMIE6K.Jpu@world.std.com> Keywords: Next SoftPC Organization: The World Public Access UNIX, Brookline, MA Distribution: USA Date: Fri, 11 Mar 1994 16:37:31 GMT I just Installed 3.2 on a Dell DGX. I ran SoftPC and it worked fine. When I attatched the PC to our network and ran SoftPC, I got a system panic that resulted in a disk crash requireing fsck to be run several times. This action is repeatable. Thanks for any clues. ebdOD
Newsgroups: comp.sys.next.bugs From: lloyd@world.std.com (Christopher Lloyd) Subject: bug in cc optimizer Message-ID: <CMIu7B.85C@world.std.com> Organization: The World Public Access UNIX, Brookline, MA Date: Fri, 11 Mar 1994 22:23:34 GMT First of all: AAAAAAAAAAAAIIIIIIIIIIIIIIIGGGGGGGGGGHHHHHH!!!!!!!! There, I feel better now. Try this at home, both sets of conditionals should succeed, however this is not the case on m68k, (pair.one==swap.one && pair.two==swap.two) fails miserably. Am I nuts? Have I been staring at code too long? I guess if I can get some confirmations I'll report it to NeXT. If someone hangs around the gcc bug lists and wants to send it in or whatever, go ahead. What a headache, Chris ---------- cut here /* cc -O bug.c -o bug Notice that the fields are switched between the two data structures, the optimizer apparently merges the && comparison of two shorts into one long comparison, and wrong-o, it doesn't work ... tested with: Reading specs from /lib/m68k/specs NeXT Computer, Inc. version cc-216.obj~13, gcc version 2.2.2 */ typedef struct PAIR { unsigned short one; unsigned short two; } PAIR; typedef struct SWAP { unsigned short two; unsigned short one; } SWAP; PAIR pair={ 1, 0 }; SWAP swap={ 0, 1 }; main() { if(pair.one==swap.one) if(pair.two==swap.two) printf("if if matches\n"); if(pair.one==swap.one && pair.two==swap.two) printf("&& matches\n"); } -- :: Christopher Lloyd :: Yrrid, Inc. :: lloyd@yrrid.com lloyd@world.std.com :: :: If I had known it was harmless, I would have killed it myself ::
Newsgroups: comp.sys.next.bugs From: ctm@ardi.com Subject: Re: bug in cc optimizer Message-ID: <CMK1Lp.Mww@cobra.cs.unm.edu> Sender: news@cobra.cs.unm.edu Organization: ARDI References: <CMIu7B.85C@world.std.com> Date: Sat, 12 Mar 1994 14:01:05 GMT In article <CMIu7B.85C@world.std.com> lloyd@world.std.com (Christopher Lloyd) writes: >First of all: AAAAAAAAAAAAIIIIIIIIIIIIIIIGGGGGGGGGGHHHHHH!!!!!!!! > >There, I feel better now. > >Try this at home, both sets of conditionals should succeed, however >this is not the case on m68k, (pair.one==swap.one && pair.two==swap.two) >fails miserably. > >Am I nuts? Have I been staring at code too long? I guess if I can >get some confirmations I'll report it to NeXT. > >If someone hangs around the gcc bug lists and wants to send it in or >whatever, go ahead. Works fine under gcc 2.5.8. Gosh, you must feel like I did when I found the bugs that NeXT introduced to the *assembler* in 3.1. --Cliff ctm@ardi.com
Newsgroups: comp.sys.next.bugs From: paul@seer.demon.co.uk (Paul Lynch) Subject: Re: Another FAX software question. Message-ID: <1994Mar12.205241.29015@seer.demon.co.uk> Sender: paul@seer.demon.co.uk Organization: P & L Systems References: <2lki24$b9@steffi.demon.co.uk> Date: Sat, 12 Mar 1994 20:52:41 GMT In article <2lki24$b9@steffi.demon.co.uk> robert@steffi.demon.co.uk (Robert Nicholson) writes: > When is the Address information physically saved in the FAX panel? > > It appears to me that a reboot/logout having made the changes to the > database loses all changes. > > Is this like the FileViewer in that you have to logout before changes > are saved? > > IMHO: In this particular case it should be either as soon as I hit the > button Fax Now or even sooner with Add/Delete/Modify. Or at the very > least through the Panels windowWillClose method. It seems to happen when the object is freed, which is usually when the application quits. I find that if I update the database and then check it with another app, the updates don't take; but when I quit the updating app, the updates are written back. I agree that for this to work in an open, cooperative manner, it should write when Add?modify is used, and it should use a locking mechanism. > Another problem which most of you area ware of is that repeated FAXing > of the same document (not using redialing) automatically removes the > previous faxes cover page information. Annoying, isn't it? Paul -- Paul Lynch P & L Systems (NeXTmail) paul@seer.demon.co.uk Tel: (0494)671501 9 Stable Lane, Seer Green, Fax: (0494)680228 Bucks, HP9 2YT, UK
Newsgroups: comp.sys.next.hardware,comp.sys.next.bugs From: mimbeck@mimbeck.hanse.de (Martin Imbeck) Subject: Re: please help me with a 1542B/Sony 541 CD-ROM Followup-To: comp.sys.next.hardware,comp.sys.next.bugs References: <2lfiuf$lsf@menudo.uh.edu> Organization: NeXTstep site Date: Fri, 11 Mar 1994 12:09:57 GMT Message-ID: <1994Mar11.120957.459@mimbeck.hanse.de> Distribution: usa David Wang (dtw07138@menudo.uh.edu) wrote: : Hi, I just got a Adaptec 1542B SCSI controller and a sony 541 single speed : CD-ROM. I was trying to install NeXT Step with it, and so far, no luck. I use the same configuration and had no problems. Perhaps you should try the default settings of the adaptec and the motherboard's bios. Or disable the floppy controller on the adaptec, I heard there is some trouble with it. -- Martin Imbeck / phone +49-40-874436 / No NeXTmail mimbeck@mimbeck.hanse.de / NeXTstep / please!
From: premise@pindar.ccs.itd.umich.edu (Sean Michael Willson) Newsgroups: comp.sys.next.bugs Subject: Workspace Crash Date: 13 Mar 1994 19:08:26 GMT Organization: University of Michigan Message-ID: <2lvoba$kjg@lastactionhero.rs.itd.umich.edu> Hello, I am a student at the University of Michigan Ann Arbor. We have a small NeXT lab on central campus that has 19 next motorollas, running NeXTSTEP version 3.0. Whenever I log on to one of these machines (It happens on all of them) I can do work for a while but then the workspace manager crashes. I cannot figure out why. I have tried replacing my NeXT file (My sys admin told me to) but that did nothing but lose everything I did in customization. I sometimes can work for a few hours switching between apps and then it will crash. Other times I will log in, open mail, then when I click on the wrokspace manager CRASH. It becomming more and more frequent and effecting my work. When it crashes it gives a message the the "Workspace manager has crashed. PLease exit and save changes if possible" or something like that. Is anyone can help at All I would owe them forever. Please I love NeXT's and I have work to do. Please help ASAP. Sean Willson premise@churchst.ccs.itd.umich.edu -- _________________________________________________________________________ |Sean Michael Willson | E-Mail | |Co-President of wwiNc. | premise@churchst.ccs.itd.umich.edu | | NeXT Mail Welcome | | "Do not believe in miracles. Rely on them." | |________________________________________________________________________|
From: juliana@semlab1.sbs.sunysb.edu (Juliana L. Silva) Newsgroups: comp.sys.next.bugs,comp.sys.next.sysadmin Subject: swapfile keeps growing .... Date: 14 Mar 1994 18:55:06 GMT Organization: Institute For Theoretical Physics Message-ID: <2m2bua$1kj@ysics.physics.sunysb.edu> I noticed my system (486/66) was getting slower and the available diskspace was "disppearing". When I checked the swapfile, it was more than 100MB long. I left the system up a couple more days to see what would happen. Well, the swapfile kept growing, my hard drive ended up full and I had to reboot the machine. Has anyone experienced that? Juliana Freire juliana@cs.sunysb.edu juliana@semlab1.sbs.sunysb.edu (<---NeXT mail ok)
Newsgroups: comp.sys.next.bugs From: thf@zelator.de (Thomas Funke) Subject: Re: bug in cc optimizer Message-ID: <1994Mar13.120407.1160@gamelan.uucp> Sender: thomas@gamelan.uucp (thomas) Organization: Disorganization References: <CMIu7B.85C@world.std.com> Date: Sun, 13 Mar 1994 12:04:07 GMT In article <CMIu7B.85C@world.std.com> lloyd@world.std.com (Christopher Lloyd) writes: > > Try this at home, both sets of conditionals should succeed, however > this is not the case on m68k, (pair.one==swap.one && gcc2.5.8 does is correct with -O4, so no need to report it to GNU. Nevertheless: A severe bug. NeXT should make available a corrected version of cc. -- ------------------------------------------------------------------ Thomas Funke ** Unix-Consultant ** thf@zelator.de Brooks's Law: Adding manpower to a late software project makes it later ------------------------------------------------------------------
From: sanguish@digifix.com Newsgroups: comp.sys.next.announce,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.bugs,comp.sys.next.programmer,comp.sys.next.marketplace Subject: Quick Guide to NEXTSTEP information on the Internet Date: 14 Mar 1994 14:22:36 -0500 Organization: Next Announcements Message-ID: <2m2dhs$9uq@digifix.digifix.com> This post is made weekly, to help 'point' users to more NEXTSTEP information Topics include: comp.sys.next newsgroups related newsgroups comp.sys.next newsgroups mailing list ftp sites NeXTanswers comp.sys.next.* newsgroups -------------------------- Comp.Sys.Next.Advocacy This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. Comp.Sys.Next.Announce Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) The NEXTSTEP FAQs are posted here monthly as well. This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Comp.Sys.Next.Bugs A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT- specific groups as well. Comp.Sys.Next.Hardware Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. Comp.Sys.Next.Marketplace NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. Comp.Sys.Next.Misc For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! Comp.Sys.Next.Programmer Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. The NEXTSTEP programmer FAQs are posted here. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. Comp.Sys.Next.Software This is a place to talk about [third party] software products that run on NEXTSTEP systems. Comp.Sys.Next.Sysadmin Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. related Newsgroups ------------------ Comp.Soft-Sys.Nextstep Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. Comp.Lang.Objective-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. Comp.Object Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original Comp.Sys.Next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News ------------------------------------------- Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep-d next-advocacy-d next-announce-d next-bugs-d next-hardware-d next-marketplace-d next-misc-d next-programmer-d next-software-d next-sysadmin-d (For a full description, send mail saying LISTS to <digestif@antigone.com>). The subscription syntax is essentially the same as LISTSERV's. To subscribe, send a message to <digestif@antigone.com> saying: SUB Listname YourName Example: SUB next-hardware-d John Doe The ftp sites ------------- cs.orst.edu: The main site for North American submissions nova.cc.purdue.edu: Lots of older stuff, but very short on disk space ftp.informatik.uni-muenchen.de: In Germany. terra.stack.urc.tue.nl (Dutch NEXTSTEP User Group) and cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp.next.com: See the below ftp.next.com and NextAnswers@next.com ------------------------------------- From the document 1000_Help from ftp.next.com Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, or you can transfer them by anonymous ftp. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to NeXTanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO <your-address> If you have any problem with the system or suggestions for improvement, please send mail to NeXTanswers-request@NeXT.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to NeXTanswers-request@NeXT.com. Written by: Eric P. Scott eps@toaster.SFSU.EDU and Scott Anguish sanguish@digifix.com Additions from: Greg Anderson (Greg_Anderson@afs.com) and Michael Pizolato (Michael_Pizolato@afs.com)
From: mbecker@cs.uml.edu (Mark Becker) Newsgroups: comp.sys.next.bugs,comp.sys.next.misc,comp.sys.next.software Subject: HELP -- Mathematica 1.2, NS 2.1, WriteNow, MALFUNCTION -- HELP [LONG!] Followup-To: comp.sys.next.bugs Date: 14 Mar 1994 21:11:14 GMT Organization: UMass-Lowell Computer Science Distribution: usa Message-ID: <2m2jti$3vs@ulowell.uml.edu> Summary: problems problems problems Keywords: Mathematica NextStep WriteNow Malfunction Hello * Please forgive the wide initial cross-posting; I wanted to hit groups that seemed relevent to the software problem I'm having. Note that the Followup-To has been set to comp.sys.next.bugs. Everything below is relevent to an '030 mono cube with NS 2.1, Mathematica 1.2, and WriteNow. This particular problem resulted in a LOT of error messages which have been compressed, uuencoded and appended to this message. A couple of days ago I was typing up a short document in WriteNow [WN], using Mathematica [Mma] 1.2 to create graphs for insertion. I was trying to get an equation from Mma into WN and things weren't going well because of a difference between fonts; one font in Mma and the other in WN. At one point I got a little frustrated. Clicked on Fonts. Clicked on All Default Styles. Mma pops up a panel: Really Reset to Default Styles? Do you realy want to reset all the Styles to their default settings? This action cannot be undone. And I clicked Ok. My thinking was that it was only for *this* session. Since I wasn't logged in as root, how much could Mma do? From there I remember clicking on a cell.. and then things got ..sick.. The entire contents of the cell started to flash and disk activity went 'way up. All Mathematica-associated objects disappeared off the screen after about thirty seconds of disk thrash. Now the following situation exists: * None of the font attributes under Font are clickable for any new window. Opening an old file darkens the font attributes so I can click on them.. but I am very leery of doing so. * /usr/adm/messages is loaded with Mathematica error messages, ~200K worth. A uuencoded compressed file of them is appended to this message. Fortunately, compress gives me about 18K of material. * Starting a new Mma session produces a an error messages in the Console: *** hashtable: count differs after rehashing; probably indicates a broken invarient: there are x and y such as isEqual(x, y) is TRUE but hash(x) != hash (y) The above is repeated for everything entered, even "5+5" . What I have done so far: A comparision of sum's between the on-disk Mma and a backup done last December. There are no differences anywhere in the /NextApps/Mathematica.app tree or in /NextLibrary/Mathematica/* from the current system to the backup on optical disk. After making a backup of ~/.NeXT, I wiped that out, logged out, logged back in, and the Mma problem is still present. There are no Mathematica-associated things in the defaults database (dread). At least none that I recognize. --- Yes, I do have a BuildDisk backup of the entire system from last December. While I could rebuild the system from that backup, that seems little extreme. Can I avoid doing it? --- The text from /usr/adm/messages is below my .signature. Hints on whats wrong, pointers, *anything* on this problem would be appreciated. Is there an archive of these things somewhere? Thanks - Mark P.S. Clicking on the Help button pops up a help window.. and the following message appears in the Console: crusty Mathematica[515]: Assertion failed: NO REAL window during lockFocus on a view. The files MathematicaHelp.ma and MathematicaHelp.mb are in /NextApps/Mathematica.app and I can load them into Mathematica. But they don't come up in a Help window. Suggestions and ideas are welcome. +----------------------------------------+--------------------------+ | Mark Becker <mbecker@cs.uml.edu> | #include <std.disclaimer>| | | No NeXT Mail yet, please | +----------------------------------------+--------------------------+ -- uuencoded compressed error messages follow -- begin 644 messages.Z M'YV03<+(`0$B!P@9,'3,B*&C1@T08^34F4,G#X@V8<:@T0%"2IDV;^B4`4&F MC)@Z9\ZD<7,&1!DW8<2P*4-&0<"!!0\F7-CP8<2)%2]FW`C"21DL5#H^:0*B MR1LW:>B\&1C#!0P0=FC$L"F0H$&$"ADZA"B1HD6,&CD:1=ITZ$$7#$%005.G M+<X;(&+DT!%CA@X:>*%,2:HW1XP=(%[`01/CRYPW=>2,*3-'L9PW:LJ,H5.Y MS9H6,6#@@$OCA90B3(H$F5*$*\ZO.\7Z+!L4+5$C4*I@+2-G3IJG(&#@@>': MJ\ZP/<D"/3N4X^(\OL>$87/QXU2+/4",AG&U39DS,?.(G..B>$ZP/,?^-"LT M+8@P=L*D81-S9G60<K"#N&%U1G7P8HA'67DW&8>>;,JQ9QM'$ZW44@PV@'"2 M&6;P-@=$3]$A'U0L!0>7#/^%-QX(;YAQWW7FP8:<>K0QY]X40TR1!`@US#!$ M#E<-D>%E;,PD!PMR"71&&73L!V08%\XQQE5A%"D<0C'0P-U5*1Z7WFS+M4=4 M$T$@]804!"'5`@XSX`##%.\E20:312IYE89R#.DD"&S4X49P51Z8W'JU-0=" M`D2D,<<:(#`14QELJ%5&9&]@Y,9D<N09VYXM:LD1H((2.D08<&04E44ST)!# M$T(`2409=J0Q&0A"L/'&&(2&)@,-$@HXAZ0K8JF@GT,0,01!P.:@E0TQM)`F M"'.0$<-[;2X)`IQRYD6GG7@6>-ZD+&:Y(`A7R$>'@R"8,15)<J1AQTA28>@= M"'*4$089>>!Z98)]NN?"O??*BR"?+A*%Z:"%'IHH"+T.H2^EVOKYKZ:<>AI4 M##7(0*JIJ*HZ4JNOQ@K#K+6.=W"VN[IWQ$N\J0H"C#*"<&JJJR*)[!EL(NLL MM$3N-^V=-WRL:[U$O4&<T&AP>4"'$'@L[T]LN1SS'[[.P<])U;;5?7YIJT MI>2:.Y(42V&%`Y!S\&;'&THQ90<.2/.+M<_+!NWT55##-])6UJHXK]K;OA3S MD]S9,.71=5NY;Z5YNY$0"$70@09O;M3\+AGMSG$A=SI0#H,9E2<DPUXYE)$V MX7XF`04(<%PFU1AO4/=2?32%.^Y*(LEA1D8CZ0UD17",)$)HI"*N..-$BO!Y MPNZ1,0<<>P_-G9E&#Q\R46X@SZS019OQM_,\<_28G6M.SS<,95P?N)X@9\_N M&R&1>&>RQ(V/[<Y*"T41;Q#!45?04*</`X%4VSTX\42Y0A"DX(0D..$('!F# MJV`%`O"LI'6T(D,8H`."%AAK"$@HPA"6`((@.($('2D":Y)"A0RJ+`A4*$(( MY"4#OOA';3#Q#AFV$(,8X*`+''G!1.3P`B*-X04QI(D+Q(`^.JB%;'.H@T;" M-9^1C(L,:6C79E#D/H6T4`8XN-H55D*&-]QA"F+C#4>@<!F9?`19</I6AWY7 ME*-0`8K><8-OGN("."1+#G88".2T)@<7$"Q#*ZF#@_A'E:O`@"&S4L@+*Q60 MWV'D6]+9`I1PJ#+!0(0-:7A)D3`I!CD(Q"*\N8P<QO@&BDPA(FF`0Y%*]X8S M>+(-+I&#*$TE&!VY021X*!(.<""L]G4E-,%!)`T40BNU-7)QCU15&"1Y2$J6 MH`1;0)PLI\(@-Y3$#`\D`V*>0*&70)$E.FJ#H\C`$8MTX9G%`>8A=9#(4%WM MF!]IDC*9&0-*$L&2"LSD+>F4ADY^,I:B)*4I4:E*TEW&E6&`92BG0DL89:@, MN=0.+Z643D,*DYCO;!(RY1G)27+DF=$LPC1'"0+NE0&;C=,F"+A9(6LZ*)SC M+"<(SEF"B@:3G<-TIS$U&D](+M.CE413/C7)3W_F!Z#4!`$42DF'4Y:KH*Q$ MJ$)'VE!;XE*7$_5E(6_:SF(RDJ?)[&@S/PI-:0:TI-8\:3:WV4V7@K-1,06! M.=%9('5>5*=?=21'?SK6H%Y2GYOLIR>/NE"2+G6@3UWE05^)U!\%U:H0Q6HO M;;K.KF94KSZEIS/+*M*SFA2E-&%K2[]Y!IB&P9HRI2EE[^K5+,$SK'RM)T?N M*51,$I63@P7E2`7:5((JMI6,+6Q5'QK174ZVKA;%*49WBMEY`A6D9DWJ9]>Z MTK:2UK2H!<$*9DK77R:WG7AA[D8S"U3:_O6V@OUG87GKU%3^5JJ-'>XM(RO1 MXWJ7JSD-;U['Z]R^0K>STDTK:%7*4F^^%*ZG):=<N5M3Y.)7(?IU+5CWJMG9 MXM.V^\2M>G>K5*:V%ZJ+36A\'TM<R5+4P97-[V7Y*U;9@N"_')YN2D5KX+>* M,\&I[>Y64PSA%?>TORXV[U`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`W)_"\,:[J MC?N\XS)8%M4?'61;%SW71P:WDI>><L<Z7=A2YOA>PB[S-9,:Z_'&^=:QJ_&> M%QOL8G=XU2%M=B(;7=<F7SO*1_QV<T>=X>P,O(25/?,7"QKO`Z8QF/F>XS+_ M?>Z2OW+0L]UOD1\^[0*G,],-7DN$0]W:D:^[U6N.UJQK7MH9[SSLZ3YVH6O[ MO(9'>\"3/O!,-[WU+$_X[D/OY]'S^^X3)[3>#TWOG=O[ZZ"7/>&)'GR`EUSM M2E^\II'_]+AC/_:])SWT;9YWK5.?ZSM?M8X7ODZ>R`!$Z0=YX2?M?<."O_CC MYG;D!W>>1G\,87_X)WAD%W$`%GVV9W&;-VVZ=V\(J'UEQWW\YVW?EWJ^MGHJ M-X".-V45F'^5MWZUUWZWIW.*YG<&B!SW9X%#!WP92'+^QX%LQW@@V'(BB$@) M.'DL5H*7YX`H"(&X5V]>5VPCJ("^5WIGUW^6IG@$AX,.E7RO1X$\"(,TUX#L MEWE$J()]-W\OEX0^^'!K%G)-J($U2'RJUW;R57X%&(972()V%X1;>&KOQWG: M=80MR!,S`#_)]H-EN'__AH9/&'Y1.'Y3Z(;G!H<*X8?[!H17)X1<F'/5MX)@ MF%Q\Z(C.5WEFV'V$B'1SUH%L6&XZ:(6-B(66%XEU*&]W*($\=XDWE8FHV(DS M>'2)9XC&QWJ)2("+B(D'J(F4-X>J>(*3N'>N>'U(^(NS*(C=1H.%"(`>*("[ M&(*FV(>H:((4YWX7AX?(N(?**(?EQ8PC9XO_MX92R&FEZ'.R"(Z!-HS9F(*5 MJ&#RYWG>^!=7P8XN%@22PQO?`ARSTT0*5@5NL$!K`"Y&\"H3H3X@<`>7T2%6 MD$EWL(,-T8/L<07E(A).X$621`,V0$D-"`(K@3B6A$WV\0)TT`9P\`)5X`14 MD`14D!I$`!HO<`4'Y$F+H2HN<`,Y@$51X@)E8$<2&3%:=)%ED)%WL)$=^7MM M-HC."(HG=XC'-XU%4@,W8`,V,`9!29%!89%1490:.2M)F8JT]XY=&(\3J(XM M!`/`"(A`I8]A(P?]>"?_.!,!.9`98Y`(>2'`P9!/T1(/608128%^`1BHZ);\ M^!MR*1]TR1$"29!X.08)N9<-Z9<0*8+#1`.+E"7T01'5(3EA,"3L\I/N(A)D M``+^\2W><2L4B`-3IX2D9YAPB9CAHI@TP9AV"2N/&9EWPI<.69D4^&I^N)E% MDIIS\)DCT2ZYTR2M<YIID)I3QI$*81#X2$FP&9>S"9"VZ9@=<I"0J9>[.9D@ M\)>!Z7/060-9-)T<49VR.9>U"0*->9?;F9<*R9N4"9C/N1<SD)FB%XSAB(%, M28XV*'Y1B8[*=V^BHDC72(?$:(?;Z(H'>4L'*0>PQ((O=Z#YN8S^V8P`JH:B M>(ZN9W[%9J&MM9]LV8YC*7W:&(&Y!P)>H(<5BI\CVGS\N7TR^)^(5XX=BH@$ M6H7D":,)ZHXH"H_PEUWS>)_+Y9KZEZ'C>*,!"I6ZN*,@VH(B^J,G^H"4.*0* MUJ(4FEQ3BIY,Z(E->8O0.(J-EXXAZJ->BHU!6I98RA%`LJ4WU:5(RHGB>'K# M%XHWJ*,?^H9<BJ9S*HQ5.H17BH<B0`7-21DB\*:PN$X'*B486J,:RJ0<FJ<# MNJ>]&*=[X:AIJJ"8U`91H1&:L09"BH=G4)Q2`Z>,FJGW^*?]":E+BGJ3*J!/ M:JF/EZKV2*5:R$^>2@>@"BNCZHJE*C<NVJ>WZJ6T:*.PBJ>R^H%26:`]6JRL M:J*YBCN:L3B^RJ9X	T4*H:(A+#BJG0.H:#=X&N:J<;&*M.RJQ0RJ?@JJG1 M.FH*2JV]*JK8ZHJ0*4N:Q*W*^:VVZJ[BNH!*Z6^1FJQ/F8OJ2JM&ZJ\D2H:S M-ZUYD#OS^JLKJJWZZJVHRA"-NJK_NH3'.K!W6K`!V(:\6*L8JZJXRF'R:JWT M.JBN2+$4L:\7^Q<F:ZQU*GSGJJSI*HWK>JG]JK$+.ZX,R&&=^JDJ*[%Q%:RG MNJ@E&ZX_"[!?6HN2BK,&J[,(:Z`S^ZY!>U8I&ZI&NW,NVZTC$;.-JI6RMHF! MJ*3FFH92&[*DZ*QG^A=D^X<,*ZU"VYQ$R[7UNJ)(&[9**[-P^ZA+Z;$W"[+1 M*++4^*PT$+>/"*BY.K2\6K1Y>[2FRK?T^*)_2[-H:[-J2[ADFH-N*Z69JKAF MV[!UNZL1&[D[M[?\NK2)"[@"^ZH?"X53:[AF"KJ7B[596+IW>ZTLJ[>3N[I^ MV[J86ZZ:^XSFJ*=4&*66*[RX*Y:-:[>/B[>]*[G"*K:9JI\R6J+D&KBP.[BR MR[9E^KG+B[UE.Z/P.HQ;R[O&.+%$4K&4F[#D*[=`&["F5[Q.^;V%V[8\^K:8 M>;):^[#5*KWK&U?WVBZWY+[`VZCQN[BMRKUI:[PY6JG)RZ[]NL"C2[?_"[&0 M.[U>V[XO:[%]J\"N6[].>+^X"+Z>N[^VV[^;BKX`?+H<G%U?"[,A?+TC?(9A MBJ.4.JL3S+.L:\'FF[5)Y;@P/,"I^[O6^Q=`K+TQZ,#V*Z;'*\&*2++!N\1S M>[ZTE[XK:\0R[,%@F\"9>I[-V['=N[GXV[G-JL++*\8;JWZ<"KU%W(J^6[TU M_!=LW+0<6[,E#,41S,-3G+!WG+U7+,0D1<0;S,4*IKI)3`.!7+Y,3+\XO*%K MF[_AJ\;$RLC^.\1P?,AR3+U)6[F7W,CRZ[1D_,`F/*8>VL-4W*BBS,`8K,FF MR\D-.L>?#,@W#*:2S+FI_,=6:\>97,B;+,"=?,1T#,KMVLH7O+VO:\I\O,,' MJ\JVW,)9_,*RK*)Q-<,@;,S]BLQ!#,FX'+6ZC+R\C+C<_,C.B[+4+,RS3,"1 M8<#;^L'OV\N8/+Q.O,<ZO*Q4"\WR7,Z#G+L9',#J.\Q=_,Y?O,C\/+]/BZRQ M>\*4G,+*&\J_S!%:W+4#C<`&?<M02[!GO,LC&\W-:X*&K,[63,RUO,\8K=#> MR]!HO+.K',81_2SI'-#KW,$$3</:O+0UP'R.W,\)+;AFK-(<?;AOF],O'=(R M/=+9I<AU3-3TO,Q/?,\Y2[OB2ZQ,_=%O',LB780D'<_/6M5MG*3$:\]-.KOZ M^]#@ZM5X[,;#:-1;+-")C,1+K=.CG,>9*];H2M:5;-:VBM:"C-`@'<Q'K=5) M#=<W[;=\O=,(7<I//=8HG,9ZC=-R[<I8_+Q8'=A>R!%*7=@'>MAS37J*;=>3 MO-)5V]61G<R$+-$QW=8S7='P#,8-4=K=W--E#,'.G,_C/-2P;<XF.-&HFUT% MG*^MG<2<+=E-[-2@'<Y2W-&]/-RF[<])Q=LQKS-6X?=(^3=OX+-66?-:Y MS=.[G=H4+=U>;--&RMRQ_=F?V,S97=84#-E%#=BJC=1O7<SEW=V)K<?I#=5X M[=#M;=CV[;3?K<%9?=G(,M[97-_6/=NG',5^K-RD_=Z5'=^"/=\E_>!-3<+Y MS=@-[=C]O=G_O81_'>'AC=F$C>`7'LG@O-'B[.!#+;JQ'>*[*^$$GMGE[>+F MC-XYK.&BK<]=;>/>?=4Q/N(-5.++[>/W7=<9?M>-S=(U#N%!WML43MVV*Y0G M_LT:#=0K+M13;N0`'J_@#>7:8^!2;KE4/L;XG>-*ON%,7N0O#=V(G$#M#-P% MO=1<3M=AG>2A'=2U2^9UKM;3+."6;98%7M,'SN95GM$+C<I9ON=4W>?/Y^6` M+N."/MVN7>9?3:=(CN9YONA3?=:.#HFTQ]9"3N.&;N:9GLLJGMQ:SN=M_N71 M'>:$/N94;<5'?N>:CMP-ONJS[N31&^AM.N3TO=RT3LIGCNI8KNJ,?M;##N)` MWNN2_NNDWM7+[MG%GN+'GNO)OM?3_NAK#=^C3N32GN#,K-]+/MI#O>V@3ME/ M_NK`7N'G+NZ+G>8[?MM3CNZ,J[O._NW!'NZ'CM(_K>C(WNG:WNJ1+N24+MSV MWL#&C>>X_LST3N8)_\HDY>9N#>?X>L#!O=01K\P8?NNICNT"C],;?]HP7?!@ M/N@6K?'P?MP?[_`L7N\$#]#/GJUB7NDC[\V(GM(`#_+;/?#2K.[Y?O+1_N[] M?MT+WL<NK^MG':.(W>4N;/+L?O!+O98\C>/&OO-)G^TX3?5^W>QQO-I17NE< M3^RG;NU8;]LO3^9CS^S=+N)"#^Y#O?;47O97?O;:_=B&+??<'NK>_O;['O<K MS_`MC_9*O]=ZG^[X_O7R3>)_/^6'?[:V?O4,GO4AG_>\KO@3SOCN[OB![_'7 M3OD]O_4QC_D$+O6:O1?ET_1V7L^"__F$K_66__/H#/5O#A%QCO%S?OH-$2%% MK^#J'=7L[<.&S?M6_?0R;_`U+]S$?^F0S_J>;_?!W]*[/_K5G/DHG_&Z7P/+ MG]9@[?R2C_2O7_F;O?U][?1\[_;L/O133OZJ/_>1;_:3'_ZA/_S4/^"3GOQ+ M#73G7>UU'_]WW^&H3__I-J]7_68<W)MR`K#J\;]$Y_^B7WE+@%VO[:V[VJ?^ MR!P$)'OOK_^!O_\G_#;;!61[YV\"5KQV)^O.V@=T?]X/_FU`![C<3N#>`WJD M3]!50*KF`C%=!F2`*S"O`<"&4`/OW5D1=7YO\UG`SO?]:AL'E'Y4J?[Y.IH7 MZRI=#U1X':\(KC<=V`$#H!*<>?;J]C5!X?8$:50*U(!&D`5VM2XXV6;?\3MY MIN\!$D$5&`:I(!(D@R2/XH$]6)?RLA\<Q'G^#KL!/S>H!F7?#^Q[Z>\`#L'> M-^YTG)X3?U;0#SXW5U?[TN!RDT[-2SWY(]I4E[132^!.NFDA@2?Q-&5J0$*` M`<B&*V&DK\21/-)(`4EWH@B,I":2&$Y22EI)+>DE%8&8%`-F4DWB%&@`)]D` M+$*5:H!/`DKWQA-6CE!(E(P24BIN49`-3D'/196L$E8*AI^0&'8E8PB62!Y9 M8G</B@Y$J`G5MX2A#2E,^R@V34+LY)YN4T&*3]UI/FU"W^1SO.%!PX!?$`>V M0?XF_-SAY2N`,E`0)A=[2`CCW:8+>/./'Q:_$!CT`F'C>SD"D?E!010'!I<A MAZN'">$;*D)@AOXHH#Z\*0F1^]G`>)A4DH`;B`^8I#1IB)80DMH%'8@,*85$ MB(',L!F.()"0(!KB((0=&[*3.F%$?(>K;^$U)8\($M.`2/Q,J#`TG40YD!+? MP$K4#'3`)9*$)A$&9&(-P1%H`QKR!9R(`G6B%-R#])"*9<3R!P)A(#Z$=A=Q MG6S%]M?]K*(RQ(H/42O>Q'MH_\#B0=R':[$?LCS7=P1MXE1DBTL06(5%L1`7 M3=T-U'D-D`]*18DX$!T6[1N!CK`=1D1(J!"]H%ELB&AQS2E&OL`8-:(/7(2' M<0[:OHNW!;OA8ER#D'&_I46[J!>N(/+KC(7-&U9&KE@5DV%H+'<\KMBH1M.( M!O%?:OR,<K'U03_!.!E+XT3D"$#0(`I!N,A.?!9K+(NN41XZ1,DH&S6'<22+ MB.^?Q<!?EQB;(TX!C<HQ,IJ[%B0,.<9OU%4B4#/.0(RH.9A>9T..##$[BD;F MR!W+(U[$@K2L!(I%]Y@;GU]@S(IVT3L6QL3W%4G57FP(FN/Q+40KIQYAX\,C MCC)`0$H\X`@(+>);)(_L1$%RO/0(&',@?I2*"?(]ZKOA""$S9'V\BNMQ.R+$ M`$D;HYYMS(\2$AE2R']W'T<CADR1SFWB,<(1^-MP'WE[D=BQ0LY#%SD9/>1^ ME([]L66=2`S)_L[C1GR,!5+-B4@$622)6XQ$;9EQ\5V_W)<?FV1SDVV%4-X= MPH"H.:SDBR.`;=$_/LCYR$Z\Y(U;@#IR.2[)#FDF?YSQFXY,L`[FQSN()?UA MPY-_>*\[TDD85Q`=)(<DDS*`3EJ]LQ@B8V-[9"=[$DSFQ?@(O+HC542/!#)- M:D=#.2+9R:.,CK`L/$K)\0@H+V7S0Y*2LE`>R`[I*4G7'ZR((Y!3\D5+F2-9 MI(7DD=81BVC(("@?5Z6L_)"$TD"F/019*A=DR3N#)A(UYL=>.2$CI:O<D>RQ M4M[*'XD9@67MJY'"$D,22Q5I+/6@J-R5I+)$-L(AV2-7(W3\E,DQ5.K*PF<K MO:61O(PR,DI:O^IX*#='J[26XQ+V=4=SZ23/V:G4E-9/50)(8_,NCQZR7).` MDEY>R0#G+!$CMXR5`G/_T;TD.>^P9<"<E<*Q5NY+=XDK7Z.2I)0(,F$.0`G8 M)U/E?Q2&,^`YGLL!F?..I9K$F!`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`@H1CU3Z?)*GLWEZLN'I0?FG[HR;KS-_1E"\\D'M9:9$G9U3 M=8),#-HS->@#]9Y!LX0R4%/Y0D7H5M.9.O1;MD[?%SJIY_>DH5^28_K0P29# M=P(2/9,#%'\:T1S*0DTHR:N=J?-VUDP8:3QU8XM,EI.3BQ+,)2K>V.;YY*+2 M4XIR4+5XBB;HC-2,Y1/[V45KQ#TWZ-6DHF(T?I)1.BA'O6:;K'5ATXX"1"KZ M1\W?T]RC4_)&MLU"FA/_IB#E>804@1K`)JH0&&EK=*`I=(H:T)U@2>]FVER4 MU\R"ELU.:CFKY0U5H=^3E)I.+!I#M2C(%)0V-).NT3F:*)5HD'2>GK''U-%3 MJDD[*)H!GJ@RBY+07PHV0:4:O:.;5)>&3P893%OI,)V6>-!U%M%9ZC6AJ0NE MB/ARDKK2G6!-T^@C!7U4U)J.T5L:2LVH]NRFL72:(E-?*D[U*#G=.7&42E93 MT&D(URD;+1/7\W^:TZF)3J/H-\63R12?NE%U.3_WZ>3LIW_QF`[2@-I.;:G: MQ*6W42$@U$#*2ZEIVVRH?TY^WC^#FC(UI@+TIQ75GLY1CQH!"2(BU9<@DZ3" M0V/Z3^NBUU2I79$_/M01"D1G`$R]I*94EHK4ERI)\R$EM:GT5$ONU+9Y4SUI M#WVG3'2;,JT=ZA@=:4A=J!U483%5M'E49^H/?9X)0:H*T::*2=4I5&6C6G5D M]L\WJD_-YZ$,JQD4I.K4KVH7T>K`=*=6U;=QS3XZ&=VJPDRH+54,6D>[NC$S M*B*5G!"2KWY4O/I4(6DR%:PEU2O&U;"'5?]"XQRB6?(?&M:H^EAI9_`4ID!4 M*P15R0I.#VMEI:K8%(8F4$K*O%JH-RVLG36JYLXK>EF=:685H(1UK4Y6L`H_ M'2HHO:JYE(454Z<J6U,K;<VG%92CDDG=:EK3:3UEJU*1N%I1$$I!"5P\5:1[ M]8D.5HK:6P%J5)6NB15O(M4R:E8K)0W`KBN5MWK5V=I6P6M,_:3PL9QV5^+X M73?KG72I==6\.DV9>EN3*@FE`6C4N`I5Y%I7\^B;W*Y\5)[V5_<Z^.#K7O6O M)A7`DL#,FE_5ZGCUK6T5P2K6^LI8<VN#C:T/UKJ"50E+7],K;HVH'(G`TD6] M>E95*0\-K2?5<X;875I=#6R)[:EN4:FNV-W:58\K>4VN)G:L$M2-NEX#*RQU ML#86PB;76NI7%>QS+71U]<=BV""K8=LJD3VD"A:P#E<E2UTSK(OUKD]6NR[6 M``M=SRJ59:FHM<D.61@K)F7LEQ6O3/;*LE>,"F6W;")%LGMUHH+9%DMBO2N; M[;`;\K[*632[7V]L7;VS]U*T:E,]*V*1)ZP\JX"VJE)8S4=H6:R5K;-KELSJ M1=):5$OI^J2SO'&O5MI5VEI'JXS=M)#53A982!M80:V.U:C4492N2E%1:+\H MP%RUIA:T/DY$>F1)J"=LM:\2C$+(6SM046V<I*O6D=?2V)SZ:#/MH12V+72< MNEDINRJ1[7+MHO8QU[[:?>ELIRIK;::>UM;RS+M:9=,LJ263U7:KRMI?^5=5 M+;7=ME!TR?99(3L9PZU8';>L--L"4?.6;KOMNA6S[?:S7MMLZE.5*KV=KG.V MV";/8*MO06B\';3SMN!"6Q")0Y/IO\VN@3;%4M*'&UYK[+U5L[M6X3)/!8M2 M$X*E>[:G%=,.W&.[6IDK60VN/1;<EMR%FRM1:0[]N-;6Y.[87U=KY^W*#;D" M]]!62I@K;N/@R2VH*;?9PE9[RUGQ;;"MK45VV9I;85@C<.V_1)G@%NE.6`]K M7^?MT`VXWM;8[ERIJV47[9NUM5>7SQ9=C!MU)2U$M8M-U]%FW9&[=<TN36VL M:7?87MJ<JVO++C#EMS'6ULI72#EWUZ[.)8XU8.]B2A3+<3UGX'6Z<]-],EW! MRVFQ+<*%NXPWU!Y/5PMUFVWD/;7E5K@VVPM+=-_KM]V\P-6YSE4!&VPY+]:] MN)^7VG)8]'H:@Z[J1;SM<W0BC&>K;+TNLZ6V.?:$2E._6W>;;>[=MX*VW]K: MWQM-B2COG;9,E_A>TUD;99>NQU6^./?X5E[<ZW8_+-J%OOIU[*;>Y%M]JR[< MQ;Y`%O5J7<"K?#>NF^VX/!#V0E"LF03O;O#-N_/VS%I<[3M^=VV6C;B%=^+* M7V(K?1<GT[V_S3+SNEZFNV?GK^>MO^`VT8I/6CMZN^S.+<#\5_RR7<"K@)FO MTM6\U!8"]UT)_'=W;04FM\T7`Q-@]=MP.ZAYZKX5-J*>8+7+@7MO!@Z]/!;8 M'MM8JSY1*-U%OAZ7!AO<3OMX<VD-T,$L5V9:U&`+A,VOUT6_/Y@$NUR'6X05 M)=5-P6@7"$??%HR#&T(3MJU/F-$V2H9@-%@P_9W`&)$+HULWF6#/[\<4PTJX MEVI%--QK!;`,1HAL6.[:X/ZK>.-PLH6K7K?F/D\NK'"G\!?NP&*1#\/@5"N" M!;$7/L!@.!!7#HV+A[/PU]W"BS@-#V'N:(B7:0C-OZZT$A?7\/N'77!.B\1M M.`0/8$T,<K,O(@;$8H$4Q]SE>XG-L`SEPC?7%(]:!)R**\?*-<*.&*7"8DD\ M5&7C+G:_$C<3VV)>S%]]\3`&QIBX9?YB.;Q[J?#T_<0;`P5KX3U<.<*N`9[% MB;@6PP"NRWIK8R&NQL38SQKC;3R(\9`>[H9<V!I'X$Z,?--Q.=:;;W@?<N'+ MJWN-KS/VO_/X';.O;PP#Z'$-;L;L^!GG8V3LBH5Q/P['['8<^V.9ZVOA,>FE MQ)7#](K=4^R)N?#J5;2Y^`Q'9(1L="&R6I+&"Y8:?^1#C(U1\6O;R`3Y"&MD MB,&1R6XM9LDI.2._8KX@D:_QB*7%)QDF6^*#*WPAL4[>Q.HV(.-CFJR/U6L\ M#L-$F23?Y&R<DR_R`C:R#1C.0N2?7(HY,45NQTG9$C?7&/R0X7!6!LJ=MR17 MY*],>QMQZSW*BI@JKV(_?)4%,EE>Q;@XS_KDFKR.V_)05LL]ER&[X:XLCUL3 M6#Z]0KD.^^6RC(7E,C6..4K9T(YEQ+R3>7!//LS[=P,'YJ7)$!CS'2[,M-(G M1^8Y?(\%LV4FS&58)<_DS[R69?%2-LF?F#3G95;,D^'O8=;`G'DRR\O*_(&W M,B$>Q7<Q,5/>NUR;?ZZ@.\>ID3:WY.TKG$/Q!<;-OI$9VV/9'#LK\Q5.NK9W MZ3KGX8R34_-SGKJ&&1U31NK,E*US4:9IR%D*FV;%C)63\V4.S3)9&)OGJAR4 M[;)GOLYX-C,?9O%LE<5R>8;/^+<@*^/MK)NEK5M>SW#9"6?GX,R?E7-D=<^4 MN4"?9^PLG[4S0N#.J!F1C&&(&X!%,5JNQ0^Z/S_=H9RAM;)OIKE1>3]WZ+\\ MD>VS6Q[1H+G-1N=OS'L,M*@]S6.Y12_HKNN(;^\GEM'L.2S#:*R,HP,T9HZ8 MD+A'KV:V;*(Y-"/^K\?Y0I]D(?UN__%R1M"S&?TT9L?[F!UT+*[/._I$WV(! MW:`)]`O2T(DW07_I*8UWRZR(OM+MN4@+YC$]H_.S:!;&;#I'`V8HW9S9R9;^ MT3F3&L?ITHRER;.6_LZLC2\CY3T]I,?S;E[3=QHZ.V+@W`D1B3J6S'0Z(#IJ M0,U=E?2-?M2Q.5+GR4EMG%<T;I8!F!H@:VH#RJG)]/LUTT%Z(0=AJUF,(;)^ M;-,5.DD+:D7\JN5TB<[2'%I5Q^4NW:BO(YB.O5&Z5OMH]#R@>[6P+M1]^E"+ M:5W-I8&TGJ;+D%I-BVFGS$S+]*2%T]`Z4TOK8$VM6_&;%M'96E1OZSKM(V'U M4Y;5#K@ORP#PFZ9Q]9HNOV;YY#'J8(A(V+6.]M,<&EXC:4]MJ>LUA([1^EI1 MG^59C:'M]9P>UY*Z3()D7:RP?_7ZW=0-VT//W%]+L)>TP;[5^/I=+VR-'"C_ M-8\&P(17/P?IS2RNW;68!MG5^E1?:Q%-LI\TPH;8'3LF%VMZS2H==@E>PS7; M5`=C$0V;2W;&%M.]>6([Y'0]J'NVRS;9P3IH-^05-:_;H:OQV"<:/\=J?EVQ M;S1];M<_.UA+[7--M8DVK;[:]UI9:^V-/9,IIHN>O/Z90V]M"]RUI;)7MIEG MVXNF;<$,M\UURM[9D+ANVVJ;G+7KM-X>U@S:66OGOXVLL;;8]ML3VI#&9\%- MH`EWDRZ^!_IEDVJ%D+C/*T:>V4X[:]KL)5P/#]"1)M:\FF93SL(=MN=V@A[= MCYL5VV:*[;5K,>I.JX;;=$?IU_U6]_6B#M%Y&TV7[@U-MQ.UBJ[1TOEWQNUH MR[M/M^^FT0.[=9_D&:"[#S;2]MO'VTVGY_W,O*'V4*[>G1IX?V/L/;P9+N?& MV=S;;GOKZ9V[K7?O)MO">(&2;N?=MQ.V^D[=NYIQ]^KW#;MW=YB>W=SX<H?N MS$V_:VBREMU^.W^'[&^=MU4UD7[>[IM9V^[D[;;[<A7ET[&[>,]N!2ZPY37N MIL8/?'UC[,.=P"DUEVW@2#F#I^X#WKXAM@BOE[57>^-F.MJ]6ZX:GM_=NC6C M:@P>KH]V"9_<+%Q\K^ZA#<(5<0[?VW59<G?N-JJSDW'>KN&1&X&;\!CNF%WS MX$;B+_J&#_$?#K@7=YX>W!>;;W-P$QZP`_<5;]Q9/(@K<1S>Q:VX\,S;83Q: MC_$I7L:[<;!<X6E<6Z]Q\-W&I_;MYHPKO&4G<2D.OE$V"$;7/=QUZ_$HOL5Q MN!_?X?LXCYOOT^W'9?C*SMN#'&U+<+]]R#\TZP[DRSN2R^U)[KX;>1.?X8-; MDQ/O^TW)T3?U-MI[O)!/\0_LR,^NZ$;EA!R`NV]6_LD?.0:'Y9*<E,]R4YZW MP3;[5N7@>VW_\;9-O7WY!I?E)ER8(W)VAL?[=71:Y+-;F5MR'KZ?$6O%%>-\ MO%=;<\O-M56X,]_F.%6-9W.:#<Z-Z@`GW]2XG%M:.3[.,[<Z-YW+'#Q_\\H= MSMDY,-?F]-R<WVTCGL[S^3KWV?><G/OSQFNM7;D['^B2=Y/K<HC]SD_L/A?9 MZ;P/&VID/KDUJ\S>W\;8HK=P(=R+(;)&%]^M_.UJY](*Q,5Y0'?GT=N.,_!J MWKR/.2=GZ"F]FPNYIIW16[H6I^A#G*17<3>^+;^Q3M?@-_VE5_28SK:]>=7& M6#8=FY_TFL[+TWE2-^DX'6?_=/@=K]_X-P_5-GRI>W0!_M`)>#K'ZJD\JFMS MKCZ^,7=&-^`37:CG=`KNQ<]X.D?K_UNM2W6V;L8QZUN'YG6ZO5YT^4W.H7@N M!]9YO5O'\T"MO#\Q?L7K"?NP9^^9?L%'NE]7Z(`]L0OV:9[(O_EC'^61G:%/ M=J%=V8_Z7XCC`%VLD_,Z?L[-ND<'[5E=M+MSTE[45SHDGK$D.J@O](K.VLLZ M1C_MB)VAU_:0;GU'>VZG[4U]I(MR[_W"R;DGI]).G$`3ILU=W-WY<2_H(EVY M#W<7/HF]\G*7V$O;*'OV>;;16;4X]NA*FP$W\^V.RR'[PZ[HX=U"D_??GM/3 M.R"OYN4=LY_W]A[<E7M\)^[5O2]S=QU.V;5[8<=8]YVZ=W3K3LN1.RA7[L9< MMF?VBB[,>;OW'>D)7JFK]HS>X&NY0:?P[%VJ5W@#;\NUL[NMW[]\PD/D#U^[ M03=?S]PDGMN&>+G>JU-\7_W=KIT:NWB`Z])G^Q"?\11:IEOP\?[?`62&;_$( M'?.J]QZ/XZ\Y5&?Q-+O(<_.NCLX]?)"OQV$=R:/X)P]\\;:,E^AQW<;C;(J[ MY&W[B3?&7+Z>AW8I#^8_=UNWZQX>RT=P+=_BS7Q==ZTR_JG;<Q'OE7DN55_@ MWA@WVWD07^,7_(TGZL/\CKLS/2_GQSR;3_*`?K!7:L^^Y_WWFO?S6S[1]W=Y MSN@+?6HG\R,>T#MX*(SH?WRG7^QYWIG'W=@NX3%]G2?KFWX:>WBP'LM-?5\> M]2GZS<M;&<_J__I\W_*HWL)']Q9?Z\W[S>;U]1W8,_?\CI1A_4[7WU]^Q,/U M1W_K6SQ=Y^D&\QL?WF$_X%_]LU?I.W[0B_IE;[\AO;/WX(]8QE]V_%[MB_UF MS^Z4GLB/>P'?JNL\$^?P%W[$KWN.WNY?_;G?R^K>TZ/X=P_=>_N^U_=@GM^K M['CO[@&^O`_V21ZU1_E#C^)WNZ[W]V!>X;=ZAA_Q$7[#-_AUOK8K^@^^G]MO M=R>@Y5X1>WS^SMF9N;9G]-.=WG_W.E_)2WZZQ^2?>.27]#GOZHM]RT?WA!WF MBX64[]T3\HA_[@-_UR=YG@_RZ[W-!_<V>N=C_%</]*V\AR?Z"G7E,WV+#^:A M?EZMSDI_KW_Q%A_@5;[/K_,%OM\_>`*]@C]^U/_ZKS[L!WV(/^(C_)&G^&T? MZ0?O)$SMC;[(E^8N7Z[R>)UOA9=^L<?[.'_1%_:T1.4A]\3W]L&0\,M]$:SX MZ[[4QXB-OXA#]&X8^4G]VT?\[;#RQWIHCST9?^;P^V)1\Q_[TG[;$:+H!^JE M'NZ;_L^O]=TZY6?]9O_J=^?3?^=-_-9/_+#?\M-\U;\/:3\*;]:W/_.#XM@? M9EWR1!K^(/WAC_U.F);4?+=O]KC_D"S^`=S\0;]8J/Z@GMW1=.Z8EBS]PL?\ MLK'[3W]++?X=/]J'_,<8N^-[F%_^B;_(1<WM/_G#>Z$O_*.Q^>_(J]_^J_\A MS_[3O^XW].`?]^?_;7[(7O`7_@V`$-SS]^M%?YJ>\L?IU7_>W^$'_4&`5)\` MJ/_]?Y<>[X?^78`$(.F7[.5_O9Y\MP#6?[G>_,?V@8#6W_%'CF5_/1WU!X[= M?\9?Q/`"[G]NUO:7_W%[*YX&&/I5#M>>CF?556UI"0[8YTV`!^!!Q@)&>RZ@ M$>C^W6#/6!!8`=Z`*:`,J`22?($?Q]<H.8$PX/:%!4I^7MWK-P7.?``@$6@! M]H!,GFF7_\U[/1_^U_NA9%Q@DY<:I25H8-'W^.V`(UD;:`:N@74@!OC]B8%G MX!.(!TI\MMX(6`1J?)->SM<RI26`H*_WO3%_E0,AF/<I&#;@'Q@%(H+C'Q#8 M"$Z"E^`1V/DE@8F@"+@(1G\.GPFX_(&"F"`,X`A6@>&>%VCU%7]:X##D!VZ` MJ^#[YXFE)4T?/Z<*EH*UX.3W!E8.75\:&`.F)6J?T[<+B@:E8#!H"PZ#O:`< M>/[1@>X=,7<%5@YNW^X7`.9_@)]X=_(-?M!@*5@-\G^'8#:8!59G:<DV^-X] M@S!`-!@&"H(6X(8G]CV`1:`YF`%.@WB@..@,/D\M!##Q#<Y^?$'AQYHY@*K> M+F@/+H%TV-)4#^J#\9L!*`#^@WJ@!(@.KGX(80=8!I9^:V`-@0E*A*T?FN</ M.G\Y(#R(_E&$=B!$N!%BA$/@0A@1NGF<7UE%_O$%(*&"QP>.A.!>)+@11H"! MX"<H_,4`DMXC:`72@WP!3*@(-G<'($U8"6*#/N$]"/_EA*_@#M@73(1DW<:7 M"OJ#(2"Y9_==?RY$40@5'H5"X2P8%5:$<-[K1Q4"A)V90'@5=H0?8$0H!*J$ M(N%&2`9Z>08A0S@6IGX:H5%X%J9Z(9E6N!9>?BNA6?@3]G]1PD1X%BJ%+J%; MJ!?>A=T@7C;"I74ZX%0H\`F#C*!@R.>1A3)A3W@8'H.)81QX]JF!&^%CJ`M& MAA.A9=@%^H.2H>PW%"J&)=Z9EQ7Z@YV@4S@'3H6GX/H7&):&[-YI>/Q!""UA M8T<:3H2I(3?X#*IF@V&6UQ9.A3>?:G@;QH),H/]5#^:"FV%B"!P&A/+2<`@8 M_H83(7'H!AJ'S:%4^!H>AUUA<CB8C7X/85BX$2:#D^$OF)N!A6EA1,@=>H96 M(6$D_ZV#_6!B.!ZR@N#@=T@%^H8X(8"&^M&%9:%1*`\*>G(.4.@.[H'UX51X M'\9X6N%^J!`VA@?A?QCJ68+R86[([/6'KZ$ZN/:-@L(?BI8`9H1UX0XH(:YF M*1R`N`M>B`JB`D@@KGX<XN^'I[E^&Z(^2,)I=2`B00C\D8B,8(CHZ'F(/*$` MZ"(F4;8?BXC[S8CU%H7((,J`."(9)AK.>J\?D[88LH45(E0H)(:&LEX/MB&F MA$3BCGA%D(38W@^(#1Z)*EY(^"&N@53BBX>\97OY8?]'J'6(.N*5B/Y]B2)B M!2<E>HDZH2<8(X*(-2$J:*-=$6FB:;@,&HD-H"C(#LJ(<6)K."<>?V0BDE@2 MHEPG89]8)3*&:R*66">BAW'AAI@G>GV4H85X*#Z(=R*(V!3JB8VBD5@"(HHH MU14Q*3**WB&H)AWRB)NB+]@*>HI8(9"X(<Z%TF"1R"?NA86@X.<EHHKGH)AH M(;**-N%F=`V^BIC@L5;[P7@'XI0(*[Z#JB*/2"NZB4O7%?$K\H>RHI'X%O*# MB6*+>"P.B(7BF#@LPH=!8F?('N*#Y9I#B!;:B!&BM2@+(E]7A&;X'-Z(WF)P MJ'B%BY\BNE@5@HL14KH8L26$,6&T:"&&@ICBQW1%L(:<XJA8&XZ#..&ZEBON MBWF8;-@B4H?,64!T1?2&MF&_2#".:O70P;@<*HRYHG-X!XZ)"Z,0IQ4YC)J@ M26@)+DOPXDY([%F(QN!E>".NA]]B$YBSG8>08GHX,N:*(6-QN#*NBR8CQZ@M MPH69(K\$,PJ'-B-XR"T>@&9;QZ@F?HQ&HH%X)G:#/2,8""PZB3GC>PB5[7W= MH.,V)-*'R>+QYS3ZB5%B"W@23HV#8I,8-<J`6*.6^"<"75>CB4@8[H92HXHX M(EJ$CH.,1BD!CV?@POGYJ(R`5)LJ+4&'<J+A);Q/C#F@W-E)L8\A7-YJ- M-2+:B/OMC6+>T;@UMA"$HSZW+0J.PA_MEC5"C72CU`@E>H#A(?KG..:(5F+D MR#5.CMEAY:@W,HF08]NX.;Z-NV#X9C0BBYHCXM@F6H-=8M.X*(J*[:'IB!T& M>AIBVO@Z*H.5HM2X.B:,<./MV!VV@K(CAK@B,HX'8/`()F:.HJ/J2#JFC:$B M[N@=]F\TXH^H)#*/F"#TZ#46@#NC`&@]KHU:8^HH0>F,Q*/V^"S&B\EC$+4K M)HE56NDX/GJ,?J/4N"S:B2ICX[@^_HSM(]<X+?*.ZF/U>#_6@`+CX%@N(H>Q M$^)X[^&/:>/_6!T&D$2<TL@OPHT&9,&8)PF0[B(5-Q^FBDBC!'D^?HU<&39H M03Z-%.3AF$#.C,QBS;A!OHAS8_DX0D:/Z&-R5T!6C^)BWE@WXHNP(SXX`]"+ M*6.S.#C"D,XC\$A#(H:#8\4XQZ6-".,"63KZD.U<\1A$!HQ,XS-(3/F,<F+N MR#4>D4:=!DDRFHM>H4!%`\Z#<.,4"4`:C!)5!*E%'I!<I!6I0&*1I>,7Z4`: M4(ACRS@N-HYF),-X,7:1I>+T.#BVD19CVJA&NI!2HX!(/M:/B*.#R$,VCGHD M^_@4YI$19"!)/PZ27*/0:#5:@@T=5]4]EH^-Y+C%%_:/PE\DZ:2ACI`DX,@K M#HW/H"5I^.V1B&0+X4FJ;JWB3?CZC9(G(IVW!HZ2!6'VN/JADF-CL"A*:I); M(B>)$\*2NJ$LN50)CV?C:,@(XI(+H@?)2@Z/OB3N!TS"B)FDN_C1G8[0(B3) M.2Z.Q60E"3IVD-[C,CD[THSVXNW"1%**WN%45R9NDHLD-NA-DI#((Q^I36J+ MDV02>4OZCN0ANSA.HI`8Y&UV$BI7$V(Y&4HJ,<NC,=D\_H[M(3W)2P:.T>0! MZ$\>CWWC/3E0?I.U9#C9_QV4Y&1!Z1K*@`PE/(D]AH^OY#[93IJ,$>7U2#FZ MDJND17DM#H49I8^80AYXO^3\V$1VD^]CO2A#B9(F)3?9"NIUX&-`*0#"E-MD MOMA/[H]D9$F)"=*4Z&0I:2NVCIUD`^E&_I(#I!!)5.Z4165.:4P*E77D4JE, M,I4_I%.9,0**C"14643.E"VD1[@#*G9<X1G9,/X%6:5VN%5:E2CB*AE6>HY0 M(6S'3(*2#Z4HN4-"AL8D#LE/XH-JY3490F:3UUU-&4."E%!D[6A,$I%E)?J7 M5_:4M6)?F%8"EJJD8-E7]HH+)1T955:2S6`4V5CNE'>D5IE6.I97Y2L96?J5 MD"5EJ4QBEH'E5EE9BI67I6<91Z:/OZ0A>5*^E'XD7%E)JI8N93_96HJ,K^5. MJ4@B@?,D;&E3TI6W)1Q%21Z`2E[AB$F:D\&EXJ@4OHD^WLUX+B:75Z1DV0T6 ME_^<X>@]0I<$70U9,U*7"=TAV58REV.D<_D,8I>.9.A(7-*2>*-E>?R!EU!> M,SE>NHOHY26I7MZ3[>4.!C_:D,)?F,=W49/EHWTY>'6.'"7ZIU\V4(\D<?E, M8I,KY6N#"?Z7#AWMR%@^E^SD1VD5-GH2I0_X+/V6`B"$B3DZE'NB#&AABI3Q MY"7'8!Z8NZ-1B?MMF'*C/;E=DIAWXT0I4ZY^*";?&&#"ER&F4EE?>I0E(\YH M[/V3TB-J.6+2F%2D=7AC7I`JIJG("$Y[7>50.6*FE-9E-DEDZI4Y9'NX9(*0 M\^5UV5+NED/ADVE71IE*)EF96.Z`J0^0*6'BEI9@E\E!QHK395*)']I(_9^8 MV5"^F">FF<E9`I=:9LW'9;J9"^9W*5=>E#;F6TE;`I=WIH/)+FH_^61]V6?6 MF,LEH'E:JI`C)F(I9T*%,M^5J5*Z4BU$HSE!DIGYI<1H7FJ8BF9A>%Y:FJ3E M>0E:;IF,YFB)5GJ:!^9F66?BA&4?DSE75IFFIO9'8;*8NN5>^6#VEAHC-DCW M%9E-97TY6[J,NN:!N6NND<!EK-EDXH.W)LKX1P:;ON:GZ!=T8;CF8RD;+9OD M99`I1[9#T&846&U.E6!CU79M-IN9Y3ZT;7:7C)TZV0TMF],DI5D_DIN!YK.) M_*V5@N1#B6YBFQDDS/=N<INAI5@P;QJ;KJ6ZR0'6DQBF$[ELQIC>Y?/T;UJ; M"""..5)V>*G1P*E<+DT*9TPI9`9##:>JB6<J7A$GE/EHMDS+)H^Y1>9)&:>R M.0-*G'[F,]9Q'IHD)<3Y<;*;VF6&.7(VEV\F=[1L3IFR)O+U<L:&XF;"R0,2 MG%_@['A<+ETSY\(I+_6<+.>I.6ZR@2`GH<EP$IT69Y(I0RV;<::F62,@G3JG M3VFC,9TXYUG97XI%5*?/&3MEG0[GM*EO#IH]YM:9"7:=.B:U.786G6%G0+1L MZIF\YM>)<[*=P*;+>7:BG*ME=;9VIIMR)RB$<RZ6KF;-V0DMFYDFV?ATZIWP MICRI;;J"6J?:B7@&G;;DT$EXTIN@9D,`>'J<CR?:N7$:4)-GX>EA-DK+YJ>Y M:$J>O"#EZ7DZG9VGZ(ESBII7I[WI#4*>G^?@^6OBD:XGSOEZ7IK+IK"Y:GIB MM2?>B1#Y!0TAO\EFJISY(,UY*V*<P6?BR7$6GXRG0DE\]IZ3IG0I.O*>T>:7 MN0E::M`GP<E\>IG\)47I;2*?EB<8>7Q>GPAEATG-<9XHH?4Y8-Z52V?Y:7QB MGNHGV8EHFIT<8??I56I%O.?YB66FGT$AZTEZ$H4D)\+Y=_*?^J?@R7L^BL>F MW)E_RI]&)OP)<+:<N^=7B(#FFNKF5@AN<HEHYO*I<7J?[*<$2E@2BR(8[WF! MSI__IP;J:"J=KI0':GU>BB2H!7J">IPB:/,Y7"*2)FC_2?!MGWGA^ED/\9Y( M9@':@-:@`6BPB(/JGC0HS#ELHF8_J.:9!?F=)B=H^#CBE^?F6V9PBI^=76'' M>S:=`J@3BGTJF(VGS:F$7IB_I[]IA8:?TF;9&8%2H3[H%THU2I^UIKQIABZA MYB8,NH9JE-GGBDF#DJ$[XA3*@H*=E^<-RA?`G;`G[XF'8J!Z*&PH@])_$>@? M"H(FH7PH[7D=NJ#OI;NYB%ZA.V<'^HB.F<YG$XH;@J%H*/GT:M*@@6<9>HF> MH:V=T*F%=J)UZ"0*BF*ARB?Y:1[2G;%E=\9[HI[:)T;TBEJ?K68J*G"ZAZPH ME8E[XJ(CJ`Y*@XZ>5>@JVHOFFP8H,.J)PJ)RJ"S*BS*B;"7P*8P^H6'H^QF! MVIX3Y]&9(&*B<>C#"7]2HR&G_\5[TII4I12JH.6B,:?(.8Y&G87E)JJ,]H@E M9K\9`_H%[&B*F8EFF^)H/.IBBI=NJ#TZ7T&A_IT:JH_>EVVH(_J/[I?0I#:J M;F:)W",^*I`^DY'H``:/EIN5:#YJ?UZ<Y"="VHYVH>^HE(9O[IERIT5JY#&A M$JG'Z9'>HR#I0BJ20J0OJ$DZB)Z`VZ>@R(4JI,"G2\IA5HU@9CW:8!J=/Z>= M)GP"E;>H3'J1PJ1>J$\JCX:B6>C_22KVH*7H43J!%J4FIU)*C@ZANZA3.HQR MI`VH5,J,MILQ:5(H=?*<[:(-2G]VI>YGR6EV9HN^)U":D9*ET.A&&8MBG6#I M`^IL=J16YUJJ>J*E!"5&NGW!HW%I,LJ6TJ64:$H:D^:E!BE<2G#RI:=H0TI] MMJ5/Z>TI<R*FZ"@J:%B"GH0I&QJ1"J1T)E,ZEM*A6R->"H2NHYBIYJB9KJ00 MXD':F3Z?C&E:2I3:HD.GO^B5&J6):*?Y=*JF2&EF^BXFG;[H.FJ()J`':6LZ M:KZFMRD$VI'JIJDG9-J;OJ55*7`JEPJG@RG`&'"FIL-IMVF;[J05:$7:G-:; MR*D1&H7ZHZ2H;!JS)9\U*78ZF'*:NRD\FIUZIHC29KJ7CJ>DZ7:ZD;:='2EZ MVH2JI[0I,5J5NJ?Y*'@:G+ZFQFA2BHP&IE4I-XJ3BIU%XU3*GO:G@RDX2H_Z MH_YIVGE\!JBF*2JZ-2&A9F?B&%WZI5ZHA`K/;:62J(6:7=:=KBCE9IZJGI=C M0EJ2`I\A:HZ7C7J=<F>)^I$&I"0J0XJA.J2:6VQ*GJJH71Z!68+&J&ZI<\J6 MTJC[:#0JEJJ;/"I`.IFVJ!YGD#JA-J)$*F@:*6Z?1BJ`:9;>I3AJ?$J5,JDH M*9):H;JHM>+4&:66I2/JE0J=9JA5:C/JI5JG_2CQ:3SVI59J1GJFGJ(VJIEZ MDRJH[.>:&F&>J&)HBOJFYJ%?J9P*AQ:D*&H#JJ>*J"QJA4J`RJ=,ZIT*B.:I M@^J4JHQNCS]IEZJF:J59*E?*J*ZH0VJ%"JDZINKHCOJ!XJ8IZJ5*+=J<DRI) M&J@^JA\JZ!FJ"I=I*I1ZJBJ.;2KYN:H>J6(JJ4JFOGQFJA"JF)JC)QR;BG[> MJ+FJ9$JAJJDY**&ZJ-JJU6A.VJONJ:WJ+7JL`JJ5*K!:I!*KW2C%:3YBHWQJ MG>JG0JO_J>*YK)JH#NKT>7ANJY3JKZJJXI1G9H:*K<*I>BBXVN49IM_JN8JG M_I_J:H\ZC\:;M2K!&:^RJKMJO;J:FISW*D'*KHJC,Z2]:ISJI2#J=!IY/IT! MJY(:/P*I!FOKZ1<DK&&I_\FO-JS[)\2ZGL:=?BK%6H5:K%+J@,JD:JQE*,?: MF'ZJ\"K(6HJ*K`UJ.CE\NJHFJVR*LD*B+^IA^K*BJ;&JJJJ<,J`?J\!:I+:L M,^K-*HK"J_2I([I$=JP8*Y,:M)*H]NEQBK`>K17J?MJG,JGYJ<OJM%JK4*N] M.K5*HREJU#JCUJ+=J9FJM9*F8N0&RCI&I\KJU]J$AJTC*P')KR:H[^K:6JHB MK&SKH0JOSIZNZ<,:MQZB$:J!2J^2G\=D"=F$#I.])'^Z??:M)B;P";@"E(*K M,OI.OJ2.*I2ZN,ZD:BG!"GH^KHWJJ.JX)JH>J^)JJ.*MZB;E.I16JUAK`QI2 M5J[-JN/JJ:JM9N?H&JZFJG8G/CFKZGTKZRVJNHJJI6OK.KO*JZ=IU\JWNJMR MJ\G)4Z:M(F;JRKMRKG+G[]J@)JM#I_%:ESZIMBNYBK,JKL,KIRJZ!JN**ENJ MO-*L6*D7>KWJJO?GC;J]^JJL:X?ZO2*K^2K?.IK^K8#ITZJXGJ]N*%=YL?:A M8"7!Z;Z*K:AKY\JSDJ9U9<KJ4SZF3V?^NKPVKK:KSVJ9VJ_RZ\":N%JO3&M& M.E@*J$6KXIK`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``O/ MOJU,K`M+SEZM/ZK<F6I.L]GJ\3G17K.*Z$6KP7*H_*Q&R\ZFLMW0,,%LYJC4 M:>"5^T6QEZ9(*S9NL5NC2OLINK0\K&P$TW*G(RS,)]+>L9CL]G73?JD#V$X; MTW)'/BU-R\L^3R+MYBJ][D-%[=N:U/ZT"-%2&['.H!B14TO2'JQ2+?W*QP*U M!6<]&[ZB9E4M1NN:=K4;;2O*U6:UM"Q(FQJ)M*0L<8K4WIS^K"EYUK*UD="= G8"=(!R@!&E`D0!23@2HAF]P!+H,G(2C0!)T06AL%"K9M[4])MJ(3 ` end
From: john@nomad.phys.lsu.edu (John Woodward) Newsgroups: comp.sys.next.bugs Subject: Re: HELP -- Mathematica 1.2, NS 2.1, WriteNow, MALFUNCTION -- HELP [LONG!] Date: 15 Mar 1994 05:22:42 GMT Organization: Louisiana State University, Baton Rouge Distribution: usa Message-ID: <2m3gn2$29m0@te6000.otc.lsu.edu> References: <2m2jti$3vs@ulowell.uml.edu> Keywords: Mathematica NextStep WriteNow Malfunction In article <2m2jti$3vs@ulowell.uml.edu>, Mark Becker <mbecker@cs.uml.edu> wrote: > >Hello * > >Please forgive the wide initial cross-posting; I wanted to hit groups >that seemed relevent to the software problem I'm having. Note that the >Followup-To has been set to comp.sys.next.bugs. > >Everything below is relevent to an '030 mono cube with NS 2.1, >Mathematica 1.2, and WriteNow. This particular problem resulted in a >LOT of error messages which have been compressed, uuencoded and appended >to this message. > You might want to check ~/Library/Mathematica for your personal defaults. Try just renaming the directory to something else and Mathematica should create a new set of defaults for you (at least, in Mathematica 2.0 it will). john -- - john@nomad.phys.lsu.edu
Newsgroups: comp.sys.next.bugs From: putaala@core.oulu.fi (Miikka Putaala) Subject: Cirrus Logic VLB Driver??? Message-ID: <PUTAALA.94Mar15140234@core.oulu.fi> Sender: news@ousrvr.oulu.fi Organization: University of Oulu Date: Tue, 15 Mar 1994 12:02:33 GMT I have installed version NeXTstep version 3.2 on my i486DX2/66Mhz equipped with 3 SCSI disks, 20 megabytes of memory, Cirrus Logic VLB graphics adapter (with 1 megabytes of memory) and a mouse. The problem is, the Cirrus Driver included in the NeXTstep release couldn't initialize the adapter to the desired resolution (1024* 768*2 60 Hz). I would also like to use the adapter in 1024*768*8 mode but the driver has no options for that kind of operation. The default VGA driver works, but shows only 800x600 pixels. Is there a replacement Cirrus driver? Where? Pointers to Intel patches/drivers? Thanks. -- -- _____________________________________________________________ Miikka Putaala Tatunkuja 6 90540 Oulu Finland putaala@rieska.oulu.fi putaala@cc.oulu.fi Phone [home] 358 (81) 362430 Phone [elsewhere] (9400) 388400
Newsgroups: comp.sys.next.bugs,comp.sys.next.misc From: karl@khaos.com (Karl Hanzel) Subject: Chess.app, player preference settings don't work Message-ID: <1994Mar15.070153.1726@nugget.rmNUG.ORG> Organization: Rocky Mountain NeXT Users' Group Date: Tue, 15 Mar 1994 07:01:53 GMT I just now noticed that the player selections in the preferences panel for Chess.app are pretty well hosed up. I know... it's only demoware. I would like to be able to set it so that two humans can play. I select that configuration, hit <set> and <yes> to restart the game with the new setting, whereupon the computer *continues to play* as the black defender (offender). Subsequent attempts to re-select this config result in the board refusing to accept (or at least do anything useful with) further mouse button events It just beeps in response to button-downs. 'Ever notice this? Ever seen a fix? Ever seen the code? Dare i pester either of the authors (Glenn Reid, Richard Williamson) with petitions for a fix? I might be able to diagnose & do something with the code myself. Did i loose the code?... i've migrated from 2.1 to 3.0, 3.1 and now 3.2. Was it ever there? Thanks for your useful suggestions via email. pax, Karl *-----> Karl Hanzel Boulder, Colorado karl@khaos.com (NeXT Mail) or hanzel@comet.ucar.edu
From: premise@churchst.ccs.itd.umich.edu (Sean Michael Willson) Newsgroups: comp.sys.next.bugs Subject: Workspace Manager Crash!!!! Date: 16 Mar 1994 02:22:10 GMT Organization: University of Michigan Distribution: world Message-ID: <2m5qgi$72i@lastactionhero.rs.itd.umich.edu> Hello, I am a student at the University of Michigan Ann Arbor. We have a small NeXT lab on central campus that has 19 next motorollas, running NeXTSTEP version 3.0. Whenever I log on to one of these machines (It happens on all of them) I can do work for a while but then the workspace manager crashes. I cannot figure out why. I have tried replacing my NeXT file (My sys admin told me to) but that did nothing but lose everything I did in customization. I sometimes can work for a few hours switching between apps and then it will crash. Other times I will log in, open mail, then when I click on the wrokspace manager CRASH. It becomming more and more frequent and effecting my work. When it crashes it gives a message the the "Workspace manager has crashed. PLease exit and save changes if possible" or something like that. Is anyone can help at All I would owe them forever. Please I love NeXT's and I have work to do. Please help ASAP. This is the THIRD time I posted this. I have only gotten one reply. Help! Sean Willson premise@churchst.ccs.itd.umich.edu P.S. Problem Still not solved!!!!! -- _________________________________________________________________________ |Sean Michael Willson | E-Mail | |Co-President of wwiNc. | premise@churchst.ccs.itd.umich.edu | | NeXT Mail Welcome | | "Do not believe in miracles. Rely on them." | |________________________________________________________________________|
From: jweiss@casbah.acns.nwu.edu (Jerry Weiss) Newsgroups: comp.sys.next.bugs Subject: Re: Workspace Manager Crash!!!! Date: 16 Mar 1994 05:11:16 GMT Organization: Northwestern University, Evanston IL USA Message-ID: <2m64dk$a8s@news.acns.nwu.edu> References: <2m5qgi$72i@lastactionhero.rs.itd.umich.edu> In article <2m5qgi$72i@lastactionhero.rs.itd.umich.edu>, Sean Michael Willson <premise@churchst.ccs.tid.umich.edu> wrote: >Hello, > I am a student at the University of Michigan Ann Arbor. We have a >small NeXT lab on central campus that has 19 next motorollas, running >NeXTSTEP version 3.0. Whenever I log on to one of these machines (It >happens on all of them) I can do work for a while but then the workspace >manager crashes. I cannot figure out why. I have tried replacing my >NeXT file (My sys admin told me to) but that did nothing but lose >everything I did in customization. I sometimes can work for a few hours >switching between apps and then it will crash. Other times I will log in, >open mail, then when I click on the wrokspace manager CRASH. It becomming >more and more frequent and effecting my work. When it crashes it gives a >message the the "Workspace manager has crashed. PLease exit and save >changes if possible" or something like that. Is anyone can help at All I >would owe them forever. Please I love NeXT's and I have work to do. >Please help ASAP. This is the THIRD time I posted this. I have only >gotten one reply. Help! > May I suggest you check with your local DNS administrator similiar, as your reply address is not valid here. I don't have any real hardhitting advice. The following suggestions are some stray thoughts. Use at your own risk. I've not seen the problem myself. Try to look at /private/adm/messages right after the crash to see any console messages you might not have seen. Do a 'dread' on your defaults database and look for strange things. Otherwise I'd ask you system for a completely new blank acct. Then slowly copy over the files and test foreoccurance. If you can't do that, create a subdirectory and put everything (including all the .files into it. Then take the critical files out one by one. Reset all your preferences manually. I'll (blindly) guess that one of your .files is scrozzled, but determining which one can be difficult. Upgrading to 3.2 wouldn't hurt (much :-), but might be pricey for 19 or so machines. -- Jerry S. Weiss j-weiss@nwu.edu Dept. Medicine, Northwestern Univ. Medical School, Chicago, Illinois %SYSTEM-S-PHALOKTARG, Phasers Locked on Target, Ready to Fire
From: ambi@kaizen.com (Mike Amirault) Newsgroups: comp.sys.next.bugs,comp.sys.next.misc Subject: 3.2 "bug" fix! Date: 16 Mar 1994 08:04:26 GMT Organization: Boston University Message-ID: <2m6eia$8af@news.bu.edu> Here's something I've run into - I have a feeling that I'm not the only one... Since upgrading to 3.2, I had the following problems: - A lot more disk access. - General system slowdown. - Crashes when I tried to do a lot of things at once. - Really slow drag/drop. I found the reason why. I have an external 680MB drive that I have the OS on and an internal 105MB drive that I use as a swapdisk. I had the swapdisk configure for a lowat of 16MB and a hiwat of 105MB. It turns out that the upgrade changed the /etc/rc.swap file and changed the hiwat to 32MB. This caused the swapfile to fill up rather quick and then either crash or start swapping to the 680MB drive and then crash. Changing the hiwat value back to 105MB has fixed everything. I hope this helps someone else in the same boat! -- Mike Amirault Kaizen Solutions, Inc. ambi@kaizen.com (617) 527-7648
From: ambi@kaizen.com (Mike Amirault) Newsgroups: comp.sys.next.bugs Subject: 3.2 "bug" fix! Date: 16 Mar 1994 08:04:39 GMT Organization: Boston University Message-ID: <2m6ein$8am@news.bu.edu> Here's something I've run into - I have a feeling that I'm not the only one... Since upgrading to 3.2, I had the following problems: - A lot more disk access. - General system slowdown. - Crashes when I tried to do a lot of things at once. - Really slow drag/drop. I found the reason why. I have an external 680MB drive that I have the OS on and an internal 105MB drive that I use as a swapdisk. I had the swapdisk configure for a lowat of 16MB and a hiwat of 105MB. It turns out that the upgrade changed the /etc/rc.swap file and changed the hiwat to 32MB. This caused the swapfile to fill up rather quick and then either crash or start swapping to the 680MB drive and then crash. Changing the hiwat value back to 105MB has fixed everything. I hope this helps someone else in the same boat! -- Mike Amirault Kaizen Solutions, Inc. ambi@kaizen.com (617) 527-7648
Newsgroups: comp.sys.next.bugs From: Nitezki@NiDat.sub.org (Peter Nitezki) Subject: Re: Another FAX software question. Message-ID: <1994Mar15.184201.2502@nidat.sub.org> Sender: nitezki@nidat.sub.org Organization: private site of Peter Nitezki, Kraichtal, Germany References: <2lki24$b9@steffi.demon.co.uk> Date: Tue, 15 Mar 1994 18:42:01 GMT In article <2lki24$b9@steffi.demon.co.uk> robert@steffi.demon.co.uk (Robert Nicholson) writes: ...munch... > Another problem which most of you area ware of is that repeated > FAXing of the same document (not using redialing) > automatically removes the previous faxes cover page > information. > This has to do with the paradigm of FAXing as the other way of printing! The cover sheet information is simply not thought to be persistent in any way. > Are these problems being addressed? > Good question! IMHO, it would mean a complete overhaul of the archiving section of the FAX software. And a shift of paradigm, too! Not merely another kind of printing, but a sort of telecom. -- Peter Nitezki | Nitezki@NiDat.sub.org # Blessed art thou who knoweth Staarenbergstr. 44 | Tel.: +49 7251 62495 # not about the pleasure and D-76703 Kraichtal | Fax : +49 7251 69215 # delight of being hooked GERMANY | NeXTmail welcome !!! # up to the Net. Peter 1,3-5
Newsgroups: comp.sys.next.bugs,comp.sys.next.hardware From: mgb@moksha.uucp (Michael Branton) Subject: 3com etherlink still buggy ? Message-ID: <CMqI26.Bx@moksha.uucp> Keywords: bad, bad ethernet board Sender: mgb@moksha.uucp (Michael Branton) Organization: Totally Disorganized Date: Wed, 16 Mar 1994 01:42:05 GMT Back in december, there were some posts indicating there were problems with the 3com ethernet board and NS/I. We recently installed NS/I on a system w/ a 3com board in it and find that all network activiity will just lock up after a while, as described in some of the earlier posts. We have a single NFS mount that is not utilized much. I did set the buffer size to 1024 as suggested in a Next Answers doc,but that didn't seem to help. Anybody know what the problem is ? Anybody know a fix ? -- -Michael mgb@thoth.stetson.edu -- -Michael
Newsgroups: comp.sys.next.bugs From: Karl_Kraft@ensuing.com Subject: Re: Chess.app, player preference settings don't work Message-ID: <CMspn4.xv@ensuing.com> Sender: karl@ensuing.com (Karl Kraft) Organization: Ensuing Technologies Inc. References: <1994Mar15.070153.1726@nugget.rmNUG.ORG> Date: Thu, 17 Mar 1994 06:21:24 GMT In article <1994Mar15.070153.1726@nugget.rmNUG.ORG> karl@khaos.com (Karl Hanzel) writes: > I just now noticed that the player selections in the preferences panel for > Chess.app are pretty well hosed up. I know... it's only demoware. > The source code to Chess.app is quite available. I believe it was recently distributed on the NeXT Educational CD. It's based on Gnu Chess. > I would like to be able to set it so that two humans can play. I select > that configuration, hit <set> and <yes> to restart the game with the new > setting, whereupon the computer *continues to play* as the black defender > (offender). The GNU code that that Chess.app is based on, does not allow two players to play. The only appropriate fix would be to disable it as a selectable combonation. -- _________ Karl Kraft Karl_Kraft@ensuing.com (NeXT mail)
From: csmith@blackplague.gmu.edu (Christian Smith) Newsgroups: comp.sys.next.bugs Subject: Re: Another FAX software question. Date: 20 Mar 1994 13:30:50 GMT Organization: George Mason University, Fairfax, Virginia, USA Message-ID: <2mhj6a$5l8@portal.gmu.edu> References: <2lki24$b9@steffi.demon.co.uk> <1994Mar15.184201.2502@nidat.sub.org> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Peter Nitezki (Nitezki@NiDat.sub.org) wrote: : In article <2lki24$b9@steffi.demon.co.uk> robert@steffi.demon.co.uk : (Robert Nicholson) writes: : ...munch... : > Another problem which most of you area ware of is that repeated : > FAXing of the same document (not using redialing) : > automatically removes the previous faxes cover page : > information. : > : This has to do with the paradigm of FAXing as the other way of printing! : The cover sheet information is simply not thought to be persistent in any : way. : > Are these problems being addressed? : > : Good question! IMHO, it would mean a complete overhaul of the archiving : section of the FAX software. And a shift of paradigm, too! Not merely : another kind of printing, but a sort of telecom. A fairly simple solution to this problem would be to compose a message that includes a cover sheet, and the send this out without an additional cover sheet. I assume since you don't want to have to add a new cover sheet that you are not interested in making changes to the cover sheet? If you have a piece of special info that goes on the cover sheet and do neded to change the addressee info on the cover sheet, consider creating a special cover sheet that includes the necessary info. You will of course have to chose this cover sheet when you send the fax, but at least you won't have to manually edit it. -- Christian Smith aka Blackplague PGP Public Key available by finger or request.
From: csmith@blackplague.gmu.edu (Christian Smith) Newsgroups: comp.sys.next.bugs Subject: Re: Workspace Manager Crash!!!! Date: 20 Mar 1994 13:35:42 GMT Organization: George Mason University, Fairfax, Virginia, USA Distribution: world Message-ID: <2mhjfe$5l8@portal.gmu.edu> References: <2m5qgi$72i@lastactionhero.rs.itd.umich.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sean Michael Willson (premise@churchst.ccs.itd.umich.edu) wrote: (problem with crashing workspace) Does this occur with only your account, or do other accounts suffer the same problem. If it is just yours, have the system admin backup you account, delete the account, create a new account, and only restore your private files, ie leave .NeXT alone. I suspect that others are having similiar problem. If so then apparently something in the Workspace portion of the system has been corrupted. I would sugest a complete reinstalation and possibly an upgrade to 3.2. -- Christian Smith aka Blackplague PGP Public Key available by finger or request.
From: Aris Theocharides <aris@xedoc.com.au> Newsgroups: comp.sys.next.bugs,comp.sys.next.hardware,comp.sys.next.programmer Subject: Cirrus Logic VLB Driver Date: 21 Mar 1994 09:08:41 GMT Organization: Connect.com.au Pty Ltd, Melbourne, Australia Distribution: world Message-ID: <2mjo6p$8gr@yarrina.connect.com.au> With the source already available on /NextDeveloper/Examples/DriverKit/CirrusLogicGD542X for the ISA driver, is there anyone out there who can easily and quickly make a version that will set the VLB card in 60Hz mode, at 1024x768x2? It seems (see copied post) that the attempt at initialisation by the current driver leaves the VLB card in the 72Hz mode, at 1024x768x2. -- Aris Theocharides Xedoc Melbourne, Australia Begin forwarded article: >I have installed version NeXTstep version 3.2 on my i486DX2/66Mhz >equipped with 3 SCSI disks, 20 megabytes of memory, Cirrus Logic >VLB graphics adapter (with 1 megabytes of memory) and a mouse. > >The problem is, the Cirrus Driver included in the NeXTstep release >couldn't initialize the adapter to the desired resolution (1024* >768*2 60 Hz). I would also like to use the adapter in 1024*768*8 >mode but the driver has no options for that kind of operation. >The default VGA driver works, but shows only 800x600 pixels. > >Is there a replacement Cirrus driver? Where? Pointers to Intel >patches/drivers? > >Thanks. > >-- >-- > _____________________________________________________________ > Miikka Putaala Tatunkuja 6 90540 Oulu Finland > putaala@rieska.oulu.fi putaala@cc.oulu.fi > Phone [home] 358 (81) 362430 > Phone [elsewhere] (9400) 388400 >
Newsgroups: comp.sys.next.bugs From: ivor@cs.mcgill.ca (Ivo ROTHSCHILD) Subject: IXStore bug Message-ID: <1994Mar21.200202.16360@sifon.cc.mcgill.ca> Keywords: IXStore, commit, bug, exception Sender: news@sifon.cc.mcgill.ca Organization: SOCS, McGill University, Montreal, Canada Date: Mon, 21 Mar 1994 20:02:02 GMT This bug has been reported to bug_next. It appears that IXStore contexts do not work properly. Specifically, when 2 contexts each make modifications and then they both try to commit, exceptions are sometimes generated during the commitTransaction method. Has anyone had any luck using Store contexts? Please email to: ivor@hasc.ca -ivo Here is some code to illustrate the bug: ----- // compile with: cc test.m -lIndexing_s #import <indexing/indexing.h> #import <stdio.h> int main() { id store, store2; int i; unsigned int myB,otherB; //store = [[IXStoreFile alloc] initWithFile:"Test.store"]; store = [[IXStore alloc] init]; [store startTransaction]; [store commitTransaction]; store2 = [store copy]; [store createBlock:&myB ofSize:100]; [store commitTransaction]; for (i=0;i<10000;i++) { NX_DURING [store2 createBlock:&otherB ofSize:100]; // With the next line in it works! // If the line is commented out, we get an exception. // [store2 commitTransaction]; [store openBlock:myB atOffset:0 forLength:100]; NX_HANDLER fprintf(stderr,"***** Exception during write = %d!!!!!\n\n", NXLocalHandler.code); NX_ENDHANDLER fprintf(stderr,"count = %d\n",i); NX_DURING [store2 commitTransaction]; [store commitTransaction]; NX_HANDLER fprintf(stderr,"***** Exception during commit = %d!!!!!\n\n", NXLocalHandler.code); break; NX_ENDHANDLER } return (0); } -----
From: rgc@cs.umd.edu (Ross Garrett Cutler) Newsgroups: comp.sys.next.sysadmin,comp.sys.next.misc,comp.sys.next.bugs Subject: Parallel port problems on ALR Evolution V Date: 21 Mar 1994 17:22:12 -0500 Organization: U of Maryland, Dept. of Computer Science, Coll. Pk., MD 20742 Distribution: world Message-ID: <2ml6mk$qen@dorito.cs.umd.edu> Hi, I have an ALR Evolution V (Pentium) with the following configuration: ATI GUP/2MB VRAM Adaptec 1542C 660 MB Fujitsu SCSI HD The system works nicely, except for the fact that the parallel port doesn't get recognized: Mar 14 19:35:05 localhost mach: IOParallelPort not allocated: controller not det ected at address 0x378 Does anyone out there with an ALR Evolution V with a similar setup have a working parallel port? BTW, the parallel port works fine under Windows NT and DOS. There are no apparent hardware conflicts. I'm using NS/I 3.2 and the latest Adaptec drivers. The parallel port really is at 0x378h (I called ALR to verify). Also, the printer is online when I boot the computer (as suggested in NeXTanswers). NeXT is unable to help me (unless I pay their $120/hour fee). NeXT does not provide the source for the parallel port driver, and the tech. support guy I talked to said that they couldn't give it to me so that *I* could try and fix their bug. Help! Ross. -- Ross Cutler University of Maryland, College Park Internet: rgc@cs.umd.edu
From: aman@futon.SFSU.EDU (Aman Khalili) Newsgroups: comp.sys.next.bugs Subject: Help: How to make a mounted DOS partition READ ONLY Date: 22 Mar 1994 01:16:24 GMT Organization: San Francisco State University Message-ID: <2mlgt8$po4@news.csus.edu> I am running NS 3.2 on an intel BOX. Unfortunetly I have a DOS partition that I would like to hide from unauthorized access. It seems that you can't simply do a chmod on a the ms_dos partition. Has anyone done this? Any help is appreciated.
From: dtw07138@menudo.uh.edu (David Wang) Newsgroups: comp.sys.next.bugs Subject: Machine Slows down....... Date: 22 Mar 1994 06:44:47 GMT Organization: University of Houston Message-ID: <2mm44v$916@menudo.uh.edu> Hello folks, I am still trying to setup my 486Dx33 computer to run NeXT Step and I am ran into a very strange problem, and I was wondering if anyone has seen this or know what I/my machine is doing wrong and possibly point me in the right direction. My machine is currently configured as follows... 486Dx33 VLB with sis chipset AMi bios 20 megs of RAM 420 WD IDE drive partitioned 5 for DOS,410 for NeXT 65 MEG old seagate st277N scsi drive, all DOS Adaptec 1542B,floppy disabled, only 2 scsi devices, 1 is the seagate, the second is a sony CD-ROM. generic 16 bit ISA IDE/Floppy controller (can't get the VLB controller to work properly) 8 bit serial/paralle card, MS rodent on com1, com2 disabled generic 14.4K modem on com2. current video card is a ET4000 W32i VLB card, though only ran in std VGA mode with the UGLY 640 X 480 2 bit grayscale. but my ATI GUP is coming.... The problem is this.. once in a while, and I've ran into this several times. my computer SLOWS down, I mean really, really slow, the annoying rotating disk doesn't even turn. The computer services the mouse only once every 10 seconds or so, so the mouse doesn't move, then jumps wildly. the computer *seems* to be serving some HUGE background process which requires 99% of the CPU timeslice, and yet, there is no such back ground process runinng!! there does not seem to be any disk access, the monitor seems to think that the computer has gone into a sleep mode or something, and trys to go into the sleep mode as well by blanking otu the screen. (as far as I know, my MB is not a greenPC thing...) my computer does have some strange behavior when I run DOZ/windoz, but it doesn't try to pretend that it is an IBM PC/XT running at the 4.77 blazing speed, as it sometimes does under NeXT Step. if I can remember correctly, the first time that I encountered this was when I was in the middle of one of the failed installs, the monitor had blanked off because during the long install process, trhe monitor blanks off, I pressed the mouse button to seem what was going on and inadvertantly clicked on the cancel button, so I had tried to install again without rebooting and that was when I first saw this strange behavior. the second time was when I was playing with terminal.app and I forget what happened exactly, but the asme thing happened. and just a few hours ago, when I was formatting a floppy and trying to install SoftPC from the CD-ROm at the same time. (I enjoyed watching the two processes going in parallel) and the same thing happened.. I brought up the process table, but as usual, it reported nothing out of hte ordinary, and I had to brutally reset the computer to regain control of my machine. I have no idea what is happening, and yet I'm supposed to be a computer engineer! (boy, this will make people want to hire me now...) Has anyone out there seem this sort of behavior? do I just have an incompatib MB? what could cause the OS to service some huge non-existant process? sometimes it seems that if I press the mouse button or press the return key repeatedly, it makes the CPU service the "valid" processes (i.e. my procesese the workspace manager,etc)else, the computer just sits there and goes to sleep entirely. What's going on? help, please/// David
Newsgroups: comp.sys.next.bugs,comp.sys.next.misc From: thompson@netcom.com (Eric Thompson) Subject: Why does /dev/pp0 keep disappearing?!?!?!?! Message-ID: <thompsonCn1ztu.Go6@netcom.com> Followup-To: comp.sys.next.bugs Organization: Saperstein, et al. Date: Tue, 22 Mar 1994 06:39:30 GMT I have had this experience with two different Intel GX machines running 3.2, with two different parallel printers attached: the device /dev/pp0 disappears, apparently randomly (but probably before/during/after some reboots or power downs). It's easy enough to recreate: # mknod /dev/pp0 c 7 0 # chmod 666 /dev/pp0 <reboot> (it doesn't work unless you reboot) But how can I keep it from disappearing? Needless to say it is very annoying. Eric Thompson thompson@netcom.com
Newsgroups: comp.sys.next.sysadmin,comp.sys.next.misc,comp.sys.next.bugs From: schwett@netcom.com (Mark Schwettmann) Subject: Re: Parallel port problems on ALR Evolution V Message-ID: <schwettCn27Ar.Avu@netcom.com> Organization: NETCOM On-line Communication Services (408 241-9760 guest) References: <2ml6mk$qen@dorito.cs.umd.edu> Date: Tue, 22 Mar 1994 09:20:51 GMT Ross, This sounds very similar to a problem I had with my Gateway System. Try changing the Paralell port to LPT*2* in the CMOS setup. It seems cheesy; but it worked for me. I did this and NeXTStep prints fine; in fact so do NT, OS/2 and DOS, even though the port has been remapped... Hope this helps... Mark
Newsgroups: comp.sys.next.bugs,comp.sys.next.misc From: thompson@netcom.com (Eric Thompson) Subject: Re: Why does /dev/pp0 keep disappearing?!?!?!?! Message-ID: <thompsonCn2nwG.1FI@netcom.com> Organization: Saperstein, et al. References: <thompsonCn1ztu.Go6@netcom.com> Date: Tue, 22 Mar 1994 15:19:28 GMT I wrote: > >But how can I keep [/dev/pp0] from disappearing? Needless to say it is >very annoying. The answer is to have the printer on before powering up the computer, and /dev/pp0 will be dynamically created. Thanks for those that answered (including that guy fron NeXT :-) Eric Thompson thompson@netcom.com
From: geom2@sfb256.iam.uni-bonn.de ( Michael Moellney ) Newsgroups: comp.sys.next.bugs,comp.sys.next.hardware Subject: Re: 3com etherlink still buggy ? Date: 22 Mar 1994 16:22:59 GMT Organization: Applied Math, University of Bonn, Germany Message-ID: <2mn613$56c@news.rhrz.uni-bonn.de> References: <CMqI26.Bx@moksha.uucp> Keywords: bad, bad ethernet board Hi, I've the same experience. Work around: do a 'ping anywhere_but_the_NFS_Server_itself' on the NFS_server This brings mine to life again (so just for a few seconds to the next hang-up!-), but over an over... Good Luck, Michael
Newsgroups: comp.sys.next.bugs From: csaldanh@mae.carleton.ca (Chris Saldanha) Subject: Windows pass OVER the dock??? Message-ID: <Cn2rrr.70t@cunews.carleton.ca> Sender: news@cunews.carleton.ca (News Administrator) Organization: Carleton University Date: Tue, 22 Mar 1994 16:43:02 GMT Has anyone noticed any problems relating to Windows passing OVER the dock, instead of under? On every other machine I've ever used, it works (the dock is on a higher level tha panels and windows) like it should. I'm running 3.2 on an el-cheapo clone 486/50 (ATI GUP) Is there some kind of dwrite that affects the 'height' of standard screen objects? --Chris Chris Saldanha, Carleton University |"Let the programmers be many, and chris@computerActive.on.ca (NeXTMail) | the managers few; then all will be csaldanh@mae.carleton.ca (NeXT/MIME) | productive." "OH! It's just a little bunny isn't it?"| -The Tao of Programming
From: stoleson@banshee.rchland.ibm.com (David Stoleson) Newsgroups: comp.sys.next.bugs Subject: Re: Windows pass OVER the dock??? Date: 22 Mar 1994 20:59:04 GMT Organization: IBM Rochester MN Distribution: world Message-ID: <2mnm6o$t90@locutus.rchland.ibm.com> References: <Cn2rrr.70t@cunews.carleton.ca> In article <Cn2rrr.70t@cunews.carleton.ca>, csaldanh@mae.carleton.ca (Chris Saldanha) writes: |> Has anyone noticed any problems relating to Windows passing OVER the |> dock, instead of under? On every other machine I've ever used, it works |> (the dock is on a higher level tha panels and windows) like it should. |> Try Alt-Click on the NeXT icon. This toggles the height of the dock, either below all windows, or above all windows (except menus). - dave stoleson@vnet.ibm.com
Newsgroups: comp.sys.next.bugs From: csaldanh@mae.carleton.ca (Chris Saldanha) Subject: Re: Windows pass OVER the dock??? Message-ID: <Cn39qA.EqH@cunews.carleton.ca> Sender: news@cunews.carleton.ca (News Administrator) Organization: Carleton University References: <Cn2rrr.70t@cunews.carleton.ca> Date: Tue, 22 Mar 1994 23:10:58 GMT Chris Saldanha (csaldanh@mae.carleton.ca) wrote: : Has anyone noticed any problems relating to Windows passing OVER the : dock, instead of under? On every other machine I've ever used, it works : (the dock is on a higher level tha panels and windows) like it should. Thank to all who replied. The trick is to ALTERNATE-Click on the NeXT Logo on the Dock. This toggles the dock from low to high. Not a bug, but a feature. :-) --Chris Chris Saldanha, Carleton University |"Let the programmers be many, and chris@computerActive.on.ca (NeXTMail) | the managers few; then all will be csaldanh@mae.carleton.ca (NeXT/MIME) | productive." "OH! It's just a little bunny isn't it?"| -The Tao of Programming
Newsgroups: comp.sys.next.bugs From: kieffer@gemstone.com (Robert W. Kieffer) Subject: Re: Windows pass OVER the dock??? Message-ID: <Cn31wB.6p0@gemstone.com> Sender: kieffer@gemstone.com (Robert W. Kieffer) Organization: Gemstone Systems, Inc. References: <Cn2rrr.70t@cunews.carleton.ca> Date: Tue, 22 Mar 1994 20:21:46 GMT In article <Cn2rrr.70t@cunews.carleton.ca> csaldanh@mae.carleton.ca (Chris Saldanha) writes: > Has anyone noticed any problems relating to Windows passing OVER the > dock, instead of under? On every other machine I've ever used, it works > (the dock is on a higher level tha panels and windows) like it should. > > I'm running 3.2 on an el-cheapo clone 486/50 (ATI GUP) > > Is there some kind of dwrite that affects the 'height' of standard screen > objects? <alternate>-clicking on the Workspace icon toggles this effect. --- Robert Kieffer - kieffer@gemstone.com - NeXTMail welcome Gemstone Systems, Inc. 19616 Redbeam Ave. Torrance, CA 90503
Newsgroups: comp.sys.next.bugs From: gerben@rna.indiv.nluug.nl Subject: Re: Windows pass OVER the dock??? Message-ID: <1994Mar23.142150.1263@rna.indiv.nluug.nl> Sender: gerben@rna.indiv.nluug.nl (Gerben Wierda) Organization: G.R.O.S.S. References: <Cn31wB.6p0@gemstone.com> Date: Wed, 23 Mar 1994 14:21:50 GMT In article <Cn31wB.6p0@gemstone.com> kieffer@gemstone.com (Robert W. Kieffer) writes: > In article <Cn2rrr.70t@cunews.carleton.ca> csaldanh@mae.carleton.ca (Chris > Saldanha) writes: > > Has anyone noticed any problems relating to Windows passing OVER the > > dock, instead of under? On every other machine I've ever used, it works > > (the dock is on a higher level tha panels and windows) like it should. > > > > I'm running 3.2 on an el-cheapo clone 486/50 (ATI GUP) > > > > Is there some kind of dwrite that affects the 'height' of standard > screen > > objects? > > <alternate>-clicking on the Workspace icon toggles this effect. > Nice! Too bad Engage! doesn't support that. SO now I have a stck of icons that are below the window and a set that is on top. Crazy. -- Gerben Wierda [NEXTSTEP RD242] gerben@rna.indiv.nluug.nl "If you don't know where you are going, any road will take you there." From the Talmud, rephrased in Lewis Carroll, "Alice in Wonderland".
From: sanguish@digifix.com Newsgroups: comp.sys.next.announce,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.bugs,comp.sys.next.programmer,comp.sys.next.marketplace Subject: Quick Guide to NEXTSTEP information on the Internet Date: 23 Mar 1994 14:42:45 -0500 Organization: Next Announcements Message-ID: <2mq63l$gm@digifix.digifix.com> This post is made weekly, to help 'point' users to more NEXTSTEP information Topics include: comp.sys.next newsgroups related newsgroups comp.sys.next newsgroups mailing list ftp sites NeXTanswers comp.sys.next.* newsgroups -------------------------- Comp.Sys.Next.Advocacy This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. Comp.Sys.Next.Announce Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) The NEXTSTEP FAQs are posted here monthly as well. This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Comp.Sys.Next.Bugs A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT- specific groups as well. Comp.Sys.Next.Hardware Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. Comp.Sys.Next.Marketplace NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. Comp.Sys.Next.Misc For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! Comp.Sys.Next.Programmer Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. The NEXTSTEP programmer FAQs are posted here. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. Comp.Sys.Next.Software This is a place to talk about [third party] software products that run on NEXTSTEP systems. Comp.Sys.Next.Sysadmin Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. related Newsgroups ------------------ Comp.Soft-Sys.Nextstep Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. Comp.Lang.Objective-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. Comp.Object Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original Comp.Sys.Next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News ------------------------------------------- Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep-d next-advocacy-d next-announce-d next-bugs-d next-hardware-d next-marketplace-d next-misc-d next-programmer-d next-software-d next-sysadmin-d (For a full description, send mail saying LISTS to <digestif@antigone.com>). The subscription syntax is essentially the same as LISTSERV's. To subscribe, send a message to <digestif@antigone.com> saying: SUB Listname YourName Example: SUB next-hardware-d John Doe The ftp sites ------------- cs.orst.edu: The main site for North American submissions nova.cc.purdue.edu: Lots of older stuff, but very short on disk space ftp.informatik.uni-muenchen.de: In Germany. terra.stack.urc.tue.nl (Dutch NEXTSTEP User Group) and cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp.next.com: See the below ftp.next.com and NextAnswers@next.com ------------------------------------- From the document 1000_Help from ftp.next.com Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, or you can transfer them by anonymous ftp. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to NeXTanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO <your-address> If you have any problem with the system or suggestions for improvement, please send mail to NeXTanswers-request@NeXT.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to NeXTanswers-request@NeXT.com. Written by: Eric P. Scott eps@toaster.SFSU.EDU and Scott Anguish sanguish@digifix.com Additions from: Greg Anderson (Greg_Anderson@afs.com) and Michael Pizolato (Michael_Pizolato@afs.com)
From: strobel@u.washington.edu (Nicolas Strobel) Newsgroups: comp.sys.next.bugs Subject: workspace and other apps preferences not being saved! Date: 24 Mar 1994 01:02:35 GMT Organization: University of Washington Message-ID: <2mqorb$68k@news.u.washington.edu> Summary: preferences set with Preferences.app not saved on logout! Keywords: preferences I got a new 1Gig disk to replace the original 330Mb disk and loaded 3.2 onto it. However, with a clean install of 3.2 I've been having problems with the defaults/preferences file .NeXT/.NeXTdefaults.L . When I run Preferences.app and set the Unix Expert flag to see all of the files, that preference is NOT saved when I log out. I have to reset it when I log back in. A bit annoying. The preferences set by Edit.app and TeXEdit.app are also not saved properly. I have to reset them when I log back in. But those preferences seem to be save in the .NeXTdefaults.L file--in fact, multiple copies of the Edit and TeXEdit preferences are saved there but are not read. The .NeXTdefaults.L file grows with every login as I reset the Workspace preferences and the Edit.app/TeXEdit.app preferences. I did not have this problem with 3.2 installed on the original 330Mb disk. However, that 3.2 was an UPGRADE from a 3.0 filesystem. Are there some 3.0 files that magically need to remain on the disk but are not there in a 3.2 system installed from scratch?? Very peculiar! Nick Strobel strobel@astro.washington.edu
From: kampp@neon (Ralf &) Newsgroups: comp.sys.next.bugs Subject: Re: swapfile keeps growing .... Date: 18 Mar 1994 12:53:15 GMT Organization: University Koblenz / Germany Message-ID: <2mc889$ddq@newshost.uni-koblenz.de> References: <2m2bua$1kj@ysics.physics.sunysb.edu> Keywords: swapfile In article <2m2bua$1kj@ysics.physics.sunysb.edu> juliana@semlab1.sbs.sunysb.edu (Juliana L. Silva) writes: > > I noticed my system (486/66) was getting slower and the available > diskspace was "disppearing". When I checked the swapfile, it was > more than 100MB long. I left the system up a couple more days to > see what would happen. Well, the swapfile kept growing, my hard drive > ended up full and I had to reboot the machine. > > Has anyone experienced that? > > > Juliana Freire > juliana@cs.sunysb.edu > juliana@semlab1.sbs.sunysb.edu (<---NeXT mail ok) Look into your /etc/swaptab file. Maybe the highwater mark (should be: lowwat=...,hiwat=...) isn't set in addition to the lowwater mark!! Simply set it to something reasonable and reboot.
Newsgroups: comp.sys.next.bugs From: gerti@tms-gmbh.de (Gerd Knops) Subject: Memory access exception in rld_load() Message-ID: <Cn628r.no@tms-gmbh.de> Sender: usenet@tms-gmbh.de Organization: tms GmbH, Regensburg, Germany Date: Thu, 24 Mar 1994 11:22:02 GMT Under certain circumstances we have a lot of trouble with a bug, that might have its cause in rld_load(). When dynamically loading classes into our program, we sometimes run into a memory access exception, that always occurs in 'layout_merged_symbols ()' deep down in rld_load(). The address, that causes the exception is usually 0x20 or 0x24. (see backlog) Slight modifications to any module (eg making the binary 2 byte longer or shorter) or changing the order, in which the modules are loaded, and everything works fine. We have this problem at least since 3.0, and also in 3.1 and 3.2. It was really bad in 3.0, until we played with objc_setMultithreaded(), which seemed to help a lot, even though objc_loadModules() is documented to be thread safe. But now it reoccurred with increasing probability. To make things worse: the problem almost never occurs in the 'debug' version. Here is, what the debugger shows (one of those rare occasions, it did occur in the 'debug' version): Program generated(1): Memory access exception on address 0x24 (protection failure). 0x502b0ba in layout_merged_symbols () (gdb) where Reading in symbols for VISIONModuleController.m...done. Reading in symbols for VISIONModule.m...done. Reading in symbols for TMSMovableMatrix.m...done. Reading in symbols for VISIONApp.m...done. Reading in symbols for 1VISION_main.m...done. #0 0x502b0ba in layout_merged_symbols () #1 0x50230d2 in layout_segments () #2 0x502913a in layout () #3 0x5027c3a in internal_rld_load () #4 0x502acfa in rld_load () #5 0x5037a30 in objc_loadModules () #6 0x14292 in -[VISIONModuleController loadClass:] (self=0x2f1358, _cmd=0x27b47, theClassName=0x3ffe8aa "TMSStrokeAttributesWell") at VISIONModuleController.m:983 #7 0x147aa in -[VISIONModuleController loadNeededClassesForModule:] (self=0x2f1358, _cmd=0x27b2b, theModule=0x3719fc) at VISIONModuleController.m:1166 #8 0x13ec6 in -[VISIONModuleController loadClass:] (self=0x2f1358, _cmd=0x27b47, theClassName=0x3087c0 "Frames") at VISIONModuleController.m:882 #9 0x1238a in handleObjcError (className=0x3087c0 "Frames") at VISIONModuleController.m:65 #10 0x5009440 in objc_getClass () Any help would be very much appreciated! Gerd
From: izumi@pinoko.berkeley.edu (Izumi Ohzawa) Newsgroups: comp.sys.next.sysadmin,comp.sys.next.bugs Subject: Warning: Too few i-nodes allocated for multi GB drives Date: 25 Mar 1994 03:13:44 GMT Organization: University of California, Berkeley Distribution: world Message-ID: <2mtkt8$s1e@agate.berkeley.edu> Attention: Those with >=2GB drives, You might want to check how many i-nodes you have for drives newly partitioned and initialized. My initial try with DEC DSP5200S (2GB) only allocated one i-node per 22kb. I actually ran out of i-nodes with only 1/3 full partition. I don't know what happened, but the log I saved shows something odd: # newfs -v /dev/rsd3a /etc/mkfs /dev/rsd3a 976897 71 21 8192 1024 32 10 60 4096 t Warning: 4096 bytes per inode impossible due to cylinder group size, using 22235 bytes per inode Reduce cylinder group size to reduce bytes per inode. Why bytes per i-node jumps to 22kb when it cannot do 4kb as requested? For now, I have worked around this by reducing the ca#, cb#.. to 8 from 32 in /etc/disktab. You might want to check the allocation of i-nodes by "df -i" before you put a lot of files onto the drive. Otherwise, you might be surprised soon even if your 'df' indicates plenty of space. Did someone break mkfs or newfs in the recent NS releases? -- Izumi Ohzawa [ $@Bg_78^=;(J ] USMail: University of California, 360 Minor Hall, Berkeley, CA 94720 Telephone: (510) 642-6440 Fax: (510) 642-3323 Internet: izumi@pinoko.berkeley.edu (NeXTMail OK)
Newsgroups: comp.sys.next.bugs From: thompson@netcom.com (Eric Thompson) Subject: Engage Re: Windows pass OVER the dock??? Message-ID: <thompsonCn7AEx.A1v@netcom.com> Organization: Saperstein, et al. References: <Cn31wB.6p0@gemstone.com> <1994Mar23.142150.1263@rna.indiv.nluug.nl> Date: Fri, 25 Mar 1994 03:16:08 GMT >> <alternate>-clicking on the Workspace icon toggles this effect. >> > >Nice! Too bad Engage! doesn't support that. SO now I have a stck of icons that >are below the window and a set that is on top. Crazy. Engage Desktop works w/alt-clicking. It's a cheap upgrade, I recommend it. Eric thompson@netcom.com (NeXTmail)
Newsgroups: comp.sys.next.bugs From: gvdl@obelix.apana.org.au Subject: Lost Services and c.s.n FAQ's (2 subjects) Message-ID: <Cn7IMu.1BJ@obelix.apana.org.au> Keywords: Services, Gone, FAQ Sender: gvdl@obelix.apana.org.au (Godfrey Martin van der Linden) Date: Fri, 25 Mar 1994 06:13:41 GMT G'day A small problem with services. I seem to have lost them. My other users and root still have a complete services menu, but I can't seem to be able to get one. I'm getting error messages in the Console like make_services: permission denied [Q] Permission denied to do what?????? In case this is a FAQ, can somebody direct me to an archive site where I can get hold of the FAQ's. I only have ftpmail access so please send the site and pathname required. Godfrey van der Linden Next-mail prefered
From: robert@steffi.demon.co.uk (Robert Nicholson) Newsgroups: comp.sys.next.bugs Subject: Why does a NeXT crawl when any user has uid 0? Date: 25 Mar 1994 13:35:16 -0000 Organization: me organised, that's a joke. Message-ID: <2mupak$472@steffi.demon.co.uk> I was doing some programming today where I had to run as a nonroot with uid 0. I noticed that everything starts to crawl when you are like that. -- "C++ is the best C++ there is." (ASCII for text only messages)
Newsgroups: comp.sys.next.bugs From: gerti@tms-gmbh.de (Gerd Knops) Subject: Re: Warning: Too few i-nodes allocated for multi GB drives Message-ID: <Cn83L0.H9@tms-gmbh.de> Sender: usenet@tms-gmbh.de Organization: tms GmbH, Regensburg, Germany References: <2mtkt8$s1e@agate.berkeley.edu> Date: Fri, 25 Mar 1994 13:46:11 GMT I had the same happening on a 44MB SyQuest, that was used for backups of source code (eg small text files). It went out of inodes, when it was half full. So after messing arround a bit with several commands, I managed to reformat it with a increased number of inodes. Gerd In article <2mtkt8$s1e@agate.berkeley.edu> izumi@pinoko.berkeley.edu (Izumi Ohzawa) writes: > Attention: Those with >=2GB drives, > > You might want to check how many i-nodes you have for drives newly > partitioned and initialized. My initial try with DEC DSP5200S (2GB) > only allocated one i-node per 22kb. I actually ran out of i-nodes > with only 1/3 full partition. I don't know what happened, but the > log I saved shows something odd: > > # newfs -v /dev/rsd3a > /etc/mkfs /dev/rsd3a 976897 71 21 8192 1024 32 10 60 4096 t > Warning: 4096 bytes per inode impossible due > to cylinder group size, using 22235 bytes per inode > Reduce cylinder group size to reduce bytes per inode. > > Why bytes per i-node jumps to 22kb when it cannot do 4kb as requested? > > For now, I have worked around this by reducing the ca#, cb#.. to 8 from > 32 in /etc/disktab. > > You might want to check the allocation of i-nodes by "df -i" before > you put a lot of files onto the drive. Otherwise, you might be > surprised soon even if your 'df' indicates plenty of space. > > Did someone break mkfs or newfs in the recent NS releases? > > -- > Izumi Ohzawa [ $@Bg_78^=;(J ] > USMail: University of California, 360 Minor Hall, Berkeley, CA 94720 > Telephone: (510) 642-6440 Fax: (510) 642-3323 > Internet: izumi@pinoko.berkeley.edu (NeXTMail OK)
Newsgroups: comp.sys.next.bugs From: yanik@planon.qc.ca (Yanik Crepeau) Subject: Re: Windows pass OVER the dock??? Message-ID: <1994Mar24.165014.18242@planon.qc.ca> Sender: yanik@planon.qc.ca (Yanik Crepeau) References: <Cn2rrr.70t@cunews.carleton.ca> Date: Thu, 24 Mar 1994 16:50:14 GMT In article <Cn2rrr.70t@cunews.carleton.ca> csaldanh@mae.carleton.ca (Chris Saldanha) writes: > Has anyone noticed any problems relating to Windows passing OVER the > dock, instead of under? On every other machine I've ever used, it works > (the dock is on a higher level tha panels and windows) like it should. > > I'm running 3.2 on an el-cheapo clone 486/50 (ATI GUP) > > Is there some kind of dwrite that affects the 'height' of standard screen > objects? > > --Chris > Chris Saldanha, Carleton University |"Let the programmers be many, and > chris@computerActive.on.ca (NeXTMail) | the managers few; then all will be > csaldanh@mae.carleton.ca (NeXT/MIME) | productive." > "OH! It's just a little bunny isn't it?"| -The Tao of Programming This is an undocumented feature of the Workspace Manager. If you press the alternate-key and you double-click the NeXT logo in the dock, the dock will change its behaviour the way you described. To restore the original behaviour, just double-click the NeXT logo again with Alt key pressed. Yanik -- Yanik Crepeau The power of OOP used to stop when Programmer the programmer typed "make". With Planon Telexpertise NEXTSTEP and PDO, that is not true E-Mail: yanik@planon.qc.ca (NeXT) anymore!
Newsgroups: comp.sys.next.bugs From: horshack@lisa.franken.de (Richard Lippmann) Subject: Crashs in NSfI/ET4000W32-VLB Message-ID: <Cn4pz0.E1@lisa.franken.de> Sender: root@lisa.franken.de (Operator) Organization: Kommunikationsnetz Franken e.V. Date: Wed, 23 Mar 1994 17:59:23 GMT I have a lot of crashs with my NSfI/486. I don't know why... I have a 486/66DX2 with an ET4000W32/VLB-Videocard. I use 1024x768 2-Bit-Greyscale. After ca. 20 Minutes I have crash where a system-window appears and I find the following in my /usr/adm/messages: Mar 23 18:29:18 lisa Workspace[294]: logged in Mar 23 18:29:28 lisa TickleServer.daemon[267]: Error raised in Listener run loop, continuing ... Mar 23 18:29:28 lisa TickleServer.daemon[267]: Error raised in Listener run loop, continuing ... Mar 23 18:29:31 lisa TickleServer.daemon[267]: Error raised in Listener run loop, continuing ... Mar 23 18:29:43 lisa last message repeated 3 times Mar 23 18:29:49 lisa TickleServer.daemon[267]: Error raised in Listener run loop, continuing ... Mar 23 18:30:06 lisa TickleServer.daemon[267]: Error raised in Listener run loop, continuing ... Mar 23 18:32:24 lisa TickleServer.daemon[267]: Error raised in Listener run loop, continuing ... Mar 23 18:33:16 lisa WindowServer[290]: WindowServer: Exception caught: PS-108.1 on host `lisa', user `horshack': Illegal instruction (SIGILL), Mar 23 18:33:16 lisa WindowServer[290]: [ eip = 0x8f0f Mar 23 18:33:16 lisa WindowServer[290]: , from 0xc172 Mar 23 18:33:16 lisa WindowServer[290]: , from 0x29010 Mar 23 18:33:16 lisa WindowServer[290]: , from 0x178f9 Mar 23 18:33:16 lisa WindowServer[290]: , from 0x17561 Mar 23 18:33:16 lisa WindowServer[290]: , from 0x1c5f2 Mar 23 18:33:16 lisa WindowServer[290]: , from 0x75a74 Mar 23 18:33:16 lisa WindowServer[290]: , from 0xc172 Mar 23 18:33:16 lisa WindowServer[290]: , from 0x56b00 Mar 23 18:33:16 lisa WindowServer[290]: ] Mar 23 18:33:16 lisa WM[295]: DPS client library error: Connection closed unexpectedly, DPSContext b2e68, data 0 Mar 23 18:33:16 lisa Mail[297]: DPS client library error: Error while writing to connection, DPSContext 6e800, data -102 Mar 23 18:33:16 lisa Preferences[296]: DPS client library error: Connection closed unexpectedly, DPSContext 4b34c, data 0 Mar 23 18:33:16 lisa Mail[297]: Exiting due to Window Server death Mar 23 18:33:16 lisa Preferences[296]: Exiting due to Window Server death Mar 23 18:33:16 lisa Terminal[298]: DPS client library error: Error while writing to connection, DPSContext 41120, data -102 Mar 23 18:33:16 lisa Terminal[298]: Exiting due to Window Server death Mar 23 18:33:31 lisa loginwindow[289]: loginwindow: Workspace exited ts 0 cd 0 rc 0 sv 0 ss 0. Mar 23 18:33:33 lisa loginwindow[289]: dps_err_write Mar 23 18:33:33 lisa last message repeated 20 times Mar 23 18:33:33 lisa loginwindow[289]: exiting due to excessive DPS errors Mar 23 18:33:43 lisa loginwindow[327]: loginwindow: Checking for DOS partitioned disk Mar 23 18:33:49 lisa loginwindow[331]: loginwindow: running /usr/lib/NextStep/Workspace.app/Workspace Mar 23 18:33:50 lisa Workspace[331]: logged in Mar 23 18:33:59 lisa TickleServer.daemon[267]: Error raised in Listener run loop, continuing ... Mar 23 18:34:00 lisa TickleServer.daemon[267]: Error raised in Listener run loop, continuing ... Mar 23 18:34:02 lisa TickleServer.daemon[267]: Error raised in Listener run loop, continuing ... Mar 23 18:34:14 lisa last message repeated 3 times Mar 23 18:34:21 lisa TickleServer.daemon[267]: Error raised in Listener run loop, continuing ... Mar 23 18:34:30 lisa TickleServer.daemon[267]: Error raised in Listener run loop, continuing ... What can I do? Is it the videocard which runs hot? Maybe it's my mouse (Microsoft serial mouse). Often when I touch my mouse after a few minutes this crash happens. Thanks Horshack -- Richard Lippmann Nur Beamte koennen wie... phone voice: +49+911+69.84.58 ... Beamte denken mail : horshack@lisa.franken.de (welcome NeXTmail!)
Newsgroups: comp.sys.next.bugs From: gvdl@obelix.apana.org.au Subject: Lost Services Still Lost(Further information) Message-ID: <Cn940r.135@obelix.apana.org.au> Sender: gvdl@obelix.apana.org.au (Godfrey Martin van der Linden) References: <Cn7IMu.1BJ@obelix.apana.org.au> Date: Sat, 26 Mar 1994 02:53:14 GMT In article <Cn7IMu.1BJ@obelix.apana.org.au> gvdl@obelix.apana.org.au writes: > G'day > > A small problem with services. I seem to have lost them. My other users and > root still have a complete services menu, but I can't seem to be able to get > one. I'm getting error messages in the Console like > This is the actual output Mar 26 10:31:21 obelix syslog: make_services: cannot read services applist file. And when I try to disable services by hand I get the following enable/disable_service: couldn't disable "Terminal": Permission denied I get this message even after the following commands cd ~ chmod -R a+rwX Library .NeXT I have tried deleting the .NeXT/services directory, still nothing. Come on you people out there surely somebody has seen this, before! Thanx in advance Godfrey van der Linden Next-mail prefered
Newsgroups: comp.sys.next.bugs From: bbum@friday.com Subject: Re: Lost Services and c.s.n FAQ's (2 subjects) Message-ID: <CnC6v0.Kq@friday.com> Sender: bbum@friday.com (Bill Bumgarner) Organization: Friday Software & Consulting; a virtual company References: <Cn7IMu.1BJ@obelix.apana.org.au> Date: Sun, 27 Mar 1994 18:47:23 GMT gvdl@obelix.apana.org.au writes > G'day > > A small problem with services. I seem to have lost them. My > other users and root still have a complete services menu, but I > can't seem to be able to get one. I'm getting error messages in > the Console like > > make_services: permission denied > > [Q] Permission denied to do what?????? Most likely to either the ~/.NeXT directory or the .services (or is it just services) file contained within .NeXT. Either that or the make_services binary itself; it is located in /usr/bin. -- b.bum <bbum@friday.com> | "blah blah blah" WhiteLight Systems, INC. | - someone somewhere
Newsgroups: comp.sys.next.bugs From: ugubser@avalon.unizh.ch (gubser urs) Subject: Bug with Floppy Device ? (itel) Message-ID: <1994Mar28.124735.2162@ifi.unizh.ch> Sender: news@ifi.unizh.ch (USENET News Admin) Organization: University of Zurich, Department of Computer Science Date: Mon, 28 Mar 1994 12:47:35 GMT Is this a bug ? When i open the /dev/rfd0b device, the workspace is unable to access the floppy. Try the following prog: #include "stdio.h" #include "stdlib.h" #include <sys/file.h> void main() { int fd; if((fd = open("/dev/rfd0b", O_RDWR, 0)) <= 0) { perror("Open call "); exit(1); } close(fd); } After this is executed, the Workspace won't recognize the disk drive ... Please help... PS: Intel only -- * Urs Gubser | Email: * * Bergheimstrasse 7 | ugubser@avalon.unizh.ch (NeXT - mail) * * Zuerich | * * Switzerland | Bix: urs *
From: wjadams@hubcap.clemson.edu (William J Adams) Newsgroups: comp.sys.next.bugs Subject: Next/FIP w/cirrus vlb, can't get out of interlaced mode Date: 28 Mar 94 23:57:16 GMT Organization: Clemson University Message-ID: <wjadams.764899036@hubcap> Hey, I just installed Next for intel processors on my machine last week, and I can't figure out how to the refresh rate. Currently it is set at 60 hz, and that puts my monitor in interlaced mode (yuch!). If I could find a way to bump it up to 72 hz I'd be in good shape (as long as everything was still in sync... I'd install linux w/X-windows on this machine at the beginning of the year, and had a nightmare of a time trying to get the video mode straightened out). Anyway, if anyone could help I'd really appreciate it. You can just email me at wjadams@hubcap.clemson.edu. Thanks Bill D. Cat
From: sanguish@digifix.com Newsgroups: comp.sys.next.announce,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.bugs,comp.sys.next.programmer,comp.sys.next.marketplace Subject: Quick Guide to NEXTSTEP information on the Internet Date: 28 Mar 1994 01:43:43 -0500 Organization: Next Announcements Message-ID: <2n5uav$9op@digifix.digifix.com> This post is made weekly, to help 'point' users to more NEXTSTEP information Topics include: comp.sys.next newsgroups related newsgroups comp.sys.next newsgroups mailing list ftp sites NeXTanswers comp.sys.next.* newsgroups -------------------------- Comp.Sys.Next.Advocacy This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. Comp.Sys.Next.Announce Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) The NEXTSTEP FAQs are posted here monthly as well. This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Comp.Sys.Next.Bugs A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT- specific groups as well. Comp.Sys.Next.Hardware Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. Comp.Sys.Next.Marketplace NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. Comp.Sys.Next.Misc For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! Comp.Sys.Next.Programmer Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. The NEXTSTEP programmer FAQs are posted here. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. Comp.Sys.Next.Software This is a place to talk about [third party] software products that run on NEXTSTEP systems. Comp.Sys.Next.Sysadmin Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. related Newsgroups ------------------ Comp.Soft-Sys.Nextstep Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. Comp.Lang.Objective-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. Comp.Object Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original Comp.Sys.Next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News ------------------------------------------- Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep-d next-advocacy-d next-announce-d next-bugs-d next-hardware-d next-marketplace-d next-misc-d next-programmer-d next-software-d next-sysadmin-d (For a full description, send mail saying LISTS to <digestif@antigone.com>). The subscription syntax is essentially the same as LISTSERV's. To subscribe, send a message to <digestif@antigone.com> saying: SUB Listname YourName Example: SUB next-hardware-d John Doe The ftp sites ------------- cs.orst.edu: The main site for North American submissions nova.cc.purdue.edu: Lots of older stuff, but very short on disk space ftp.informatik.uni-muenchen.de: In Germany. terra.stack.urc.tue.nl (Dutch NEXTSTEP User Group) and cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp.next.com: See the below ftp.next.com and NextAnswers@next.com ------------------------------------- From the document 1000_Help from ftp.next.com Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, or you can transfer them by anonymous ftp. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to NeXTanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO <your-address> If you have any problem with the system or suggestions for improvement, please send mail to NeXTanswers-request@NeXT.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to NeXTanswers-request@NeXT.com. Written by: Eric P. Scott eps@toaster.SFSU.EDU and Scott Anguish sanguish@digifix.com Additions from: Greg Anderson (Greg_Anderson@afs.com) and Michael Pizolato (Michael_Pizolato@afs.com)
Newsgroups: comp.sys.next.bugs From: Nitezki@NiDat.sub.org (Peter Nitezki) Subject: Re: Warning: Too few i-nodes allocated for multi GB drives Message-ID: <1994Mar29.205755.2282@nidat.sub.org> Sender: nitezki@nidat.sub.org Organization: private site of Peter Nitezki, Kraichtal, Germany References: <2mtkt8$s1e@agate.berkeley.edu> Date: Tue, 29 Mar 1994 20:57:55 GMT In article <2mtkt8$s1e@agate.berkeley.edu> izumi@pinoko.berkeley.edu (Izumi Ohzawa) writes: > Attention: Those with >=2GB drives, > > You might want to check how many i-nodes you have for drives newly > partitioned and initialized. My initial try with DEC DSP5200S > (2GB) only allocated one i-node per 22kb. I actually ran out of > i-nodes with only 1/3 full partition. I don't know what > happened, but the log I saved shows something odd: > > # newfs -v /dev/rsd3a /etc/mkfs /dev/rsd3a 976897 71 21 8192 > 1024 32 10 60 4096 t Warning: 4096 bytes per inode impossible due > to cylinder group size, using 22235 bytes per inode Reduce > cylinder group size to reduce bytes per inode. > > Why bytes per i-node jumps to 22kb when it cannot do 4kb as > requested? > > For now, I have worked around this by reducing the ca#, cb#.. to 8 > from 32 in /etc/disktab. > > You might want to check the allocation of i-nodes by "df -i" > before you put a lot of files onto the drive. Otherwise, you might > be surprised soon even if your 'df' indicates plenty of space. > > Did someone break mkfs or newfs in the recent NS releases? > Well, I don't know whether someone broke mkfs, but I'm quite sure he did it long time ago (at least long before NS :-) Newfs and anything atop of it, like BuildDisk, uses heuristics to determine i-node quantity and block and frag size. Unfortunately, not every combination works and even less of them fly! Newfs is no AI, its a simple heuristic, so don't expect miracles from it. When it was first devised the mention of 2GB disks would have qualified everyone as a loony, believe me! In the times of early BSD development even a 300MB drive cost a fortune and was delivered one in truck. But enough of war stories and horror tales. I just want to state that everyone who wants to have partitions in the Giga range should do newfs manually and he will have to test quite a number of parameter settings for good results. Same for people making a file system for their news server. Actually, I only have recently converted to using newfs. I used to be a strict mkfs believer... -- Peter Nitezki | Nitezki@NiDat.sub.org # Blessed art thou who knoweth Staarenbergstr. 44 | Tel.: +49 7251 62495 # not about the pleasure and D-76703 Kraichtal | Fax : +49 7251 69215 # delight of being hooked GERMANY | NeXTmail welcome !!! # up to the Net. Peter 1,3-5
From: blake015@mc.duke.edu(Denise Blakeley) Newsgroups: comp.sys.next.bugs Subject: BackSpace bug? Date: 29 Mar 1994 22:34:38 GMT Organization: Duke University; Durham, N.C., USA Message-ID: <2naadu$5bn@news.duke.edu> Has anyone else noticed this in BackSpace v1.02? I have it set to be a screen locker via BackSpace's own preferences, so of course this means I had to make up a password. I launch BackSpace and set it to screen-locking. Later, I bump the mouse so that it prompts me for the password. But if I just sit there for a few seconds without doing anything, it unlocks the screen anyway! Denise -- Denise Blakeley | PROGRAM, tr. v., An activity similar Duke Med Center Info Systems | to banging one's head against a wall, Durham, NC | but with fewer opportunities for (919) 286-6468 W | reward. blake015@mc.duke.edu | NeXTMail welcome!
From: izumi@pinoko.berkeley.edu (Izumi Ohzawa) Newsgroups: comp.sys.next.bugs,comp.sys.next.sysadmin Subject: Re: Warning: Too few i-nodes allocated for multi GB drives Date: 30 Mar 1994 01:08:28 GMT Organization: University of California, Berkeley Distribution: world Message-ID: <2najec$s39@agate.berkeley.edu> References: <1994Mar29.205755.2282@nidat.sub.org> In article <1994Mar29.205755.2282@nidat.sub.org> Nitezki@NiDat.sub.org (Peter Nitezki) writes: >In article <2mtkt8$s1e@agate.berkeley.edu> izumi@pinoko.berkeley.edu >(Izumi Ohzawa) writes: >> Attention: Those with >=2GB drives, >> >> You might want to check how many i-nodes you have for drives newly >> partitioned and initialized. My initial try with DEC DSP5200S >> (2GB) only allocated one i-node per 22kb. >> ... >> Did someone break mkfs or newfs in the recent NS releases? > > >Well, I don't know whether someone broke mkfs, but I'm quite sure he did >it long time ago (at least long before NS :-) > ... >When it was first devised the mention of 2GB disks >would have qualified everyone as a loony, believe me! Perhaps so, but things did work not so long ago with 1GB drive/partition (Fujitsu M2266S), and our new 2GB drive (DEC 5200S) is divided into two 1GB partitions each of which is smaller than the 1-partition Fujitsu. I got a tip from someone that 'newfs' indeed became broken with Release 3.1 and did not get fixed for 3.2, and that I might try the binary from release 3.0. I intend to that sometime, when I can't find anything better to do. Thanks NeXT for keeping this type of info from the eyes of lowey users who are not about to call Tech Support to get humiliated. -- Izumi Ohzawa [ $@Bg_78^=;(J ] USMail: University of California, 360 Minor Hall, Berkeley, CA 94720 Telephone: (510) 642-6440 Fax: (510) 642-3323 Internet: izumi@pinoko.berkeley.edu (NeXTMail OK)
Newsgroups: comp.sys.next.bugs Organization: Antigone Press gateway, San Francisco Return-Path: <valmar!sas@decster.uta.edu> Message-ID: <m0pm09S-0001xsC@valmar> Date: Wed, 30 Mar 94 07:17 CST From: sas@valmar.netcom.com (Shawn Alva Stone) Subject: Re: BackSpace bug? Denise: I've been bummed out about Backspace for a long while. Somehow - I know not how - I got some of the display options duplicated. So now I have 2 flexicubes, 2 sperms, 2 fading images, etc. When I choose the wrong one of any of the duplicated options, I'm given some kind of "You're a moron" message. The worst of it is, though, that I've never been able to get the 'world' or 'earth' (whatever it's called) to work. (I seem to remember trying to hack on it and inadvertently duplicating some of the other display options). I grew disgusted with the app long ago and did the only thing I knew to do: I compressed that sucker and am now using Lockscreen! --- Ciao, Shawn <sas@halite.uta.edu> "Does anybody really know what time it is?" - R. Lamm, '70s
From: samurai@wsc.com (Darcy Brockbank) Newsgroups: comp.sys.next.bugs Subject: <mach-task> back to haunt me Date: 31 Mar 1994 01:24:18 GMT Organization: WSC Investment Services, Inc. Disclaimer: The views represented within this posting are not necessarily those held by WSC Investment Services, Inc. Message-ID: <2nd8o2$dc@cerberus.wsc.com> Hello all... We're running several different kinds of PCs here, and each of them is exhibiting the mach-task flakiness. Normally, I wouldn't care so much, except that one of them seems to be sucking on resources: USER PID %CPU %MEM VSIZE RSIZE TT STAT TIME COMMAND root -1 0.4 21.7 52.2M 6.95M ? U 11:06 <mach-task> root -1 0.0 0.0 0K 0K ? ?W< 0:00 <mach-task> root -1 0.0 0.0 0K 0K ? SW 0:00 <mach-task> root -1 0.0 0.0 0K 0K ? S < 0:00 <mach-task> root -1 0.0 0.0 0K 0K ? S 0:00 <mach-task> root -1 0.0 0.0 0K 0K ? S 0:00 <mach-task> root -1 0.0 0.0 0K 0K ? S 0:00 <mach-task> root -1 0.0 0.0 0K 0K ? S 0:00 <mach-task> root -1 0.0 0.0 0K 0K ? S 0:00 <mach-task> Now, I booted some machines up as single user, and the sucky mach-task appears to exist at that time, so perhaps it is the kernel losing a thread early on. The questions are: 1) is anyone else seeing this behaviour? 2) does anyone know what's causing the bad mach-task? I believe it has something to do with the initial device configuration, but that's only a hunch. Anyway, these PC machines are different (one is from eCesys, and the other from GEC, one has DPT, the other Adaptec, one has 32M RAM, one has 16M RAM, etc.) and the only thing they have in common is that they're running NEXTSTEP 3.2... Any info is appreciated. I'll summarize. - darcy
From: aman@futon.SFSU.EDU (Aman Khalili) Newsgroups: comp.sys.next.bugs Subject: HELP: Kernel Trap on FPU on NS/I 3.2 Date: 31 Mar 1994 05:48:10 GMT Organization: San Francisco State University Message-ID: <2ndo6q$7u8@news.csus.edu> I have had two instances where at random I have gotten a kernel trap while using the Edit.app. I wasn't doing anything but editing a file. Error: kernel trap floating point execution error ( 3, 0, 0) Total system crash. Couldn't restart, halt, nothing would work. I had to Cold boot the machine. Has anyone else had these kind of problems with Intel boxes? Any help is appreciated.
From: michael@cogito.iaee.tuwien.ac.at (Michael Suessner) Newsgroups: comp.sys.next.bugs Subject: Boot problems with DPT2022 Date: 31 Mar 1994 11:08:06 GMT Organization: Technical University Vienna, Austria Message-ID: <2neaum$gur@email.tuwien.ac.at> I have a brand new DPT 2022 HBA an I cannot boot if there is a disk in a removable drive like CD-ROM or a Syquest drive. The boot process stops mostly with a kernel trap error. It was not possible to install NeXTstep using this card. I have read a mail from a guy having similar boot problems. Windows/DOS works fine with the card, so it seems that there is a BUG in the NS DPT 2012 driver. Is there anybody, who has no troubles with the DPT card? Salute Suessner Michael
From: me@vigor.anatomy.upenn.edu (Joe Panico) Newsgroups: comp.sys.next.bugs Subject: Re: Boot problems with DPT2022 Date: 31 Mar 1994 13:06:26 GMT Organization: University of Pennsylvania Distribution: world Message-ID: <2nehsi$c6b@netnews.upenn.edu> References: <2neaum$gur@email.tuwien.ac.at> In article <2neaum$gur@email.tuwien.ac.at> michael@cogito.iaee.tuwien.ac.at (Michael Suessner) writes: > I have a brand new DPT 2022 HBA an I cannot boot if there is a disk in a removable drive like CD-ROM or a Syquest drive. The boot process stops mostly with a kernel trap error. It was not possible to install NeXTstep using this card. > > I have read a mail from a guy having similar boot problems. > > Windows/DOS works fine with the card, so it seems that there is a BUG in the NS DPT 2012 driver. > > Is there anybody, who has no troubles with the DPT card? > > Salute > Suessner Michael Hi, My configuration: DECpc 466 MTE dpt 2022 attached to dpt: 500Mb SCSI HD, 200MB SCSI HD, NEC CD-84, Syquest Marlin (105MB) It boots no problem, and all SCSI devices work as advertised. Joe Panico joe@retina.anatomy.upenn.edu
From: dekorte@scri.fsu.edu (Steve Dekorte) Newsgroups: comp.sys.next.bugs Subject: Can't print in Portrait format! Date: 31 Mar 1994 16:56:27 GMT Organization: Florida State University Message-ID: <2nevbr$60d@mailer.fsu.edu> I've run into a serious bug. Using Edit, I can not print in Portrait format - only Landscape. Every time I go into Page Layout and set it to Portrait, and hit 'OK' the next time I print or go back into Page Layout, it has been reset! Anyone know how to fix this? Steve Dekorte
From: ccpaulh@monad.missouri.edu (H. Paul Hammann) Newsgroups: comp.sys.next.sysadmin,comp.sys.next.bugs Subject: System accounting problems under Black NS 3.2 Date: 29 Mar 1994 01:48:45 GMT Organization: University of Missouri - Columbia Distribution: World Message-ID: <2n81eb$ieh@golf.ustores.missouri.edu> Hello all, for some reason accounting does not seem to work under NS 3.2. When I run /usr/etc/accton I get no error messages but the file /usr/adm/acct is not created. When I try to specify the file I get an error. "accton: No such file or directory" in either case system accounting is not started. Strange... Any help or hints would be appreciated. -- H. Paul Hammann ccpaulh@monad.missouri.edu
From: theharv@csld.ucr.edu (Brian Harvey) Newsgroups: comp.sys.next.sysadmin,comp.sys.next.bugs Subject: Re: System accounting problems under Black NS 3.2 Followup-To: comp.sys.next.sysadmin,comp.sys.next.bugs Date: 29 Mar 1994 17:01:07 GMT Organization: University of California, Riverside Distribution: World Message-ID: <2n9msj$j7k@galaxy.ucr.edu> References: <2n81eb$ieh@golf.ustores.missouri.edu> H. Paul Hammann (ccpaulh@monad.missouri.edu) wrote: : Hello all, : for some reason accounting does not seem to work under NS 3.2. When I run : /usr/etc/accton I get no error messages but the file /usr/adm/acct is not : created. When I try to specify the file I get an error. "accton: No such file : or directory" in either case system accounting is not started. Strange... Any : help or hints would be appreciated. You need to create the file first. Simply do: touch /usr/adm/acct and then /usr/etc/accton /usr/adm/acct to turn on the accounting. ---------------- Brian Harvey * theharv@csld.ucr.edu University of California, Riverside * brian@cs.ucr.edu C.S.L.D. System Administrator (NeXT Mail preferred everywhere!)
From: gt6963c@prism.gatech.edu (John) Newsgroups: comp.sys.next.bugs,comp.sys.next.sysadmin Subject: Some system problems (NS/Intel) Date: 31 Mar 1994 21:09:15 -0500 Organization: Georgia Institute of Technology Distribution: world Message-ID: <2nfvob$1pa@acme.gatech.edu> I've been having a few problems.. first, every so often I'll boot and my scsi card or cd-rom causes a system panic, and when I reboot the cd-rom isn't recognized.. so I have to deinstall the scsi driver and then reinstall it and reboot.. (just rebooting doesn't work) and then it works for a while. I've got an Adaptek 1542c (not cf) and a Sony CDU-561. (the text boot sequence recognizes the scsi card but not hte cd-rom when it's having this problem). Second.. the since I reinstalled my os from scratch a couple weeks ago, the SoftPC demo wont run. It gets to the browser where you pick a dos or windows setup, and I click ok..then a few minutes later I get a panel that says "This session of SoftPC cannot use full screen display" and has quit or continue buttons.. no matter which I pick, the app exits. Third, and MOST annoying, my dos partition (on the second hard drive) is only valid for the first log in. When I boot up, the first time I log in (no matter what account), I can see my dos partition. If I log out, and log back in to any account (including the same one), I can't find my dos partion, or mount it, or anything. It tells me the device file isn't found, even tho I can do an ls -l on it. (putting an entry into the fstab or the netinfo fstab doesn't seem to matter..and the printed and on-line documentation are out of date to pre-intel hardware). Thanks for any help. John Rudd -- John "Kzin" Rudd gt6963c@prism.getech.edu (ex-kzin@cc.gatech.edu) ===========Intel: Putting the backward in backward compatable.=============== "Tonya, the new cologne for men. One whif makes women weak in the knees" ===========================DOOM(1.2) ON YOU!=================================
From: croehrig@celegans.psych.ubc.ca (Chris Roehrig) Newsgroups: comp.sys.next.bugs Subject: Re: HELP: Kernel Trap on FPU on NS/I 3.2 Date: 1 Apr 1994 02:41:10 GMT Organization: Computer Science, University of B.C., Vancouver, B.C., Canada Message-ID: <2ng1k6$cbb@cs.ubc.ca> References: <2ndo6q$7u8@news.csus.edu> In article <2ndo6q$7u8@news.csus.edu> aman@futon.SFSU.EDU (Aman Khalili) writes: > I have had two instances where at random I have gotten > a kernel trap while using the Edit.app. Make sure that you don't have a PowerSave feature enabled in your computer's CMOS setup (POWER MANAGEMENT FEATURES, or something). If there's no bus activity for a certain amount of time, this will power the CPU down to 8 MHz, and UNIX doesn't seem to like that much. :-) Took me a while to track my random crashes down to this. PC's: ya gotta love 'em! :-/ -- Chris Roehrig (croehrig@celegans.psych.ubc.ca) Invertebrate Learning Group, University of British Columbia, Canada
Newsgroups: comp.sys.next.bugs From: allan@narwal.ali.bc.ca (Allan Noordvyk) Subject: Re: Boot problems with DPT2022 Message-ID: <1994Mar31.200031.1190@narwal.ali.bc.ca> Organization: A.L.I. Technologies Distribution: na Date: Thu, 31 Mar 1994 20:00:31 GMT In article <2neaum$gur@email.tuwien.ac.at> michael@cogito.iaee.tuwien.ac.at (Michael Suessner) writes: > I have a brand new DPT 2022 HBA an I cannot boot if there is a > disk in a removable drive like CD-ROM or a Syquest drive. The boot > process stops mostly with a kernel trap error. It was not possible > to install NeXTstep using this card. > We use DPTs all over the place here and they work fine. There are a couple of gotchas you have to watch out for: 1. Watch your SCSI termination. DOS will just pretend that everything is OK and merrily jumble occasional bits of your data. NS is more discerning. 2. Make sure you have the proper ROMs on the card. Even if it says NeXTSTEP on the outside of the box it ain't necessarily so. Note what is written on your ROMs and give the very friendly people at DPT a call. They will be happy to send you an update. 3. Make sure you set the EISA configuration properly. The controller *MUST* be set at Interrupt 15 *EDGE* and the controller should be set-up as SECONDARY - NON-BOOTABLE (I don't know why -- that's just what the NS docs say). 4. The NS DPT driver does not like devices to issue SCSI time-outs. This does not mean that there has to be a disk in your CD-ROM drive. I do this all of the time. The only time this problem came up here was when using NeXTSTEP to run a WORM disk juke box and the users were ignoring the plaintive beeping coming from the unit (it wanted them to insert a new disk and eventually gave up waiting). Happy hacking... Allan Noordvyk
Newsgroups: comp.sys.next.bugs,comp.sys.next.misc From: boonlow@kits.sfu.ca (Boon Chong Benjamin Low) Subject: USR wouldn't do 14.4 transfer Message-ID: <boonlow.765224146@sfu.ca> Sender: news@sfu.ca (seymour news) Organization: Simon Fraser University, Burnaby, B.C., Canada Date: Fri, 1 Apr 1994 18:15:46 GMT Hey there, I have a USR 14,400 external fax modem and I am having tons's of problem from it under NS. Whenever I tried any serial communication, things like serial loging or SLIP connection, I will keep losing characters if I set the baudrate on the communication over 9600 bps. MY screen will get all screwed up, characters will be lost e.t.c I tried Modulator, TipTop and set the baudrate above 9600 bps, all the problem will be there if the baudrate is set greater than 9600bps. Could it be NS doesn't support my modem or SW problems ? I have an Intel 486 with LocalBus motherboard. However, the modem is external and my I/O controller card is only ISA and not VESA. According to be friend, this might constitute to the problem of losing characters and if I get a VESA controller card with the UART 16500 chip, my problem would go away. BTW, I have a UART 16500 chip already on my present ISA controller/serial card. I am running NS3.2 on a Intel machine. Only my ATI video card is VESA, the rest are all ISA. I tried to asked around, my sysadmin e.t.c but no people seem to be able to help me. I have no problems with running at 14.4 under DOS or Windows, only NS. Therefore, I would really appreciate it if someone can take the time to drop me some adive/help. Thanks. - Ben.
From: samurai@wsc.com (Darcy Brockbank) Newsgroups: comp.sys.next.bugs Subject: mach-task solved? Date: 1 Apr 1994 20:00:30 GMT Organization: WSC Investment Services, Inc. Disclaimer: The views represented within this posting are not necessarily those held by WSC Investment Services, Inc. Message-ID: <2nhugu$be3@cerberus.wsc.com> I got word from someone (who didn't say if I should/should not mention them, so I won't) that NeXT says: 1) all NS-Intel machines show the seemingly resource sucking mach-task... and it shows up before single-user boot finishes. 2) It's not really sucking up resources. 3) Size is always reported at Real RAM + 16M 4) NeXT says it is innocuous (um, so I guess they can get rid of this "feature" then ;-). 5) Don't worry about that Radon gas seeping into your house... Well, they didn't *really* say #5... that's supposed to be a joke :-). - darcy
Newsgroups: comp.sys.next.bugs From: newman@string.harvard.edu (Michael Newman) Subject: Re: mach-task solved? Message-ID: <Cnqz47.Avx@das.harvard.edu> Sender: usenet@das.harvard.edu Organization: Division of Applied Sciences, Harvard University References: <1994Apr2.104010.6728@seer.demon.co.uk> Date: Mon, 4 Apr 1994 18:24:06 GMT > > If so, why do they sometimes claim 50%+ CPU? (and yes, they use it!). > > Paul I thought I remembered seeing someone post something once about the mach-task's being something to do with the serial port....(don't ask me what that means) On our black hardware, I have never seen any of them use ANY of the CPU time, never mind 50%. Is this a difference between black and white? Fortunately our machines never stay up long enough for too many mach-task's to accumulate. We are always managing to crash them somehow :@) Mike Newman
From: mow@marsu.tynet.sub.org (Markus Wenzel) Newsgroups: comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.hardware Subject: EATA timeout report #2 Date: 4 Apr 1994 21:21:44 +0100 Organization: Palumbian Research Labs Message-ID: <2npsso$9p@marsu.tynet.sub.org> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit EATA timeout report #2 ====================== Description of the problem: --------------------------- The DPT EISA controllers 2022/9x and 2012/9x cause disastrous crahes under NeXTSTEP Intel 3.x. These crashes are initiated by an "EATA timeout" error message in /usr/adm/messages, and a few seconds later any disk access on the SCSI bus is blocked. If this happens when file system relevant information is written to disk, the resulting damage to the filesystem is often so big that a complete restore from tape is the only way left. Equivalent crashes were reported by: ------------------------------------ Rene Kulschewski <rene@rkt.in-berlin.de> martin@cube.han.de (Martin Ortlepp) Rainer Staringer <rainer@fml.tuwien.ac.at> John B. Hampshire II <hamps@richibucto.jpl.nasa.gov> ramesh@bodhi.esys.cwru.edu (Ramesh Dodamani) Varun Mitroo <mitroo@magnus.acs.ohio-state.edu> Milo Velimirovic <milov@fingers.acs.uwlax.edu> ivo@next.agsm.ucla.edu (Ivo Welch) "Thomas J. Herbert" <THERBERT@umiami.ir.miami.edu> Johann Adalbert Stoeger <hans@tms-gmbh.de> Markus Wenzel <mow@marsu.tynet.sub.org> New information since report #1: -------------------------------- Martin Ortlepp and I tested new BIOS and firmware revisions from DPT. We got SmartROM version 002-D and BIOS version 005E. Martin had one EATA timeout after about one week of testing which could be recovered. I set all EISA parameters according to NA document #1457, cache to write-through and transfer rate to 10 MB/s. The first EATA timeout occurred three days later. Then I reduced transfer rate to 5 MB/s and hoped once more for the best. The next EATA timeout occurred four days later... Fortunately, both timeouts didn't do any damage to my filesystem, mainly because I pressed AltGr-Num quickly enough. Both timeouts happened immediately after login. A sysadmin from the University of Berlin told me that´he administrates about 20 NS/I workstations, and also has DPT hassles. But he said it depends on the machine in which he plugs the same DPT controller. Some work, some don't. Conclusions from all this: -------------------------- 1. One year after NeXTSTEP/Intel 3.1 PR1 there is only _ONE_ EISA controller supported by NeXT as an EISA controller. All others (Adaptec, BusLogic) only work in 1542 compatibility mode. 2. This one and only EISA combination 2012/2022/2122 and NS/I DPT driver does not work reliably. It must be considered even dangerous to run a NS/I server machine with this combination. Severe data loss is possible. 3. A NeXT representative from NeXT Europe told me at CeBit expo that they also had problems with DPT controllers at NeXT Corp. Their consequence was to use Adaptecs for the machines presented at the expo. 4. DPT was very helpful and responsive, but they just don't know what the reason is. I think they were not able to reproduce the error situation. 5. I have not heard anything from NeXT if someone is busily working on it. 6. PC hardware is bullshit. 7. I'm sick of being a beta tester for device drivers approved in the hardware combatability guide!!! I lost money, data, confidence and patience. Regards, Markus. -- .sig got a SIGKILL signal.
Newsgroups: comp.sys.next.bugs From: dabbah@tron.bwi.wec.com (Albert Dabbah) Subject: Re: mach-task solved? Message-ID: <1994Apr5.060100.12537@tron.bwi.wec.com> Organization: Westinghouse Electric Corporation, Baltimore MD. References: <1994Apr2.104010.6728@seer.demon.co.uk> <Cnqz47.Avx@das.harvard.edu> Date: Tue, 5 Apr 1994 06:01:00 GMT In article <Cnqz47.Avx@das.harvard.edu> newman@string.harvard.edu (Michael Newman) writes: > >> >> If so, why do they sometimes claim 50%+ CPU? (and yes, they use it!). >> >> Paul > >I thought I remembered seeing someone post something once about the >mach-task's being something to do with the serial port....(don't ask me >what that means) > >On our black hardware, I have never seen any of them use ANY of the CPU >time, never mind 50%. Is this a difference between black and white? > >Fortunately our machines never stay up long enough for too many >mach-task's to accumulate. We are always managing to crash them >somehow :@) > > >Mike Newman I had it happen on my cube last week. I am running version 3.2. And that was the only time I have seen it occur. Albert -- dabbah@tron.wec.com ! #include <standard_disclaimer>
From: chris@helser54.res.iastate.edu (Chris Wong) Newsgroups: comp.sys.next.bugs,comp.sys.next.misc,comp.sys.next.software Subject: NS/FIP 3.2 can't read Mac disk. Date: 5 Apr 1994 06:44:12 GMT Organization: Iowa State University, Ames, Iowa Message-ID: <2nr1bs$k5i@news.iastate.edu> Keywords: Mac, NeXTSTEP, disk How on the earth can I get NS/FIP 3.2 to read Mac disks? Please help. It reads the Mac disks formatted by itself. But, it can't read the Mac disks formatter by a Mac. Any clue? Chris
From: colbert1@news.delphi.com (COLBERT13@DELPHI.COM) Newsgroups: comp.sys.next.bugs,comp.sys.next.misc Subject: Re: USR wouldn't do 14.4 transfer Date: 5 Apr 1994 01:05:17 -0400 Organization: Delphi Internet Services Corporation Message-ID: <2nqrid$re@news.delphi.com> References: <boonlow.765224146@sfu.ca> boonlow@kits.sfu.ca (Boon Chong Benjamin Low) writes: >Hey there, > I have a USR 14,400 external fax modem and I am having tons's of problem from it under NS. Whenever I tried any serial communication, things like >serial loging or SLIP connection, I will keep losing characters >if I set the baudrate on the communication over 9600 bps. MY screen will get all screwed up, characters will be lost e.t.c I tried Modulator, TipTop >and set the baudrate above 9600 bps, all the problem will be there if the > baudrate is set greater than 9600bps. >Could it be NS doesn't support my modem or SW problems ? >I have an Intel 486 with LocalBus motherboard. However, the modem is >external and my I/O controller card is only ISA and not >VESA. According to be friend, this might constitute to the problem of >losing characters and if I get a VESA controller card with the >UART 16500 chip, my problem would go away. BTW, I have a UART 16500 chip >already on my present ISA controller/serial card. >I am running NS3.2 on a Intel machine. Only my ATI video card is VESA, >the rest are all ISA. >I tried to asked around, my sysadmin e.t.c but no people seem to be able to >help me. I have no problems with running at 14.4 under DOS or Windows, only >NS. >Therefore, I would really appreciate it if someone can take the time to >drop me some adive/help. >Thanks. >- Ben. The root o your problem dates back to the original IBM pc . they originaly used a 8250 uart to control serial transmissions, whic required polling by the CPU to avoid overrun. The 8250 and the later 440 (both still commonly used) provide no buffer, onot a problem at up to 9600 bps, but tends to lose data under certain conditions ie. when the cpu has many other tasks it has to perform. I Havn't heard of a local bus serial port, kindda sounds like overrkill. a ISA bus w/ a >2 Mb transfer should be more than enough, but a serial board with a 550 should solve yuor problem, if you cold find one NeXT will work with. remember 55 saves lives ;) colbert13@delph.com no ascii sig couldn't think of anything to say
From: eric@skatter.usask.ca Newsgroups: comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.hardware Subject: Re: EATA timeout report #2 Date: 5 Apr 1994 15:46:51 GMT Organization: University of Saskatchewan Message-ID: <2ns15b$iu3@tribune.usask.ca> References: <2npsso$9p@marsu.tynet.sub.org> > 2. This one and only EISA combination 2012/2022/2122 and NS/I DPT driver > does not work reliably. It must be considered even dangerous to run > a NS/I server machine with this combination. Severe data loss is > possible. > I'd like to point out that `your mileage may vary'. I have been running a DPT 2022 (with no cache) for 3 months now, heavy usage, and have had exactly 0 of these EATA Timeouts. -- Eric Norum eric@skatter.usask.ca Saskatchewan Accelerator Laboratory University of Saskatchewan Saskatoon, Canada. NeXTMail accepted.
From: knobi@.architektur.uni-kassel.de (Frank Knobloch) Newsgroups: comp.sys.next.bugs Subject: Login-problems on an Intel with NeXT-Step 3.2 Date: 5 Apr 1994 16:38:28 GMT Organization: IAG-CAD Zentrum der UGH Kassel Distribution: world Message-ID: <2ns464$cs2@hrz-ws11.hrz.uni-kassel.de> Hi! I get in trouble after the installation of NeXT-Step 3.2 on an Intel. The Intel-machine is a Netinfo client and a origin NeXT works as a Netinfo-server. The origin NeXT mounts a /home partition from a Sun 4/470 In these partitions are the accounts. All accounts are installed in the /-domain and there is no problem to login on the NeXT. But when i try to login on the Intel nothing happens. I entered the password and the wheel is spinning arround maybe till christmas. When i login (on the Intel) as root its no problem to create a new account on /home/.... But its also impossible to login with this new created account on the Intel. Its no problem on the NeXT. I'm really appreciated for any kind of help! Ciao Knobi
From: next2@info2.rus.uni-stuttgart.de (Markus Wenzel) Newsgroups: comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.hardware Subject: Re: EATA timeout report #2 Date: 5 Apr 94 20:53:16 GMT Organization: Comp.Center (RUS), U of Stuttgart, FRG Message-ID: <next2.765579196@info2.rus.uni-stuttgart.de> References: <2npsso$9p@marsu.tynet.sub.org> <2ns15b$iu3@tribune.usask.ca> eric@skatter.usask.ca writes: >I'd like to point out that `your mileage may vary'. I have been running >a DPT 2022 (with no cache) for 3 months now, heavy usage, and have had >exactly 0 of these EATA Timeouts. Just be happy about it. But I think more than 10 persons with similar problems are enough to point out that something has to be done against it. If your system were unstable, you'd know what I meant with 'unreliable'. BTW: What devices did you connect to your 2022? Fast-SCSI-II harddisk drives? My theory is that it only happens with new fast drives. Regards, Markus. -- Markus Wenzel System administration, Consulting, Networking mow@marsu.tynet.sub.org on... NeXTSTEP / Unix / Novell / Windows NT IRC: Marsu Expert in quantum bogodynamics
Newsgroups: comp.sys.next.bugs From: cousens@biztech.com Subject: Re: BackSpace bug? Message-ID: <1994Apr1.132032.26012@biztech.com> Sender: news@biztech.com Organization: Biztech, Inc. References: <2naadu$5bn@news.duke.edu> Date: Fri, 1 Apr 1994 13:20:32 GMT Denise, Yes, I concur. Back in the 2.2 days, I saw this behavior, also. Does it happen in 3.2 too? I haven't run Backspace in so long.... I agree with Shawn (sas@valmar.netcom.com)... If you see the problem occurring in 3.2, drop me a line, and I'll have a look at it. Rik Cousens (cousens@biztech.com) DIn article <2naadu$5bn@news.duke.edu> blake015@mc.duke.edu(Denise Blakeley) writes: > Has anyone else noticed this in BackSpace v1.02? > > I have it set to be a screen locker via BackSpace's own preferences, so of > course this means I had to make up a password. > > I launch BackSpace and set it to screen-locking. Later, I bump the mouse so > that it prompts me for the password. But if I just sit there for a few seconds > without doing anything, it unlocks the screen anyway! > > Denise > -- > Denise Blakeley | PROGRAM, tr. v., An activity similar > Duke Med Center Info Systems | to banging one's head against a wall, > Durham, NC | but with fewer opportunities for > (919) 286-6468 W | reward. > blake015@mc.duke.edu | NeXTMail welcome!
From: mow@marsu.tynet.sub.org (Markus Wenzel) Newsgroups: comp.sys.next.bugs Subject: Re: Boot problems with DPT2022 Date: 2 Apr 1994 11:56:58 +0100 Organization: Palumbian Research Labs Distribution: na Message-ID: <2njj1q$40k@marsu.tynet.sub.org> References: <1994Mar31.200031.1190@narwal.ali.bc.ca> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit allan@narwal.ali.bc.ca (Allan Noordvyk) writes: >We use DPTs all over the place here and they work fine. There are >a couple of gotchas you have to watch out for: Lucky man! I still suffer from the EATA timeout problem. >2. Make sure you have the proper ROMs on the card. Even if it > says NeXTSTEP on the outside of the box it ain't necessarily > so. Note what is written on your ROMs and give the very > friendly people at DPT a call. They will be happy to send > you an update. The update to 2D/5E made the timeouts occur less frequently, but it did not cure the problem completely. >4. The NS DPT driver does not like devices to issue SCSI > time-outs. This leads to the question what one can do to keep one's SCSI devices from issueing these goddamn' timeouts!! I think the experts at NeXT and DPT should stick together for some weeks and _solve_ the EATA timeout bug, finally!! And they should not keep the result inhouse until 3.3 release date! More news in EATA timeout report #2, after my testing cycle is finished. Regards, Markus. -- .sig got a SIGKILL signal.
From: xcarn@mcmuse.mc.maricopa.edu (Michael Carnright) Newsgroups: comp.sys.next.bugs Subject: help with computer rebooting Date: 2 Apr 1994 16:03:11 GMT Organization: Maricopa Community Colleges Message-ID: <2nk4vv$47t@names.maricopa.edu> I work with several Black NeXT computers that are networked together. On several of these when they are powered down they reboot. Is there a preferance I can change to keep them from automaticly rebooting? A second question I have is that on several of the login windows there is a phone icon. They are not set up to use this application (I think). Why has this Icon appeared? Does it have something to do with the Phone connector app. ? Thank you in advance for any help you can give me on these problems.
Newsgroups: comp.sys.next.bugs From: flash!jon@myxa.com Subject: Re: BackSpace bug? Message-ID: <Cnn53x.8q@dsinc!flash> Sender: jon@dsinc!flash (Jonathan Hendry) Organization: Who Needs It? References: <1994Apr1.132032.26012@biztech.com> Date: Sat, 2 Apr 1994 16:43:08 GMT I think this happens when either you or root has a null password. Both your own password and the root password will unlock the screen, in addition to your Backspace password. Just make sure both accounts have a password. "It's not a bug, it's a feature" >In article <1994Apr1.132032.26012@biztech.com> cousens@biztech.com writes: > Denise, > > Yes, I concur. Back in the 2.2 days, I saw this behavior, also. > Does it happen in 3.2 too? I haven't run Backspace in so long.... > I agree with Shawn (sas@valmar.netcom.com)... If you see the problem > occurring in 3.2, drop me a line, and I'll have a look at it. > > Rik Cousens (cousens@biztech.com) > > DIn article <2naadu$5bn@news.duke.edu> blake015@mc.duke.edu(Denise > Blakeley) writes: > > Has anyone else noticed this in BackSpace v1.02? > > > > I have it set to be a screen locker via BackSpace's own preferences, so > of > > course this means I had to make up a password. > > > > I launch BackSpace and set it to screen-locking. Later, I bump the > mouse so > > that it prompts me for the password. But if I just sit there for a few > seconds > > without doing anything, it unlocks the screen anyway! > > > > Denise -- Jonathan W. Hendry Inexpensive NeXTSTEP Consulting tjhendry@mcs.drexel.edu For Your "Not-So-Mission-Critical" Apps
From: james@aleph.ee.ualberta.ca (James Strohschein) Newsgroups: comp.sys.next.bugs,comp.sys.next.software Subject: Icon patch? Date: 6 Apr 1994 19:55:27 GMT Organization: Computer and Network Services, U of Alberta, Edmonton, Canada Message-ID: <2nv43f$imc@quartz.ucs.ualberta.ca> Keywords: Icon, patch Does anybody know where can I find a patch for Icon? I have the Release 2.0 Icon, and want to run it under NS 3.2. Thanks. -- James Strohschein james@aleph.ee.ualberta.ca (NeXTMail) Electrical Engineering (403) 492-4935 University of Alberta (403) 492 -1181 (fax)
Newsgroups: comp.sys.next.bugs From: brad@instep.wimsey.bc.ca Subject: Re: mach-task solved? Message-ID: <1994Apr6.192900.20997@instep.wimsey.bc.ca> Sender: brad@instep.wimsey.bc.ca (Bradley Head) Organization: InStep Mobile Communications Inc. References: <Cnqz47.Avx@das.harvard.edu> Date: Wed, 6 Apr 1994 19:29:00 GMT In article <Cnqz47.Avx@das.harvard.edu> newman@string.harvard.edu (Michael Newman) writes: > I thought I remembered seeing someone post something once about the > mach-task's being something to do with the serial port....(don't ask me > what that means) > > On our black hardware, I have never seen any of them use ANY of the CPU > time, never mind 50%. Is this a difference between black and white? > Speaking of serial port and cpu usage. Look no further than getty. Since NS 3.2, if you have a getty entry turned on for one of the serial ports, but don't have nay device attached, getty will consume upwards of 50% of the cpu. brad. -- Bradley Head Software Developer, InStep Mobile Communications Inc. brad@instep.wimsey.bc.ca (NeXTmail accepted) 604 872-7116 fax: 604 872-7125
Newsgroups: comp.sys.next.bugs From: sr@rdbois.fdn.org (serge_ruby) Subject: Re: USR wouldn't do 14.4 transfer Message-ID: <1994Apr3.210120.9213@rdbois.uucp> Sender: sr@rdbois.uucp (serge_ruby) Organization: S.RUBY References: <boonlow.765224146@sfu.ca> Date: Sun, 3 Apr 1994 21:01:20 GMT In article <boonlow.765224146@sfu.ca> boonlow@kits.sfu.ca (Boon Chong Benjamin Low) writes: > > Hey there, > > I have a USR 14,400 external fax modem and I am having tons's of problem from it under NS. Whenever I tried any serial communication, things like > serial loging or SLIP connection, I will keep losing characters > if I set the baudrate on the communication over 9600 bps.> Therefore, I would really appreciate it if someone can take the time to > drop me some adive/help. > I don't know on NS for Intel processors, but if it is the same as with NS for Motorola processor you can use speed over 9600 bps only if you use hardware handshaking, ie device cufa or cufb instead of cua or cub. Hope this helps Serge
From: sanguish@digifix.com Newsgroups: comp.sys.next.announce,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.bugs,comp.sys.next.programmer,comp.sys.next.marketplace Subject: Quick Guide to NEXTSTEP information on the Internet Date: 4 Apr 1994 00:51:30 -0400 Organization: Next Announcements Message-ID: <2no6ci$8bn@digifix.digifix.com> This post is made weekly, to help 'point' users to more NEXTSTEP information Topics include: NEXTSTEP World Wide Web Product Information Server online comp.sys.next newsgroups related newsgroups comp.sys.next newsgroups mailing list ftp sites NeXTanswers New Information --------------- NEXTSTEP World Wide Web Product Information Server online A product directory built around the World Wide Web system, this will allow full multimedia announcements by NEXTSTEP developers, as well as the ability to browse the available products for NEXTSTEP. This service is online now, and can be reached at http://digifix.digifix.com/ it can be reached using OmniWeb (available from ftp.omnigroup.com) or Mosaic. The entries are coming in quite quickly, and I'll be moving the entire NEXTSTEP Third Party Catalog contents in as soon as they become available. Additionally the NEXTSTEP Product Information Mail Server will be stocked full file files in the next week... you can get information on using the mail server at ns-products@digifix.com Suggestions or comments can be directed to me at sanguish@digifix.com comp.sys.next.* newsgroups -------------------------- Comp.Sys.Next.Advocacy This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. Comp.Sys.Next.Announce Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) The NEXTSTEP FAQs are posted here monthly as well. This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Comp.Sys.Next.Bugs A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT- specific groups as well. Comp.Sys.Next.Hardware Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. Comp.Sys.Next.Marketplace NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. Comp.Sys.Next.Misc For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! Comp.Sys.Next.Programmer Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. The NEXTSTEP programmer FAQs are posted here. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. Comp.Sys.Next.Software This is a place to talk about [third party] software products that run on NEXTSTEP systems. Comp.Sys.Next.Sysadmin Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. related Newsgroups ------------------ Comp.Soft-Sys.Nextstep Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. Comp.Lang.Objective-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. Comp.Object Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original Comp.Sys.Next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News ------------------------------------------- Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep-d next-advocacy-d next-announce-d next-bugs-d next-hardware-d next-marketplace-d next-misc-d next-programmer-d next-software-d next-sysadmin-d (For a full description, send mail saying LISTS to <digestif@antigone.com>). The subscription syntax is essentially the same as LISTSERV's. To subscribe, send a message to <digestif@antigone.com> saying: SUB Listname YourName Example: SUB next-hardware-d John Doe The ftp sites ------------- cs.orst.edu: The main site for North American submissions nova.cc.purdue.edu: Lots of older stuff, but very short on disk space ftp.informatik.uni-muenchen.de: In Germany. terra.stack.urc.tue.nl (Dutch NEXTSTEP User Group) and cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp.next.com: See the below ftp.next.com and NextAnswers@next.com ------------------------------------- From the document 1000_Help from ftp.next.com Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, or you can transfer them by anonymous ftp. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to NeXTanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO <your-address> If you have any problem with the system or suggestions for improvement, please send mail to NeXTanswers-request@NeXT.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to NeXTanswers-request@NeXT.com. Written by: Eric P. Scott eps@toaster.SFSU.EDU and Scott Anguish sanguish@digifix.com Additions from: Greg Anderson (Greg_Anderson@afs.com) and Michael Pizolato (Michael_Pizolato@afs.com)
Newsgroups: comp.sys.next.bugs,comp.sys.next.hardware From: allan@narwal.ali.bc.ca (Allan Noordvyk) Subject: Minimizing problems with DPT SCSI controllers Message-ID: <1994Apr5.233653.1057@narwal.ali.bc.ca> Organization: A.L.I. Technologies Distribution: na Date: Tue, 5 Apr 1994 23:36:53 GMT In response to my earlier reply to a posting by a chap who couldn't get his sytem to boot with a DPT 2022 SCSI adapter in it, some people posted and e-mailed me indicating that they were still having intermittent problems when running NS/I with this controller and were suprised by the lack of problems I claimed. Anyhow, here is the some hard-won wisdom on the subject using these dang things with fewer problems. Note that I said fewer problems -- I would say none, but sometimes its hard to know which hardware or software component is responsible for the once in a blue moon system lock-up. First off, NeXT recommends disabling the Command Queueing option in the DPT's EISA configuration. I think this is to protect you from devices that don't quite do this properly. It may be superstition to do this on all machines -- but why tempt fate ;-) One thing you should be sure to do is DISABLE the Command Time-Out option of the DPT in its EISA configuration. Otherwise, if one of your hard-drives is taking too long to execute a DASD read or write command, the controller will reset the SCSI bus and retry the command. This will seriously confuse the OS. This is especially true if you have multiple SCSI devices like me (ie. at least two hard-drives going simultaneously most of the time). Although I don't seem to have any serious DPT problems with my home and office NS/I machines, our tech-support people here haven't completely managed to eliminate glitches from our customer systems and thus we are switching to using the new Bustek 747 controllers on all new shipping systems. I think that the customer's DPTs are having more problems than our in-house machines because our customer systems frequently have Iomega Bernoulli drives and/or external WORM jukeboxes on long cables. The Iomega's have an awful SCSI implementation which gave me no end of trouble when I had one in my box. I was more than glad to be rid of the damn thing after a couple of weeks. The WORM jukeboxes seem to introduce cabling noise which the DPT's don't like too much. My two machines are quite stable even when I use them with a portable external SCSI-1 drive. I keep the cable short though, and have external cable detection turned ON in the EISA config. Termination-wise, I always put the terminating internal hard-drive at the end of the cable, rather than at the last used connector. Again this may be superstition, but it works for me. Since I have external cable detection turned on (and thus will only run at 5 MB/s when one is detected) I always remove my external devices when I don't expect them to be used for a while and then terminate the non-SCSI-2 connectors with an external terminator. Since my only external device is my portable hard-drive and I only use this as a digital briefcase to take work home, this hasn't been a big hardship. Your situation may be different. Also VERY IMPORTANT, make sure you are set to be using a WRITE-THROUGH cache in the EISA configuration. The default of WRITE-BACK is unsafe for any unix system without a journaled file-system (BOY WOULD I LIKE A JOURNALED FILE-SYSTEM! NeXT?! VERITAS?! ANYONE?!). I also like to up the read-ahead to 80% as well. Hope this helps.... Allan Noordvyk, Software Artisan ALI Technologies
From: nemiroal@bus.orst.edu Newsgroups: comp.sys.next.bugs,comp.sys.next.hardware Subject: Re: Minimizing problems with DPT SCSI controllers Date: Wed, 6 Apr 1994 22:58:10 Organization: College of Business, Oregon State University, Corvallis Distribution: na Message-ID: <nemiroal.112.0016F8E4@bus.orst.edu> References: <1994Apr5.233653.1057@narwal.ali.bc.ca> In article <1994Apr5.233653.1057@narwal.ali.bc.ca> allan@narwal.ali.bc.ca (Allan Noordvyk) writes: >From: allan@narwal.ali.bc.ca (Allan Noordvyk)>Subject: Minimizing problems with DPT SCSI controllers >Date: Tue, 5 Apr 1994 23:36:53 GMT >In response to my earlier reply to a posting by a chap who couldn't get his >sytem to boot with a DPT 2022 SCSI adapter in it, some people posted and >e-mailed me indicating that they were still having intermittent problems >when running NS/I with this controller and were suprised by the lack of >problems I claimed. Anyhow, here is the some hard-won wisdom on the >subject using these dang things with fewer problems. Note that I said >fewer problems -- I would say none, but sometimes its hard to know which >hardware or software component is responsible for the once in a blue moon >system lock-up. When I upgraded the ROMs, my random boot-up problems went away. Something you might want to try.
From: therbert@umiami.ir.miami.edu Newsgroups: comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.hardware Subject: Re: EATA timeout report #2 Message-ID: <1994Apr5.214338.17359@umiami.ir.miami.edu> Date: 5 Apr 94 21:43:38 EDT References: <2npsso$9p@marsu.tynet.sub.org> <2ns15b$iu3@tribune.usask.ca> Organization: Univ of Miami IR In article <2ns15b$iu3@tribune.usask.ca>, eric@skatter.usask.ca writes: >> 2. This one and only EISA combination 2012/2022/2122 and NS/I DPT driver >> does not work reliably. It must be considered even dangerous to run >> a NS/I server machine with this combination. Severe data loss is >> possible. >> > I'd like to point out that `your mileage may vary'. I have been running > a DPT 2022 (with no cache) for 3 months now, heavy usage, and have had > exactly 0 of these EATA Timeouts. > > -- > Eric Norum eric@skatter.usask.ca > Saskatchewan Accelerator Laboratory > University of Saskatchewan > Saskatoon, Canada. NeXTMail accepted. As reported to some of the users having problems - I have NO EATA timeouts even with heavy usage as a server AS LONG AS I RUN ONLY A SINGLE SCSI DEVICE - THE INTERNAL DEC 1.5GB DRIVE. If I run either my Wang DAT or my Sony CD drive or both, I get a timeout every 7-10 days. In almost six months of total operation, 4.5 of which were with two or more SCSI devices, I have had maybe 15 timeouts, of which only one caused serious problems - In that case I had a panic and minor filesystem damage which was repaired by a manual fsck and replacement of a link. Still, considering the number of similar reported problems and the fact that this problem seems to occur on just the kind of machines which are likely to be network servers, I am very nervous - Even if one out of 10 or even 100 such users suffers panics or filesystem damage, the problem is serious. Yes, there seems to be a lot of variation in frequency among users but, I view the timeout problem as one of the most serious I have encountered with Unix or NeXTstep systems - An erratic server is dangerous and I worry - Not to the extent of staying awake all night but nevertheless I am very concerned. Thomas J. Herbert Department of Biology University of Miami
From: csmith@blackplague.gmu.edu (Christian Smith) Newsgroups: comp.sys.next.bugs,comp.sys.next.misc,comp.sys.next.software Subject: Re: NS/FIP 3.2 can't read Mac disk. Followup-To: comp.sys.next.bugs,comp.sys.next.misc,comp.sys.next.software Date: 6 Apr 1994 11:45:30 GMT Organization: George Mason University, Fairfax, Virginia, USA Message-ID: <2nu7cq$f3g@portal.gmu.edu> References: <2nr1bs$k5i@news.iastate.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Chris Wong (chris@helser54.res.iastate.edu) wrote: : How on the earth can I get NS/FIP 3.2 to read Mac disks? : Please help. : It reads the Mac disks formatted by itself. But, it can't read the Mac : disks formatter by a Mac. Is it possible that the Mac disks formatted by the Mac are not formatted high density? I.e., if the Mac does not have a HDSD floppy drive then it will format HD disks as DD disks (800K instead of 1.44M). Other then this, you should have no trouble reading Mac formatted disks on an NSI 3.2 machine. -- Christian Smith aka Blackplague PGP Public Key available by finger or request.
From: csmith@blackplague.gmu.edu (Christian Smith) Newsgroups: comp.sys.next.bugs Subject: Re: Login-problems on an Intel with NeXT-Step 3.2 Date: 6 Apr 1994 11:50:55 GMT Organization: George Mason University, Fairfax, Virginia, USA Distribution: world Message-ID: <2nu7mv$f3g@portal.gmu.edu> References: <2ns464$cs2@hrz-ws11.hrz.uni-kassel.de> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Frank Knobloch (knobi@.architektur.uni-kassel.de) wrote: : I get in trouble after the installation of NeXT-Step 3.2 on an Intel. : the Intel-machine is a Netinfo client and a origin NeXT works as a : Netinfo-server. The origin NeXT mounts a /home partition from a : Sun 4/470 In these partitions are This is most likely the problem. I think you should try mounting the home directory directly from the SUN to the Intel (I assume you are trying to mount it to the Intel from the NeXT Net-Info server). Use NFS Manager to fix this, and don't forget to give the Intel machine mounting privs on the Sun. -- Christian Smith aka Blackplague PGP Public Key available by finger or request.
From: next2@info2.rus.uni-stuttgart.de (Markus Wenzel) Newsgroups: comp.sys.next.bugs,comp.sys.next.hardware Subject: Re: Minimizing problems with DPT SCSI controllers Date: 7 Apr 94 20:01:25 GMT Organization: Comp.Center (RUS), U of Stuttgart, FRG Distribution: na Message-ID: <next2.765748885@info2.rus.uni-stuttgart.de> References: <1994Apr5.233653.1057@narwal.ali.bc.ca> allan@narwal.ali.bc.ca (Allan Noordvyk) writes: [... A lot of good advice omitted ...] I disabled "Timeout enable" in EISA config. I removed my SyQuest drive from the SCSI chain because it behaved as bad as the mentioned Bernoulli drives. I have no external devices and a rather short internal cable. I set the transfer rate to 5 MB/s. I have active termination on the last internal hard disk. I set the cache option to "Write back". I still got EATA timeouts. What else could I do? Shouldn't it be possible for the NeXT DPT driver to detect these timeouts and react in a certain manner? Regards, Markus. -- Markus Wenzel System administration, Consulting, Networking mow@marsu.tynet.sub.org on... NeXTSTEP / Unix / Novell / Windows NT IRC: Marsu Expert in quantum bogodynamics
From: chris@helser54.res.iastate.edu (Chris Wong) Newsgroups: comp.sys.next.bugs Subject: Re: Minimizing problems with DPT SCSI controllers Date: 7 Apr 1994 21:31:14 GMT Organization: Iowa State University, Ames, Iowa Message-ID: <2o1u32$r78@news.iastate.edu> References: <next2.765748885@info2.rus.uni-stuttgart.de> In article <next2.765748885@info2.rus.uni-stuttgart.de> next2@info2.rus.uni-stuttgart.de (Markus Wenzel) writes: >:) I still got EATA timeouts. What else could I do? >:) Shouldn't it be possible for the NeXT DPT driver to detect these timeouts >:) and react in a certain manner? Make sure that you have the latest ROM on your DPT card. Some people have problems using DPT with NS with the old ROM. After they upgraded the ROM, everything is fine. I never have any problem since I plug the board in. I guess I have the latest ROM. Chris
Newsgroups: comp.sys.next.bugs From: norton@cmf.nrl.navy.mil (David Norton) Subject: Another BackSpace bug Message-ID: <Cnr67F.3oI@ra.nrl.navy.mil> Sender: usenet@ra.nrl.navy.mil Organization: Naval Research Lab, Washington, DC Distribution: w Date: Mon, 4 Apr 1994 20:57:14 GMT If BackSpace has blanked the screen, and the Filemover then pops up an error message (such as "disk full" - stop; proceed after makeing room; or "file is a link" - copy, cancel, link ...) BackSpace will no longer read the keystrokes, so you have no way of entering a password to get back to the system. You can log in from elsewhere, and kill BackSpace, but the you lose the cursor. Has anyone else seen this? know of a way around it? thanks -dave
Newsgroups: comp.sys.next.bugs From: rfi@jeans.fokus.gmd.de (Robert Fischer) Subject: Re: Icon patch? Message-ID: <1994Apr8.065534.24499@fokus.gmd.de> Sender: news@fokus.gmd.de (News system) Organization: GMD-Fokus References: <2nv43f$imc@quartz.ucs.ualberta.ca> Date: Fri, 8 Apr 1994 06:55:34 GMT In article <2nv43f$imc@quartz.ucs.ualberta.ca> james@aleph.ee.ualberta.ca (James Strohschein) writes: > Does anybody know where can I find a patch for Icon? I have the Release > 2.0 Icon, and want to run it under NS 3.2. > > Thanks. > > -- > James Strohschein james@aleph.ee.ualberta.ca (NeXTMail) > Electrical Engineering (403) 492-4935 > University of Alberta (403) 492 -1181 (fax) Hi, I'll mail it to you, if you don't got it yet. It's really buggy, but for little work its good enough. BTW: We have bought Appsoft Image and this is buggy as well. Are there any versions newer than 1.0 (Yes, I know that Appsoft has gone)? Robert. ------------------------------------------------------------------------ Robert Fischer @ GMD-Fokus -------- __o ------- _`\<,_ fischer@fokus.gmd.de ------- (*)/ (*) ## NeXT-Mail welcome ## ------------------------------------------------------------------------
From: next2@info2.rus.uni-stuttgart.de (Markus Wenzel) Newsgroups: comp.sys.next.bugs Subject: Re: Minimizing problems with DPT SCSI controllers Date: 8 Apr 94 10:17:23 GMT Organization: Comp.Center (RUS), U of Stuttgart, FRG Message-ID: <next2.765800243@info2.rus.uni-stuttgart.de> References: <next2.765748885@info2.rus.uni-stuttgart.de> <2o1u32$r78@news.iastate.edu> chris@helser54.res.iastate.edu (Chris Wong) writes: >Make sure that you have the latest ROM on your DPT card. Some people have >problems using DPT with NS with the old ROM. After they upgraded the ROM, >everything is fine. DPT sent me the lastest revision only two weeks ago, thus I use 2D/5E. This is the most recent one, I suppose. Regards, Markus. -- Markus Wenzel System administration, Consulting, Networking mow@marsu.tynet.sub.org on... NeXTSTEP / Unix / Novell / Windows NT IRC: Marsu Expert in quantum bogodynamics
From: wolf@lanl.gov (David R Wolf) Newsgroups: comp.sys.next.bugs Subject: webster Date: 8 Apr 1994 16:24:45 GMT Organization: Los Alamos National Laboratory Distribution: world Message-ID: <2o40gd$j20@newshost.lanl.gov> here's a strange one: seems it's always possible to crash webster (NS 3.0) by searching for the word "a". -- ======================================================================= David R. Wolf wolf@lanl.gov LANL, MS P940, 87545 (505) 667-3813 =======================================================================
From: erik@rat.se (Erik Heimdahl) Newsgroups: comp.sys.next.bugs Subject: Problems with IB+NFS Date: 8 Apr 1994 15:33:24 -0400 Organization: The Ohio State University Sender: root@magnus.acs.ohio-state.edu Message-ID: <9404081648.AA04080@_rat.se> My IB keeps crashing (about once a day) when I try to save my .nib and it is very annoying. IB is running locally, but my home-directory is NFS mounted. The error I get is: NFS read error ESTALE to host iceman fh 600 0 80000 c87d3d08 94d10000 80000 116c2b26 408f0000 InterfaceBuilder[520]: NFS read error on pagein: stale file handle Anyone recognise this or know how to get the name of the file NFS is having problems with ?? I have black hardware and NS 3.1 /erik
From: samurai@wsc.com (Darcy Brockbank) Newsgroups: comp.sys.next.bugs,comp.sys.next.sysadmin Subject: Serious TCP bug?!?!? Date: 8 Apr 1994 17:47:16 GMT Organization: WSC Investment Services, Inc. Disclaimer: The views represented within this posting are not necessarily those held by WSC Investment Services, Inc. Message-ID: <2o45b4$52i@cerberus.wsc.com> I've documented the following two situations. They've occurred under similar circumstances, in two different installations: 1) #### A) NS-3.2 running on a 66MHz PC, with an Intel EtherExpress 16bit ISA card. This machine is running as a firewall, and runs INN as well as SOCKD (socks). B) NS-3.2 running on a 66MHz PC, on an internal net (can reach A via a router). B has an EtherExpress as well. Occasionaly when there are high rates of communication between A and B the following occurs: + Communications between A and B hang + A does not respond to rlogin/telnet from B + A does not return 'ping' requests from B + A allows login at console + B is to A as A is to B :-). + Ethernet card on B is not active. + After about 5 to 10 minutes, connections between A and B can be made again, without any interferance from a sysadmin. 2) #### A) NS-3.2 running on a 25MHz Motorola NeXT. This machine is running SLIP, and netinfo (network server). Nothing else. B) NS-3.2 runnong on a 66MHz PC, with an Intel EtherExpress. Occasionaly when there are high rates of communication between A and B the following occurs: + Communications between A and B hang + A does not respond to rlogin/telnet from B + A **DOES** return 'ping' requests from B + A allows login at console + B is to A as A is to B :-). + Ethernet card on B is not active. + THIS NEVER OCCURS UNLESS THE PC TRIES TO TALK TO THE NeXT. + After about 5 to 10 minutes, connections between A and B can be made again, without any interferance from a sysadmin. It would seem to me (and it's very possible that I'm wrong) that this is a TCP problem. The problem would seem to be introduced by the presence of the EtherExpress. However, it seems that the problem shows up in the host opposite to the EtherExpress since it happens with Motorola as well. This doesn't seem very scientific... it's as far as I can figure it out though. Questions: 1) Is anyone else seeing this? 2) Am I incorrect in thinking that this could be a TCP problem? Of course, I have to ask this on the net, since bug_next@next.com never elicits a response, and I don't have the money available to pay for an answer from NeXT... If this is a bug, then it's very serious... you can't have net connections hanging all the time. - darcy
From: eps@futon.SFSU.EDU (Eric P. Scott) Newsgroups: comp.sys.next.misc,comp.sys.next.bugs Subject: Re: NS/FIP 3.2 can't read Mac disk. Followup-To: comp.sys.next.misc Date: 9 Apr 1994 08:11:00 GMT Organization: San Francisco State University Message-ID: <2o5nuk$5ui@nic-nac.CSU.net> References: <2nr1bs$k5i@news.iastate.edu> <CnsJBp.Kxx@cunews.carleton.ca> In article <CnsJBp.Kxx@cunews.carleton.ca> csaldanh@mae.carleton.ca (Chris Saldanha) writes: >NeXTSTEP will only read High Density (HD) 1.4MB Mac floppies. 720K double >densities will not work. When you try to mount a 720K, NS will say its >not initialized, and ask you if you want to format it. 720K Mac diskettes *do* work. I just tried it. Well, actually, it didn't work the first time: the machine churned for a bit and spit the disk out. "That's not supposed to happen!" So I called up the Console window: Disk is Write Protected probing for CDROM probing for DOS /usr/filesystems/mac.fs/mac.name: Permission denied /usr/filesystems/mac.fs/mac.label: Permission denied probing for mac ... kern_loader: kern_loader: Link failed ... kern_loader: kern_loader: Link failed ... kern_loader: kern_loader: server macfs won't link ... kern_loader: kern_loader: server macfs won't link Not good. Here's what's happened: the machine I used (a NeXTstation) is a NetBoot client, which doesn't have root access to / (for damn good reason!). It was trying to scribble in /usr/filesystems/mac.fs--a BIG no-no. So I logged into the server as root and did # cd /usr/filesystems/mac.fs # ln -s /etc/mac.name . # ln -s /etc/mac.label . # ln -s /etc/macfs_loadable . (This way the different clients couldn't clobber each other's files.) THEN I could read 720K Mac floppies. [I'm crossposting this to comp.sys.next.bugs, since NeXT did screw up ever so slightly.] [Ob. FAQ reminder: you _did_ install the "720K Floppy Disk Formatter" extension on your Mac, right? If so, when you initialize a DD floppy you're offered a choice between "Macintosh 800K" and "Macintosh HFS Interchange Format." You need to select the latter. If you see buttons titled "One-Sided" and "Two-Sided," find a copy of System Update 2.0.1. If you don't have one, it's available for anonymous FTP from ftp.apple.com in dts/mac/sys.soft/7.system.updates/. You will need a blank HD floppy for this.] -=EPS=-
From: eps@futon.SFSU.EDU (Eric P. Scott) Newsgroups: comp.sys.next.bugs Subject: Why Johnny Can't RTFM (revised for NEXTSTEP 3.2) Date: 9 Apr 1994 09:03:03 GMT Organization: San Francisco State University Message-ID: <2o5r07$7gr@nic-nac.CSU.net> The following man pages do not nroff correctly because they do not end with a newline: file.1 genstrings.1 otool.1 libudp.3 lockf.3 monitor.3 mount.3r NSWSd.8c adduser.8 autonfsmount.8 bootparamd.8 newclient.8 rdate.8C rpcinfo.8 The following man pages source files that are only available if NEXTSTEP Developer is installed: getttyent.3 acct.5 ar.5 types.5 The following man pages need to be manually processed with tbl: mf.1 tbl.1 fd.4 tty.4 zs.4 The following man pages need to be manually processed with eqn: eqn.1 eqnchar.7 (note that spline.1g has already been eqn'ed and should not be run through eqn a second time.) *** /usr/bin/eqn is BROKEN in NEXTSTEP 3.2 for Motorola Processors--use the version distributed with NEXTSTEP 3.0 (bug introduced in NEXTSTEP 3.1) *** /usr/bin/eqn is BROKEN in NEXTSTEP 3.2 for Intel Processors-- no known workaround Reproduce: eqn /usr/man/man1/eqn.1 blows up with Segmentation fault -=EPS=-
Newsgroups: comp.sys.next.bugs From: js@euler.hnv.icem.de(Juergen Sell) Subject: Re: webster Message-ID: <Co0nEq.Aq@euler.hnv.icem.de> Sender: js@euler.hnv.icem.de (Juergen Sell) Organization: Ink Unknown References: <2o40gd$j20@newshost.lanl.gov> Date: Sat, 9 Apr 1994 23:47:11 GMT In article <2o40gd$j20@newshost.lanl.gov> wolf@lanl.gov (David R Wolf) writes: > here's a strange one: > > seems it's always possible to crash webster (NS 3.0) by searching for the > word "a". Works fine on my slab under NS3.0. Takes 1, 2 minutes and finally produces a long list of words starting with 'a'. No crash here. Juergen --- Juergen Sell E-Mail js@icem.de ( NeXTMail ok ) Deisterstr. 18 Fax ++49-511-440617 BRD 30449 Hannover Fon ++49-511-440688 == What time do we live in when all the word 'revolution' makes you think of == is a new generation of soap powder ?
From: sanguish@digifix.com Newsgroups: comp.sys.next.announce,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.bugs,comp.sys.next.programmer,comp.sys.next.marketplace Subject: Quick Guide to NEXTSTEP information on the Internet Date: 11 Apr 1994 02:24:34 -0400 Organization: Next Announcements Message-ID: <2oaqf2$f6e@digifix.digifix.com> This post is made weekly, to help 'point' users to more NEXTSTEP information Topics include: NEXTSTEP WWW Third Party Product Information Server NEXTSTEP Third Party Product Mail Server comp.sys.next newsgroups related newsgroups comp.sys.next newsgroups mailing list ftp sites NeXTanswers NEXTSTEP WWW Third Party Product Information Server --------------------------------------------------- The Server has information from many NEXTSTEP Third Party Developers. As well, there are - FTP links to the March 1994 NextAnswers MiniExamples - the ONLY only Hypertext version of the NEXTSTEP Expo Information - links to the big NEXTSTEP FTP sites - information on the NEXTSTEP Newsgroups Using a Mosaic compatible WWW reader, point to <http://digifix.digifix.com/index.html> [Under NEXTSTEP try OmniWeb.app (available from ftp.omnigroup.com:/pub/software)] NEXTSTEP Third Party Product Mail Server ---------------------------------------- Finally online, the Mail Server will allow you to retrieve information on NEXTSTEP products via NextMail (other formats coming soon). Accessing the Server The information is free, and the service is supported by sponsors from the NEXTSTEP Community. To find out what information is available and how to use the server send email to ns-products@digifix.com with index or help in the message. Index and Help queries are returned as plain ASCII, all others as NextMail. comp.sys.next.* newsgroups -------------------------- Comp.Sys.Next.Advocacy This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. Comp.Sys.Next.Announce Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) The NEXTSTEP FAQs are posted here monthly as well. This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Comp.Sys.Next.Bugs A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT- specific groups as well. Comp.Sys.Next.Hardware Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. Comp.Sys.Next.Marketplace NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. Comp.Sys.Next.Misc For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! Comp.Sys.Next.Programmer Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. The NEXTSTEP programmer FAQs are posted here. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. Comp.Sys.Next.Software This is a place to talk about [third party] software products that run on NEXTSTEP systems. Comp.Sys.Next.Sysadmin Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. related Newsgroups ------------------ Comp.Soft-Sys.Nextstep Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. Comp.Lang.Objective-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. Comp.Object Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original Comp.Sys.Next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News ------------------------------------------- Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep-d next-advocacy-d next-announce-d next-bugs-d next-hardware-d next-marketplace-d next-misc-d next-programmer-d next-software-d next-sysadmin-d (For a full description, send mail saying LISTS to <digestif@antigone.com>). The subscription syntax is essentially the same as LISTSERV's. To subscribe, send a message to <digestif@antigone.com> saying: SUB Listname YourName Example: SUB next-hardware-d John Doe The ftp sites ------------- cs.orst.edu: The main site for North American submissions nova.cc.purdue.edu: Lots of older stuff, but very short on disk space ftp.informatik.uni-muenchen.de: In Germany. terra.stack.urc.tue.nl (Dutch NEXTSTEP User Group) and cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp.next.com: See the below ftp.next.com and NextAnswers@next.com ------------------------------------- From the document 1000_Help from ftp.next.com Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, or you can transfer them by anonymous ftp. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to NeXTanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO <your-address> If you have any problem with the system or suggestions for improvement, please send mail to NeXTanswers-request@NeXT.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to NeXTanswers-request@NeXT.com. Written by: Eric P. Scott eps@toaster.SFSU.EDU and Scott Anguish sanguish@digifix.com Additions from: Greg Anderson (Greg_Anderson@afs.com) and Michael Pizolato (Michael_Pizolato@afs.com)
Newsgroups: comp.sys.next.bugs,comp.sys.next.programmer,comp.sys.next.software From: gery@ares.fdn.org (Gery Divry) Subject: Renderman Matrices product Message-ID: <1994Apr11.091815.371@ares.fdn.org> Keywords: Renderman Bug Sender: news@ares.fdn.org Organization: Ares - Lyon, France. Date: Mon, 11 Apr 1994 09:18:15 GMT After several tests it seems that the composition of two transformations (Transform) is bugged in renderman. I was thinking that this bug was fixed. ex : ##RenderMan RIB-Structure 1.0 #version 3.03 ##System NeXTSTEP Release 3 ##Creator ZZvolume on NeXTSTEP - (C) 1993 Gery.Divry Display "Image_000.tiff" "tiff" "rgb" Format 624 514 1.00 CropWindow 0.0000000 1.0000000 0.0000000 1.0000000 Option "shadow" "bias0" 0.005 Option "shadow" "bias1" 0.01 ShadingRate 4.000000 ShadingInterpolation "smooth" Projection "perspective" "fov" 13.661610 Orientation "lh" Identity Rotate 0.000000 0.0 0.0 1.0 Rotate -65.905157 1.0 0.0 0.0 Rotate 26.565051 0.0 1.0 0.0 Translate -0.200000 -1.160000 0.400000 Declare "coneangle" "float" Declare "conedeltaangle" "float" Declare "beamdistribution" "float" Declare "intensity" "float" Declare "lightcolor" "color" Declare "from" "point" Declare "to" "point" LightSource "ambientlight 0" "ambientlight" "intensity" [0.25] LightSource "distantlight Soleil" "distantlight" "from" [-0.539658 0.303675 -0.785207] "to" [0 0 0] "intensity" [0.85] MacroBegin "chaise" AttributeBegin Color 1.000 0.451 0.235 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ -0.01750 0.00000 -0.04637 -0.01750 0.04500 -0.04637 -0.01750 0.00000 -0.05037 -0.01750 0.04500 -0.05037 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.00000 -0.05037 -0.01750 0.04500 -0.05037 -0.02150 0.00000 -0.05037 -0.02150 0.04500 -0.05037 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.00000 -0.05037 -0.02150 0.04500 -0.05037 -0.02150 0.00000 -0.04637 -0.02150 0.04500 -0.04637 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.00000 -0.04637 -0.02150 0.04500 -0.04637 -0.01750 0.00000 -0.04637 -0.01750 0.04500 -0.04637 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.00000 -0.04637 -0.01750 0.00000 -0.05037 -0.02150 0.00000 -0.04637 -0.02150 0.00000 -0.05037 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.04500 -0.04637 -0.02150 0.04500 -0.05037 -0.01750 0.04500 -0.04637 -0.01750 0.04500 -0.05037 ] AttributeEnd AttributeEnd AttributeBegin Color 1.000 0.451 0.235 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ 0.02150 0.00000 -0.04637 0.02150 0.04500 -0.04637 0.02150 0.00000 -0.05037 0.02150 0.04500 -0.05037 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02150 0.00000 -0.05037 0.02150 0.04500 -0.05037 0.01750 0.00000 -0.05037 0.01750 0.04500 -0.05037 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.00000 -0.05037 0.01750 0.04500 -0.05037 0.01750 0.00000 -0.04637 0.01750 0.04500 -0.04637 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.00000 -0.04637 0.01750 0.04500 -0.04637 0.02150 0.00000 -0.04637 0.02150 0.04500 -0.04637 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02150 0.00000 -0.04637 0.02150 0.00000 -0.05037 0.01750 0.00000 -0.04637 0.01750 0.00000 -0.05037 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.04500 -0.04637 0.01750 0.04500 -0.05037 0.02150 0.04500 -0.04637 0.02150 0.04500 -0.05037 ] AttributeEnd AttributeEnd AttributeBegin Color 1.000 0.451 0.235 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ 0.02150 0.00000 -0.00737 0.02150 0.04500 -0.00737 0.02150 0.00000 -0.01137 0.02150 0.04500 -0.01137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02150 0.00000 -0.01137 0.02150 0.04500 -0.01137 0.01750 0.00000 -0.01137 0.01750 0.04500 -0.01137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.00000 -0.01137 0.01750 0.04500 -0.01137 0.01750 0.00000 -0.00737 0.01750 0.04500 -0.00737 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.00000 -0.00737 0.01750 0.04500 -0.00737 0.02150 0.00000 -0.00737 0.02150 0.04500 -0.00737 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02150 0.00000 -0.00737 0.02150 0.00000 -0.01137 0.01750 0.00000 -0.00737 0.01750 0.00000 -0.01137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.04500 -0.00737 0.01750 0.04500 -0.01137 0.02150 0.04500 -0.00737 0.02150 0.04500 -0.01137 ] AttributeEnd AttributeEnd AttributeBegin Color 1.000 0.451 0.235 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ -0.01750 0.00000 -0.00737 -0.01750 0.04500 -0.00737 -0.01750 0.00000 -0.01137 -0.01750 0.04500 -0.01137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.00000 -0.01137 -0.01750 0.04500 -0.01137 -0.02150 0.00000 -0.01137 -0.02150 0.04500 -0.01137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.00000 -0.01137 -0.02150 0.04500 -0.01137 -0.02150 0.00000 -0.00737 -0.02150 0.04500 -0.00737 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.00000 -0.00737 -0.02150 0.04500 -0.00737 -0.01750 0.00000 -0.00737 -0.01750 0.04500 -0.00737 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.00000 -0.00737 -0.01750 0.00000 -0.01137 -0.02150 0.00000 -0.00737 -0.02150 0.00000 -0.01137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.04500 -0.00737 -0.02150 0.04500 -0.01137 -0.01750 0.04500 -0.00737 -0.01750 0.04500 -0.01137 ] AttributeEnd AttributeEnd AttributeBegin Color 1.000 0.451 0.235 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ 0.02150 0.05000 -0.00738 0.02150 0.09941 0.00133 0.02150 0.05000 -0.01144 0.02150 0.10011 -0.00261 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02150 0.05000 -0.01144 0.02150 0.10011 -0.00261 0.01750 0.05000 -0.01144 0.01750 0.10011 -0.00261 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.05000 -0.01144 0.01750 0.10011 -0.00261 0.01750 0.05000 -0.00738 0.01750 0.09941 0.00133 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.05000 -0.00738 0.01750 0.09941 0.00133 0.02150 0.05000 -0.00738 0.02150 0.09941 0.00133 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02150 0.05000 -0.00738 0.02150 0.05000 -0.01144 0.01750 0.05000 -0.00738 0.01750 0.05000 -0.01144 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.09941 0.00133 0.01750 0.10011 -0.00261 0.02150 0.09941 0.00133 0.02150 0.10011 -0.00261 ] AttributeEnd AttributeEnd AttributeBegin Color 1.000 0.451 0.235 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ -0.01750 0.05000 -0.00738 -0.01750 0.09941 0.00133 -0.01750 0.05000 -0.01144 -0.01750 0.10011 -0.00261 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.05000 -0.01144 -0.01750 0.10011 -0.00261 -0.02150 0.05000 -0.01144 -0.02150 0.10011 -0.00261 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.05000 -0.01144 -0.02150 0.10011 -0.00261 -0.02150 0.05000 -0.00738 -0.02150 0.09941 0.00133 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.05000 -0.00738 -0.02150 0.09941 0.00133 -0.01750 0.05000 -0.00738 -0.01750 0.09941 0.00133 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.05000 -0.00738 -0.01750 0.05000 -0.01144 -0.02150 0.05000 -0.00738 -0.02150 0.05000 -0.01144 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.09941 0.00133 -0.02150 0.10011 -0.00261 -0.01750 0.09941 0.00133 -0.01750 0.10011 -0.00261 ] AttributeEnd AttributeEnd AttributeBegin Color 1.000 0.451 0.235 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ 0.02150 0.04500 -0.00738 0.02150 0.05000 -0.00738 0.02250 0.04500 -0.05137 0.02250 0.05000 -0.05137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02250 0.04500 -0.05137 0.02250 0.05000 -0.05137 -0.02250 0.04500 -0.05137 -0.02250 0.05000 -0.05137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02250 0.04500 -0.05137 -0.02250 0.05000 -0.05137 -0.02150 0.04500 -0.00738 -0.02150 0.05000 -0.00738 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.04500 -0.00738 -0.02150 0.05000 -0.00738 0.02150 0.04500 -0.00738 0.02150 0.05000 -0.00738 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02150 0.04500 -0.00738 0.02250 0.04500 -0.05137 -0.02150 0.04500 -0.00738 -0.02250 0.04500 -0.05137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.05000 -0.00738 -0.02250 0.05000 -0.05137 0.02150 0.05000 -0.00738 0.02250 0.05000 -0.05137 ] AttributeEnd AttributeEnd AttributeBegin Color 0.171 0.269 0.150 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ -0.01750 0.09762 0.00000 -0.01750 0.09779 -0.00098 0.01750 0.09762 0.00000 0.01750 0.09779 -0.00098 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.09762 0.00000 0.01750 0.09779 -0.00098 0.01750 0.08284 -0.00260 0.01750 0.08302 -0.00359 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.08284 -0.00260 0.01750 0.08302 -0.00359 -0.01750 0.08284 -0.00260 -0.01750 0.08302 -0.00359 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.08284 -0.00260 -0.01750 0.08302 -0.00359 -0.01750 0.09762 0.00000 -0.01750 0.09779 -0.00098 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.09762 0.00000 0.01750 0.09762 0.00000 -0.01750 0.08284 -0.00260 0.01750 0.08284 -0.00260 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.08302 -0.00359 0.01750 0.08302 -0.00359 -0.01750 0.09779 -0.00098 0.01750 0.09779 -0.00098 ] AttributeEnd AttributeEnd MacroEnd WorldBegin Surface "plastic" AttributeBegin Color 0.000 0.000 0.000 Opacity 0.000 0.000 0.000 Attribute "identifier" "name" ["chaise2_1048581"] TransformBegin Transform [ 0.000000 0.000000 1.000000 0.000000 0.000000 1.000000 0.000000 0.000000 -1.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 1.000000 ] AttributeBegin Color 0.000 0.000 0.000 Opacity 0.000 0.000 0.000 Attribute "identifier" "name" ["chaise_0"] TransformBegin Transform [ 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 ] MacroInstance "chaise" TransformEnd AttributeEnd TransformEnd AttributeEnd AttributeBegin Color 0.000 0.000 0.000 Opacity 0.000 0.000 0.000 Attribute "identifier" "name" ["chaise2_1048578"] TransformBegin Transform [ 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 ] AttributeBegin Color 0.000 0.000 0.000 Opacity 0.000 0.000 0.000 Attribute "identifier" "name" ["chaise_0"] TransformBegin Transform [ 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 ] MacroInstance "chaise" TransformEnd AttributeEnd TransformEnd AttributeEnd WorldEnd Only one chair appear.... i should have a second one with a 90deg angle The composition of two Identity Matrix is Ok ( fortunatly) but an Identity with an other type of matrix dont seem to work. Gery DIVRY ( ZZVolume Daddy ) ARES Publisher 8, rue Victor Lagrange Phone: (+33) 72 80 16 30 69007 LYON Fax: (+33) 72 80 16 32 France Email: gery@ares.fdn.org Earth, Solar System, Galaxy MW1 NeXT Mail accepted
Newsgroups: comp.sys.next.bugs From: gery@ares.fdn.org (Gery Divry) Subject: Renderman Matrices product Message-ID: <1994Apr11.091841.435@ares.fdn.org> Keywords: Renderman Bug Sender: news@ares.fdn.org Organization: Ares - Lyon, France. Date: Mon, 11 Apr 1994 09:18:41 GMT After several tests it seems that the composition of two transformations (Transform) is bugged in renderman. I was thinking that this bug was fixed. ex : ##RenderMan RIB-Structure 1.0 #version 3.03 ##System NeXTSTEP Release 3 ##Creator ZZvolume on NeXTSTEP - (C) 1993 Gery.Divry Display "Image_000.tiff" "tiff" "rgb" Format 624 514 1.00 CropWindow 0.0000000 1.0000000 0.0000000 1.0000000 Option "shadow" "bias0" 0.005 Option "shadow" "bias1" 0.01 ShadingRate 4.000000 ShadingInterpolation "smooth" Projection "perspective" "fov" 13.661610 Orientation "lh" Identity Rotate 0.000000 0.0 0.0 1.0 Rotate -65.905157 1.0 0.0 0.0 Rotate 26.565051 0.0 1.0 0.0 Translate -0.200000 -1.160000 0.400000 Declare "coneangle" "float" Declare "conedeltaangle" "float" Declare "beamdistribution" "float" Declare "intensity" "float" Declare "lightcolor" "color" Declare "from" "point" Declare "to" "point" LightSource "ambientlight 0" "ambientlight" "intensity" [0.25] LightSource "distantlight Soleil" "distantlight" "from" [-0.539658 0.303675 -0.785207] "to" [0 0 0] "intensity" [0.85] MacroBegin "chaise" AttributeBegin Color 1.000 0.451 0.235 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ -0.01750 0.00000 -0.04637 -0.01750 0.04500 -0.04637 -0.01750 0.00000 -0.05037 -0.01750 0.04500 -0.05037 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.00000 -0.05037 -0.01750 0.04500 -0.05037 -0.02150 0.00000 -0.05037 -0.02150 0.04500 -0.05037 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.00000 -0.05037 -0.02150 0.04500 -0.05037 -0.02150 0.00000 -0.04637 -0.02150 0.04500 -0.04637 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.00000 -0.04637 -0.02150 0.04500 -0.04637 -0.01750 0.00000 -0.04637 -0.01750 0.04500 -0.04637 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.00000 -0.04637 -0.01750 0.00000 -0.05037 -0.02150 0.00000 -0.04637 -0.02150 0.00000 -0.05037 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.04500 -0.04637 -0.02150 0.04500 -0.05037 -0.01750 0.04500 -0.04637 -0.01750 0.04500 -0.05037 ] AttributeEnd AttributeEnd AttributeBegin Color 1.000 0.451 0.235 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ 0.02150 0.00000 -0.04637 0.02150 0.04500 -0.04637 0.02150 0.00000 -0.05037 0.02150 0.04500 -0.05037 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02150 0.00000 -0.05037 0.02150 0.04500 -0.05037 0.01750 0.00000 -0.05037 0.01750 0.04500 -0.05037 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.00000 -0.05037 0.01750 0.04500 -0.05037 0.01750 0.00000 -0.04637 0.01750 0.04500 -0.04637 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.00000 -0.04637 0.01750 0.04500 -0.04637 0.02150 0.00000 -0.04637 0.02150 0.04500 -0.04637 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02150 0.00000 -0.04637 0.02150 0.00000 -0.05037 0.01750 0.00000 -0.04637 0.01750 0.00000 -0.05037 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.04500 -0.04637 0.01750 0.04500 -0.05037 0.02150 0.04500 -0.04637 0.02150 0.04500 -0.05037 ] AttributeEnd AttributeEnd AttributeBegin Color 1.000 0.451 0.235 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ 0.02150 0.00000 -0.00737 0.02150 0.04500 -0.00737 0.02150 0.00000 -0.01137 0.02150 0.04500 -0.01137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02150 0.00000 -0.01137 0.02150 0.04500 -0.01137 0.01750 0.00000 -0.01137 0.01750 0.04500 -0.01137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.00000 -0.01137 0.01750 0.04500 -0.01137 0.01750 0.00000 -0.00737 0.01750 0.04500 -0.00737 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.00000 -0.00737 0.01750 0.04500 -0.00737 0.02150 0.00000 -0.00737 0.02150 0.04500 -0.00737 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02150 0.00000 -0.00737 0.02150 0.00000 -0.01137 0.01750 0.00000 -0.00737 0.01750 0.00000 -0.01137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.04500 -0.00737 0.01750 0.04500 -0.01137 0.02150 0.04500 -0.00737 0.02150 0.04500 -0.01137 ] AttributeEnd AttributeEnd AttributeBegin Color 1.000 0.451 0.235 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ -0.01750 0.00000 -0.00737 -0.01750 0.04500 -0.00737 -0.01750 0.00000 -0.01137 -0.01750 0.04500 -0.01137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.00000 -0.01137 -0.01750 0.04500 -0.01137 -0.02150 0.00000 -0.01137 -0.02150 0.04500 -0.01137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.00000 -0.01137 -0.02150 0.04500 -0.01137 -0.02150 0.00000 -0.00737 -0.02150 0.04500 -0.00737 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.00000 -0.00737 -0.02150 0.04500 -0.00737 -0.01750 0.00000 -0.00737 -0.01750 0.04500 -0.00737 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.00000 -0.00737 -0.01750 0.00000 -0.01137 -0.02150 0.00000 -0.00737 -0.02150 0.00000 -0.01137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.04500 -0.00737 -0.02150 0.04500 -0.01137 -0.01750 0.04500 -0.00737 -0.01750 0.04500 -0.01137 ] AttributeEnd AttributeEnd AttributeBegin Color 1.000 0.451 0.235 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ 0.02150 0.05000 -0.00738 0.02150 0.09941 0.00133 0.02150 0.05000 -0.01144 0.02150 0.10011 -0.00261 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02150 0.05000 -0.01144 0.02150 0.10011 -0.00261 0.01750 0.05000 -0.01144 0.01750 0.10011 -0.00261 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.05000 -0.01144 0.01750 0.10011 -0.00261 0.01750 0.05000 -0.00738 0.01750 0.09941 0.00133 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.05000 -0.00738 0.01750 0.09941 0.00133 0.02150 0.05000 -0.00738 0.02150 0.09941 0.00133 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02150 0.05000 -0.00738 0.02150 0.05000 -0.01144 0.01750 0.05000 -0.00738 0.01750 0.05000 -0.01144 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.09941 0.00133 0.01750 0.10011 -0.00261 0.02150 0.09941 0.00133 0.02150 0.10011 -0.00261 ] AttributeEnd AttributeEnd AttributeBegin Color 1.000 0.451 0.235 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ -0.01750 0.05000 -0.00738 -0.01750 0.09941 0.00133 -0.01750 0.05000 -0.01144 -0.01750 0.10011 -0.00261 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.05000 -0.01144 -0.01750 0.10011 -0.00261 -0.02150 0.05000 -0.01144 -0.02150 0.10011 -0.00261 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.05000 -0.01144 -0.02150 0.10011 -0.00261 -0.02150 0.05000 -0.00738 -0.02150 0.09941 0.00133 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.05000 -0.00738 -0.02150 0.09941 0.00133 -0.01750 0.05000 -0.00738 -0.01750 0.09941 0.00133 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.05000 -0.00738 -0.01750 0.05000 -0.01144 -0.02150 0.05000 -0.00738 -0.02150 0.05000 -0.01144 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.09941 0.00133 -0.02150 0.10011 -0.00261 -0.01750 0.09941 0.00133 -0.01750 0.10011 -0.00261 ] AttributeEnd AttributeEnd AttributeBegin Color 1.000 0.451 0.235 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ 0.02150 0.04500 -0.00738 0.02150 0.05000 -0.00738 0.02250 0.04500 -0.05137 0.02250 0.05000 -0.05137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02250 0.04500 -0.05137 0.02250 0.05000 -0.05137 -0.02250 0.04500 -0.05137 -0.02250 0.05000 -0.05137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02250 0.04500 -0.05137 -0.02250 0.05000 -0.05137 -0.02150 0.04500 -0.00738 -0.02150 0.05000 -0.00738 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.04500 -0.00738 -0.02150 0.05000 -0.00738 0.02150 0.04500 -0.00738 0.02150 0.05000 -0.00738 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02150 0.04500 -0.00738 0.02250 0.04500 -0.05137 -0.02150 0.04500 -0.00738 -0.02250 0.04500 -0.05137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.05000 -0.00738 -0.02250 0.05000 -0.05137 0.02150 0.05000 -0.00738 0.02250 0.05000 -0.05137 ] AttributeEnd AttributeEnd AttributeBegin Color 0.171 0.269 0.150 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ -0.01750 0.09762 0.00000 -0.01750 0.09779 -0.00098 0.01750 0.09762 0.00000 0.01750 0.09779 -0.00098 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.09762 0.00000 0.01750 0.09779 -0.00098 0.01750 0.08284 -0.00260 0.01750 0.08302 -0.00359 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.08284 -0.00260 0.01750 0.08302 -0.00359 -0.01750 0.08284 -0.00260 -0.01750 0.08302 -0.00359 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.08284 -0.00260 -0.01750 0.08302 -0.00359 -0.01750 0.09762 0.00000 -0.01750 0.09779 -0.00098 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.09762 0.00000 0.01750 0.09762 0.00000 -0.01750 0.08284 -0.00260 0.01750 0.08284 -0.00260 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.08302 -0.00359 0.01750 0.08302 -0.00359 -0.01750 0.09779 -0.00098 0.01750 0.09779 -0.00098 ] AttributeEnd AttributeEnd MacroEnd WorldBegin Surface "plastic" AttributeBegin Color 0.000 0.000 0.000 Opacity 0.000 0.000 0.000 Attribute "identifier" "name" ["chaise2_1048581"] TransformBegin Transform [ 0.000000 0.000000 1.000000 0.000000 0.000000 1.000000 0.000000 0.000000 -1.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 1.000000 ] AttributeBegin Color 0.000 0.000 0.000 Opacity 0.000 0.000 0.000 Attribute "identifier" "name" ["chaise_0"] TransformBegin Transform [ 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 ] MacroInstance "chaise" TransformEnd AttributeEnd TransformEnd AttributeEnd AttributeBegin Color 0.000 0.000 0.000 Opacity 0.000 0.000 0.000 Attribute "identifier" "name" ["chaise2_1048578"] TransformBegin Transform [ 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 ] AttributeBegin Color 0.000 0.000 0.000 Opacity 0.000 0.000 0.000 Attribute "identifier" "name" ["chaise_0"] TransformBegin Transform [ 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 ] MacroInstance "chaise" TransformEnd AttributeEnd TransformEnd AttributeEnd WorldEnd Only one chair appear.... i should have a second one with a 90deg angle The composition of two Identity Matrix is Ok ( fortunatly) but an Identity with an other type of matrix dont seem to work. Gery DIVRY ( ZZVolume Daddy ) ARES Publisher 8, rue Victor Lagrange Phone: (+33) 72 80 16 30 69007 LYON Fax: (+33) 72 80 16 32 France Email: gery@ares.fdn.org Earth, Solar System, Galaxy MW1 NeXT Mail accepted
Newsgroups: comp.sys.next.bugs From: gery@ares.fdn.org (Gery Divry) Subject: Renderman Matrices product Message-ID: <1994Apr11.091935.559@ares.fdn.org> Keywords: Renderman Bug Sender: news@ares.fdn.org Organization: Ares - Lyon, France. Date: Mon, 11 Apr 1994 09:19:35 GMT After several tests it seems that the composition of two transformations (Transform) is bugged in renderman. I was thinking that this bug was fixed. ex : ##RenderMan RIB-Structure 1.0 #version 3.03 ##System NeXTSTEP Release 3 ##Creator ZZvolume on NeXTSTEP - (C) 1993 Gery.Divry Display "Image_000.tiff" "tiff" "rgb" Format 624 514 1.00 CropWindow 0.0000000 1.0000000 0.0000000 1.0000000 Option "shadow" "bias0" 0.005 Option "shadow" "bias1" 0.01 ShadingRate 4.000000 ShadingInterpolation "smooth" Projection "perspective" "fov" 13.661610 Orientation "lh" Identity Rotate 0.000000 0.0 0.0 1.0 Rotate -65.905157 1.0 0.0 0.0 Rotate 26.565051 0.0 1.0 0.0 Translate -0.200000 -1.160000 0.400000 Declare "coneangle" "float" Declare "conedeltaangle" "float" Declare "beamdistribution" "float" Declare "intensity" "float" Declare "lightcolor" "color" Declare "from" "point" Declare "to" "point" LightSource "ambientlight 0" "ambientlight" "intensity" [0.25] LightSource "distantlight Soleil" "distantlight" "from" [-0.539658 0.303675 -0.785207] "to" [0 0 0] "intensity" [0.85] MacroBegin "chaise" AttributeBegin Color 1.000 0.451 0.235 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ -0.01750 0.00000 -0.04637 -0.01750 0.04500 -0.04637 -0.01750 0.00000 -0.05037 -0.01750 0.04500 -0.05037 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.00000 -0.05037 -0.01750 0.04500 -0.05037 -0.02150 0.00000 -0.05037 -0.02150 0.04500 -0.05037 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.00000 -0.05037 -0.02150 0.04500 -0.05037 -0.02150 0.00000 -0.04637 -0.02150 0.04500 -0.04637 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.00000 -0.04637 -0.02150 0.04500 -0.04637 -0.01750 0.00000 -0.04637 -0.01750 0.04500 -0.04637 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.00000 -0.04637 -0.01750 0.00000 -0.05037 -0.02150 0.00000 -0.04637 -0.02150 0.00000 -0.05037 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.04500 -0.04637 -0.02150 0.04500 -0.05037 -0.01750 0.04500 -0.04637 -0.01750 0.04500 -0.05037 ] AttributeEnd AttributeEnd AttributeBegin Color 1.000 0.451 0.235 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ 0.02150 0.00000 -0.04637 0.02150 0.04500 -0.04637 0.02150 0.00000 -0.05037 0.02150 0.04500 -0.05037 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02150 0.00000 -0.05037 0.02150 0.04500 -0.05037 0.01750 0.00000 -0.05037 0.01750 0.04500 -0.05037 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.00000 -0.05037 0.01750 0.04500 -0.05037 0.01750 0.00000 -0.04637 0.01750 0.04500 -0.04637 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.00000 -0.04637 0.01750 0.04500 -0.04637 0.02150 0.00000 -0.04637 0.02150 0.04500 -0.04637 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02150 0.00000 -0.04637 0.02150 0.00000 -0.05037 0.01750 0.00000 -0.04637 0.01750 0.00000 -0.05037 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.04500 -0.04637 0.01750 0.04500 -0.05037 0.02150 0.04500 -0.04637 0.02150 0.04500 -0.05037 ] AttributeEnd AttributeEnd AttributeBegin Color 1.000 0.451 0.235 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ 0.02150 0.00000 -0.00737 0.02150 0.04500 -0.00737 0.02150 0.00000 -0.01137 0.02150 0.04500 -0.01137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02150 0.00000 -0.01137 0.02150 0.04500 -0.01137 0.01750 0.00000 -0.01137 0.01750 0.04500 -0.01137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.00000 -0.01137 0.01750 0.04500 -0.01137 0.01750 0.00000 -0.00737 0.01750 0.04500 -0.00737 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.00000 -0.00737 0.01750 0.04500 -0.00737 0.02150 0.00000 -0.00737 0.02150 0.04500 -0.00737 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02150 0.00000 -0.00737 0.02150 0.00000 -0.01137 0.01750 0.00000 -0.00737 0.01750 0.00000 -0.01137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.04500 -0.00737 0.01750 0.04500 -0.01137 0.02150 0.04500 -0.00737 0.02150 0.04500 -0.01137 ] AttributeEnd AttributeEnd AttributeBegin Color 1.000 0.451 0.235 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ -0.01750 0.00000 -0.00737 -0.01750 0.04500 -0.00737 -0.01750 0.00000 -0.01137 -0.01750 0.04500 -0.01137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.00000 -0.01137 -0.01750 0.04500 -0.01137 -0.02150 0.00000 -0.01137 -0.02150 0.04500 -0.01137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.00000 -0.01137 -0.02150 0.04500 -0.01137 -0.02150 0.00000 -0.00737 -0.02150 0.04500 -0.00737 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.00000 -0.00737 -0.02150 0.04500 -0.00737 -0.01750 0.00000 -0.00737 -0.01750 0.04500 -0.00737 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.00000 -0.00737 -0.01750 0.00000 -0.01137 -0.02150 0.00000 -0.00737 -0.02150 0.00000 -0.01137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.04500 -0.00737 -0.02150 0.04500 -0.01137 -0.01750 0.04500 -0.00737 -0.01750 0.04500 -0.01137 ] AttributeEnd AttributeEnd AttributeBegin Color 1.000 0.451 0.235 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ 0.02150 0.05000 -0.00738 0.02150 0.09941 0.00133 0.02150 0.05000 -0.01144 0.02150 0.10011 -0.00261 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02150 0.05000 -0.01144 0.02150 0.10011 -0.00261 0.01750 0.05000 -0.01144 0.01750 0.10011 -0.00261 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.05000 -0.01144 0.01750 0.10011 -0.00261 0.01750 0.05000 -0.00738 0.01750 0.09941 0.00133 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.05000 -0.00738 0.01750 0.09941 0.00133 0.02150 0.05000 -0.00738 0.02150 0.09941 0.00133 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02150 0.05000 -0.00738 0.02150 0.05000 -0.01144 0.01750 0.05000 -0.00738 0.01750 0.05000 -0.01144 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.09941 0.00133 0.01750 0.10011 -0.00261 0.02150 0.09941 0.00133 0.02150 0.10011 -0.00261 ] AttributeEnd AttributeEnd AttributeBegin Color 1.000 0.451 0.235 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ -0.01750 0.05000 -0.00738 -0.01750 0.09941 0.00133 -0.01750 0.05000 -0.01144 -0.01750 0.10011 -0.00261 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.05000 -0.01144 -0.01750 0.10011 -0.00261 -0.02150 0.05000 -0.01144 -0.02150 0.10011 -0.00261 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.05000 -0.01144 -0.02150 0.10011 -0.00261 -0.02150 0.05000 -0.00738 -0.02150 0.09941 0.00133 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.05000 -0.00738 -0.02150 0.09941 0.00133 -0.01750 0.05000 -0.00738 -0.01750 0.09941 0.00133 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.05000 -0.00738 -0.01750 0.05000 -0.01144 -0.02150 0.05000 -0.00738 -0.02150 0.05000 -0.01144 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.09941 0.00133 -0.02150 0.10011 -0.00261 -0.01750 0.09941 0.00133 -0.01750 0.10011 -0.00261 ] AttributeEnd AttributeEnd AttributeBegin Color 1.000 0.451 0.235 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ 0.02150 0.04500 -0.00738 0.02150 0.05000 -0.00738 0.02250 0.04500 -0.05137 0.02250 0.05000 -0.05137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02250 0.04500 -0.05137 0.02250 0.05000 -0.05137 -0.02250 0.04500 -0.05137 -0.02250 0.05000 -0.05137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02250 0.04500 -0.05137 -0.02250 0.05000 -0.05137 -0.02150 0.04500 -0.00738 -0.02150 0.05000 -0.00738 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.04500 -0.00738 -0.02150 0.05000 -0.00738 0.02150 0.04500 -0.00738 0.02150 0.05000 -0.00738 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02150 0.04500 -0.00738 0.02250 0.04500 -0.05137 -0.02150 0.04500 -0.00738 -0.02250 0.04500 -0.05137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.05000 -0.00738 -0.02250 0.05000 -0.05137 0.02150 0.05000 -0.00738 0.02250 0.05000 -0.05137 ] AttributeEnd AttributeEnd AttributeBegin Color 0.171 0.269 0.150 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ -0.01750 0.09762 0.00000 -0.01750 0.09779 -0.00098 0.01750 0.09762 0.00000 0.01750 0.09779 -0.00098 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.09762 0.00000 0.01750 0.09779 -0.00098 0.01750 0.08284 -0.00260 0.01750 0.08302 -0.00359 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.08284 -0.00260 0.01750 0.08302 -0.00359 -0.01750 0.08284 -0.00260 -0.01750 0.08302 -0.00359 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.08284 -0.00260 -0.01750 0.08302 -0.00359 -0.01750 0.09762 0.00000 -0.01750 0.09779 -0.00098 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.09762 0.00000 0.01750 0.09762 0.00000 -0.01750 0.08284 -0.00260 0.01750 0.08284 -0.00260 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.08302 -0.00359 0.01750 0.08302 -0.00359 -0.01750 0.09779 -0.00098 0.01750 0.09779 -0.00098 ] AttributeEnd AttributeEnd MacroEnd WorldBegin Surface "plastic" AttributeBegin Color 0.000 0.000 0.000 Opacity 0.000 0.000 0.000 Attribute "identifier" "name" ["chaise2_1048581"] TransformBegin Transform [ 0.000000 0.000000 1.000000 0.000000 0.000000 1.000000 0.000000 0.000000 -1.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 1.000000 ] AttributeBegin Color 0.000 0.000 0.000 Opacity 0.000 0.000 0.000 Attribute "identifier" "name" ["chaise_0"] TransformBegin Transform [ 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 ] MacroInstance "chaise" TransformEnd AttributeEnd TransformEnd AttributeEnd AttributeBegin Color 0.000 0.000 0.000 Opacity 0.000 0.000 0.000 Attribute "identifier" "name" ["chaise2_1048578"] TransformBegin Transform [ 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 ] AttributeBegin Color 0.000 0.000 0.000 Opacity 0.000 0.000 0.000 Attribute "identifier" "name" ["chaise_0"] TransformBegin Transform [ 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 ] MacroInstance "chaise" TransformEnd AttributeEnd TransformEnd AttributeEnd WorldEnd Only one chair appear.... i should have a second one with a 90deg angle The composition of two Identity Matrix is Ok ( fortunatly) but an Identity with an other type of matrix dont seem to work. Gery DIVRY ( ZZVolume Daddy ) ARES Publisher 8, rue Victor Lagrange Phone: (+33) 72 80 16 30 69007 LYON Fax: (+33) 72 80 16 32 France Email: gery@ares.fdn.org Earth, Solar System, Galaxy MW1 NeXT Mail accepted
Newsgroups: comp.sys.next.bugs From: gery@ares.fdn.org (Gery Divry) Subject: Renderman bug Message-ID: <1994Apr11.092007.680@ares.fdn.org> Sender: news@ares.fdn.org Organization: Ares - Lyon, France. Date: Mon, 11 Apr 1994 09:20:07 GMT After several tests it seems that the composition of two transformations (Transform) is bugged in renderman. I was thinking that this bug was fixed. ex : ##RenderMan RIB-Structure 1.0 #version 3.03 ##System NeXTSTEP Release 3 ##Creator ZZvolume on NeXTSTEP - (C) 1993 Gery.Divry Display "Image_000.tiff" "tiff" "rgb" Format 624 514 1.00 CropWindow 0.0000000 1.0000000 0.0000000 1.0000000 Option "shadow" "bias0" 0.005 Option "shadow" "bias1" 0.01 ShadingRate 4.000000 ShadingInterpolation "smooth" Projection "perspective" "fov" 13.661610 Orientation "lh" Identity Rotate 0.000000 0.0 0.0 1.0 Rotate -65.905157 1.0 0.0 0.0 Rotate 26.565051 0.0 1.0 0.0 Translate -0.200000 -1.160000 0.400000 Declare "coneangle" "float" Declare "conedeltaangle" "float" Declare "beamdistribution" "float" Declare "intensity" "float" Declare "lightcolor" "color" Declare "from" "point" Declare "to" "point" LightSource "ambientlight 0" "ambientlight" "intensity" [0.25] LightSource "distantlight Soleil" "distantlight" "from" [-0.539658 0.303675 -0.785207] "to" [0 0 0] "intensity" [0.85] MacroBegin "chaise" AttributeBegin Color 1.000 0.451 0.235 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ -0.01750 0.00000 -0.04637 -0.01750 0.04500 -0.04637 -0.01750 0.00000 -0.05037 -0.01750 0.04500 -0.05037 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.00000 -0.05037 -0.01750 0.04500 -0.05037 -0.02150 0.00000 -0.05037 -0.02150 0.04500 -0.05037 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.00000 -0.05037 -0.02150 0.04500 -0.05037 -0.02150 0.00000 -0.04637 -0.02150 0.04500 -0.04637 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.00000 -0.04637 -0.02150 0.04500 -0.04637 -0.01750 0.00000 -0.04637 -0.01750 0.04500 -0.04637 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.00000 -0.04637 -0.01750 0.00000 -0.05037 -0.02150 0.00000 -0.04637 -0.02150 0.00000 -0.05037 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.04500 -0.04637 -0.02150 0.04500 -0.05037 -0.01750 0.04500 -0.04637 -0.01750 0.04500 -0.05037 ] AttributeEnd AttributeEnd AttributeBegin Color 1.000 0.451 0.235 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ 0.02150 0.00000 -0.04637 0.02150 0.04500 -0.04637 0.02150 0.00000 -0.05037 0.02150 0.04500 -0.05037 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02150 0.00000 -0.05037 0.02150 0.04500 -0.05037 0.01750 0.00000 -0.05037 0.01750 0.04500 -0.05037 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.00000 -0.05037 0.01750 0.04500 -0.05037 0.01750 0.00000 -0.04637 0.01750 0.04500 -0.04637 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.00000 -0.04637 0.01750 0.04500 -0.04637 0.02150 0.00000 -0.04637 0.02150 0.04500 -0.04637 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02150 0.00000 -0.04637 0.02150 0.00000 -0.05037 0.01750 0.00000 -0.04637 0.01750 0.00000 -0.05037 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.04500 -0.04637 0.01750 0.04500 -0.05037 0.02150 0.04500 -0.04637 0.02150 0.04500 -0.05037 ] AttributeEnd AttributeEnd AttributeBegin Color 1.000 0.451 0.235 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ 0.02150 0.00000 -0.00737 0.02150 0.04500 -0.00737 0.02150 0.00000 -0.01137 0.02150 0.04500 -0.01137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02150 0.00000 -0.01137 0.02150 0.04500 -0.01137 0.01750 0.00000 -0.01137 0.01750 0.04500 -0.01137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.00000 -0.01137 0.01750 0.04500 -0.01137 0.01750 0.00000 -0.00737 0.01750 0.04500 -0.00737 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.00000 -0.00737 0.01750 0.04500 -0.00737 0.02150 0.00000 -0.00737 0.02150 0.04500 -0.00737 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02150 0.00000 -0.00737 0.02150 0.00000 -0.01137 0.01750 0.00000 -0.00737 0.01750 0.00000 -0.01137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.04500 -0.00737 0.01750 0.04500 -0.01137 0.02150 0.04500 -0.00737 0.02150 0.04500 -0.01137 ] AttributeEnd AttributeEnd AttributeBegin Color 1.000 0.451 0.235 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ -0.01750 0.00000 -0.00737 -0.01750 0.04500 -0.00737 -0.01750 0.00000 -0.01137 -0.01750 0.04500 -0.01137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.00000 -0.01137 -0.01750 0.04500 -0.01137 -0.02150 0.00000 -0.01137 -0.02150 0.04500 -0.01137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.00000 -0.01137 -0.02150 0.04500 -0.01137 -0.02150 0.00000 -0.00737 -0.02150 0.04500 -0.00737 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.00000 -0.00737 -0.02150 0.04500 -0.00737 -0.01750 0.00000 -0.00737 -0.01750 0.04500 -0.00737 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.00000 -0.00737 -0.01750 0.00000 -0.01137 -0.02150 0.00000 -0.00737 -0.02150 0.00000 -0.01137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.04500 -0.00737 -0.02150 0.04500 -0.01137 -0.01750 0.04500 -0.00737 -0.01750 0.04500 -0.01137 ] AttributeEnd AttributeEnd AttributeBegin Color 1.000 0.451 0.235 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ 0.02150 0.05000 -0.00738 0.02150 0.09941 0.00133 0.02150 0.05000 -0.01144 0.02150 0.10011 -0.00261 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02150 0.05000 -0.01144 0.02150 0.10011 -0.00261 0.01750 0.05000 -0.01144 0.01750 0.10011 -0.00261 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.05000 -0.01144 0.01750 0.10011 -0.00261 0.01750 0.05000 -0.00738 0.01750 0.09941 0.00133 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.05000 -0.00738 0.01750 0.09941 0.00133 0.02150 0.05000 -0.00738 0.02150 0.09941 0.00133 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02150 0.05000 -0.00738 0.02150 0.05000 -0.01144 0.01750 0.05000 -0.00738 0.01750 0.05000 -0.01144 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.09941 0.00133 0.01750 0.10011 -0.00261 0.02150 0.09941 0.00133 0.02150 0.10011 -0.00261 ] AttributeEnd AttributeEnd AttributeBegin Color 1.000 0.451 0.235 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ -0.01750 0.05000 -0.00738 -0.01750 0.09941 0.00133 -0.01750 0.05000 -0.01144 -0.01750 0.10011 -0.00261 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.05000 -0.01144 -0.01750 0.10011 -0.00261 -0.02150 0.05000 -0.01144 -0.02150 0.10011 -0.00261 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.05000 -0.01144 -0.02150 0.10011 -0.00261 -0.02150 0.05000 -0.00738 -0.02150 0.09941 0.00133 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.05000 -0.00738 -0.02150 0.09941 0.00133 -0.01750 0.05000 -0.00738 -0.01750 0.09941 0.00133 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.05000 -0.00738 -0.01750 0.05000 -0.01144 -0.02150 0.05000 -0.00738 -0.02150 0.05000 -0.01144 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.09941 0.00133 -0.02150 0.10011 -0.00261 -0.01750 0.09941 0.00133 -0.01750 0.10011 -0.00261 ] AttributeEnd AttributeEnd AttributeBegin Color 1.000 0.451 0.235 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ 0.02150 0.04500 -0.00738 0.02150 0.05000 -0.00738 0.02250 0.04500 -0.05137 0.02250 0.05000 -0.05137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02250 0.04500 -0.05137 0.02250 0.05000 -0.05137 -0.02250 0.04500 -0.05137 -0.02250 0.05000 -0.05137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02250 0.04500 -0.05137 -0.02250 0.05000 -0.05137 -0.02150 0.04500 -0.00738 -0.02150 0.05000 -0.00738 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.04500 -0.00738 -0.02150 0.05000 -0.00738 0.02150 0.04500 -0.00738 0.02150 0.05000 -0.00738 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.02150 0.04500 -0.00738 0.02250 0.04500 -0.05137 -0.02150 0.04500 -0.00738 -0.02250 0.04500 -0.05137 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.02150 0.05000 -0.00738 -0.02250 0.05000 -0.05137 0.02150 0.05000 -0.00738 0.02250 0.05000 -0.05137 ] AttributeEnd AttributeEnd AttributeBegin Color 0.171 0.269 0.150 Opacity 1.000 1.000 1.000 AttributeBegin Patch "bilinear" "P" [ -0.01750 0.09762 0.00000 -0.01750 0.09779 -0.00098 0.01750 0.09762 0.00000 0.01750 0.09779 -0.00098 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.09762 0.00000 0.01750 0.09779 -0.00098 0.01750 0.08284 -0.00260 0.01750 0.08302 -0.00359 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ 0.01750 0.08284 -0.00260 0.01750 0.08302 -0.00359 -0.01750 0.08284 -0.00260 -0.01750 0.08302 -0.00359 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.08284 -0.00260 -0.01750 0.08302 -0.00359 -0.01750 0.09762 0.00000 -0.01750 0.09779 -0.00098 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.09762 0.00000 0.01750 0.09762 0.00000 -0.01750 0.08284 -0.00260 0.01750 0.08284 -0.00260 ] AttributeEnd AttributeBegin Patch "bilinear" "P" [ -0.01750 0.08302 -0.00359 0.01750 0.08302 -0.00359 -0.01750 0.09779 -0.00098 0.01750 0.09779 -0.00098 ] AttributeEnd AttributeEnd MacroEnd WorldBegin Surface "plastic" AttributeBegin Color 0.000 0.000 0.000 Opacity 0.000 0.000 0.000 Attribute "identifier" "name" ["chaise2_1048581"] TransformBegin Transform [ 0.000000 0.000000 1.000000 0.000000 0.000000 1.000000 0.000000 0.000000 -1.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 1.000000 ] AttributeBegin Color 0.000 0.000 0.000 Opacity 0.000 0.000 0.000 Attribute "identifier" "name" ["chaise_0"] TransformBegin Transform [ 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 ] MacroInstance "chaise" TransformEnd AttributeEnd TransformEnd AttributeEnd AttributeBegin Color 0.000 0.000 0.000 Opacity 0.000 0.000 0.000 Attribute "identifier" "name" ["chaise2_1048578"] TransformBegin Transform [ 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 ] AttributeBegin Color 0.000 0.000 0.000 Opacity 0.000 0.000 0.000 Attribute "identifier" "name" ["chaise_0"] TransformBegin Transform [ 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 ] MacroInstance "chaise" TransformEnd AttributeEnd TransformEnd AttributeEnd WorldEnd Only one chair appear.... i should have a second one with a 90deg angle The composition of two Identity Matrix is Ok ( fortunatly) but an Identity with an other type of matrix dont seem to work. Gery DIVRY ( ZZVolume Daddy ) ARES Publisher 8, rue Victor Lagrange Phone: (+33) 72 80 16 30 69007 LYON Fax: (+33) 72 80 16 32 France Email: gery@ares.fdn.org Earth, Solar System, Galaxy MW1 NeXT Mail accepted
Newsgroups: comp.sys.next.bugs From: gery@ares.fdn.org (Gery Divry) Subject: Re: Renderman bug Message-ID: <1994Apr11.110445.996@ares.fdn.org> Sender: news@ares.fdn.org Organization: Ares - Lyon, France. References: <1994Apr11.092007.680@ares.fdn.org> Date: Mon, 11 Apr 1994 11:04:45 GMT In article <1994Apr11.092007.680@ares.fdn.org> gery@ares.fdn.org (Gery Divry) writes: : :After several tests :it seems that the composition of two transformations (Transform) is bugged :in renderman. I was thinking that this bug was fixed. : :ex : :Only one chair appear.... i should have a second one with a 90deg angle : :The composition of two Identity Matrix is Ok ( fortunatly) but an Identity :with an other type of matrix dont seem to work. : OOPS all My excuse to the renderman Team I was out of my shoes ... I shoud have use ConcatTransform and no Transform. How stupid I am .... It should work nicely now Gery DIVRY ( ZZVolume Daddy ) ARES Publisher 8, rue Victor Lagrange Phone: (+33) 72 80 16 30 69007 LYON Fax: (+33) 72 80 16 32 France Email: gery@ares.fdn.org Earth, Solar System, Galaxy MW1 NeXT Mail accepted
From: flexus!rfschtkt@maze.ruca.ua.ac.be (Raf Schietekat) Newsgroups: comp.sys.next.bugs Subject: Re: webster Date: 11 Apr 1994 12:08:31 -0500 Organization: UTexas Mail-to-News Gateway Sender: daemon@cs.utexas.edu Message-ID: <9404101448.AA01325@flexus> LS, Just a quick reminder, folks. There's a big file of ``Known Bugs in NextStep'', at cs.orst.edu:~ftp/pub/next/documents/KBNS.* (with * a wildcard), currently incarnation 32.0 (the first attempt at describing release 3.2). Among a wealth (well...) of major bugs and trivia, it describes the Webster bug. >>>>> KBNS.11.0.016_o3.013.2o Bug_NeXT.43680 Webster can't handle ``a'' Description: o3.013.1c: Webster crashes when asked to explain ``a'' (without the quotation marks<). c3.2o: Webster reports ``word not found'' when asked to explain ``a'', even though all the other letters seem to be explained. How to verify: Obvious. Urgency: Well, it's not a crash anymore< Confirmed: Robert_Brown (found), Subrata_Sircar (o3.013.1o, reported to KBNS), Raf_Schietekat (o3.0o_o3.2o) Report history: KBNS.32.1.rev <<<<< Juergen is probably searching for prefixes. Raf Schietekat, RfSchtkt@maze.ruca.ua.ac.be, Flanders, Belgium (real, i.e., with triangle in the Deliver button) NeXTmail preferred Addressing limitations: no !, %, .at or .uucp (? I should test this again)
From: ccwf@plato.klab.caltech.edu (Charles Fu) Newsgroups: comp.sys.next.bugs,comp.sys.next.programmer Subject: vfork does not work as advertised Date: 11 Apr 1994 18:15:33 GMT Organization: FTL Enterprises, Inc. Message-ID: <2oc446$ch8@gap.cco.caltech.edu> Under 3.1 and 3.2, there really is no vfork. For example, under 3.2 Motorola, vfork is implemented as vfork 0x50085b4 bral 0x5008578 <fork> 0x50085ba nop Either vfork needs to be fixed to work as documented or the man page for vfork should be changed to indicate that vfork is not really a vfork but just a cover for fork. Here are some key differences between a properly implemented vfork and fork: 1) vfork suspends the parent until the child dies; 2) parent and child share the same context--memory is not set to copy-on-write, so the child must be very careful not to mess up the context eventually be used by the parent. I've never relied upon the second property. I'm annoyed because I just spent a great deal of time tracking down a kernel panic that was due to expecting property (1) to suspend the parent so I didn't have to worry about a race condition (easily fixed in this instance with a wait). -ccwf
From: mgrbgc@cornflower (Bruce G. Calder) Newsgroups: comp.sys.next.bugs Subject: Workspace Errors Date: 12 Apr 1994 00:53:01 GMT Organization: Rose-Hulman Institute of Technology Message-ID: <2ocrdd$nog@master.cs.rose-hulman.edu> At Rose-Hulman Institute of Technology, we have 5 labs of NeXTs. Some users have had the problem of clicking between their favorite application and the workspace causing the Workspace Error panel to pop up which informs them that they are to logout. It only seems to happen to certain users frequently. During the course of a 3 hour test, one student had 4 Workspace errors. Does anybody know some of the causes of Workspace errors. Our site runs NextStep 3.1 with AFS 3.3 and a few NFS utilities. All of the user accounts reside on the AFS file servers. AFS has proven to be difficult at times. It's file caching from the network causes NeXT's to panic on sndplays of more than 3 seconds. Thanks, Bruce Calder mgrbgc@nextwork.rose-hulman.edu Unix Administrator
Newsgroups: comp.sys.next.bugs,comp.sys.next.programmer From: cedman@princeton.edu (Carl Edman) Subject: Re: vfork does not work as advertised In-Reply-To: ccwf@plato.klab.caltech.edu's message of 11 Apr 1994 18:15:33 GMT To: ccwf@plato.klab.caltech.edu (Charles Fu) Message-ID: <CEDMAN.94Apr11191005@capitalist.princeton.edu> Originator: news@nimaster Sender: news@Princeton.EDU (USENET News System) Organization: Princeton University References: <2oc446$ch8@gap.cco.caltech.edu> Date: Mon, 11 Apr 1994 23:10:05 GMT In article <2oc446$ch8@gap.cco.caltech.edu> ccwf@plato.klab.caltech.edu (Charles Fu) writes: Under 3.1 and 3.2, there really is no vfork. For example, under 3.2 Motorola, vfork is implemented as vfork 0x50085b4 bral 0x5008578 <fork> 0x50085ba nop Either vfork needs to be fixed to work as documented or the man page for vfork should be changed to indicate that vfork is not really a vfork but just a cover for fork. If I may quote the man page for vfork: "This system call will be eliminated when proper system sharing mechanisms are implemented. Users should not depend on the memory sharing semantics of vfork as it will, in that case, be made synonymous to fork." Carl Edman
From: eps@futon.SFSU.EDU (Eric P. Scott) Newsgroups: comp.sys.next.bugs,comp.sys.next.programmer Subject: Re: vfork does not work as advertised Followup-To: comp.unix.questions Date: 12 Apr 1994 05:54:13 GMT Organization: San Francisco State University Message-ID: <2odd25$3ue@nic-nac.CSU.net> References: <2oc446$ch8@gap.cco.caltech.edu> Not a bug. vfork() is merely advisory; *any* UNIX system is free to substitute "full" fork() for it (but never the other way around). [I.e. you should not depend on the parent being suspended until the child calls _exit() or execve().] Read the vfork(2) man page. Let me rephrase this more strongly: even on systems that implement vfork() distinctly from fork(), properly written code will always tolerate being compiled with cc -Dvfork=fork Followups to comp.unix.questions. -=EPS=-
From: nike@indirect.com (Laurence Canter) Newsgroups: comp.sys.next.bugs,pgh.food Subject: Green Card Lottery- Final One? Date: 12 Apr 1994 08:01:47 GMT Organization: Canter & Siegel Message-ID: <2odkhb$3go@herald.indirect.com> Green Card Lottery 1994 May Be The Last One! THE DEADLINE HAS BEEN ANNOUNCED. The Green Card Lottery is a completely legal program giving away a certain annual allotment of Green Cards to persons born in certain countries. The lottery program was scheduled to continue on a permanent basis. However, recently, Senator Alan J Simpson introduced a bill into the U. S. Congress which could end any future lotteries. THE 1994 LOTTERY IS SCHEDULED TO TAKE PLACE SOON, BUT IT MAY BE THE VERY LAST ONE. PERSONS BORN IN MOST COUNTRIES QUALIFY, MANY FOR FIRST TIME. The only countries NOT qualifying are: Mexico; India; P.R. China; Taiwan, Philippines, North Korea, Canada, United Kingdom (except Northern Ireland), Jamaica, Domican Republic, El Salvador and Vietnam. Lottery registration will take place soon. 55,000 Green Cards will be given to those who register correctly. NO JOB IS REQUIRED. THERE IS A STRICT JUNE DEADLINE. THE TIME TO START IS NOW!! For FREE information via Email, send request to cslaw@indirect.com -- ***************************************************************** Canter & Siegel, Immigration Attorneys 3333 E Camelback Road, Ste 250, Phoenix AZ 85018 USA cslaw@indirect.com telephone (602)661-3911 Fax (602) 451-7617
Newsgroups: comp.sys.next.bugs From: A.R.J.B.Simonis@twi.tudelft.nl (Armand Simonis) Subject: Bug in Text object with -sizeToFit ? Message-ID: <Co570v.5r5@dutiws.twi.tudelft.nl> Keywords: Text Sender: news@dutiws.twi.tudelft.nl (TWI News Administration) Organization: Delft University of Technology Date: Tue, 12 Apr 1994 10:41:18 GMT Hi, We seem to have a problem with [ Text sizeToFit ]. If you insert a string with a '\n' with the message -setText:(char*) in a Text object after the creation of this object, you get strange results: - sizeToFit doesn't work properly. - the text you see isn't wrapped in the same way as the text really (?) present in the text object itself. If you compile textbug.m , you can see for yourself ! Any explanation ( is this a bug, or are we buggy ? ) or help is greatly appreciated before we send it to NeXT. Thanx, Armand & Dimitri. BTW under my signature : Makefile & textbug.m ===== Armand Simonis == A.R.J.B.Simonis@twi.tudelft.nl ============= Delft University of Technology || Epona Software Technology Fac.Applied Math & Computer Science || Beuckelaar3, 2681 NG Monster The Netherlands -------------------------------------------------------------------- 'Every correct and minimal consuming algorithm must be elegant' ==================================================================== == >8 == Makefile =========================== >8 ====================== textbug : textbug.m cc -Wall -o textbug textbug.m -lNeXT_s -lsys_s = >8 textbug.m === >8 ================================================== #import <appkit/appkit.h> static char* Text1P = "aaaa\n bbbbb cccc dddd eee ffff gggg hhhh iiii" " jjjj kkkk"; static char* Text2P = "aaaa bbbbb cccc dddd eee ffff gggg hhhh iiii" " jjjj kkkk"; id CreateText(NXRect FrameRect, char* TextP ) { id TextID; NXSize MaxSize={300,1e10}; NXSize MinSize={1,1}; TextID = [ [Text alloc] initFrame:&FrameRect text:"" alignment:NX_LEFTALIGNED ]; [ TextID setHorizResizable:YES ]; [ TextID setVertResizable:YES ]; [ [ [ TextID setOpaque:YES] setEditable:YES ] setCharWrap:NO ]; [ TextID setMaxSize:&MaxSize ]; [ TextID setMinSize:&MinSize ]; [ TextID setSel:0:0 ]; [ TextID setText:TextP ]; // [ TextID calcLine ]; // Doesn't matter whether calcLine message is sent or not [ TextID sizeToFit ]; // sometimes this doesn't calculate a correct height !! return TextID; } void Start(void) { id WindowID,ViewID,MenuID; NXRect Rect, Frame1, Frame2; NXSetRect(&Rect,100,100,700,400); WindowID = [ [ Window alloc ] initContent: &Rect style : NX_TITLEDSTYLE backing : NX_BUFFERED buttonMask : NX_MINIATURIZEBUTTONMASK defer : NO ]; [ WindowID setTitle: "Text Bug Demo" ]; [ WindowID display ]; // menu.... MenuID = [ [ Menu alloc] initTitle: "Text Bug" ]; [ MenuID addItem: "Quit" action: @selector(terminate:) keyEquivalent: 'q' ]; [ MenuID sizeToFit ]; [ NXApp setMainMenu: MenuID ]; // view... ViewID = [ [ View alloc ] initFrame:&Rect ]; [ [ ViewID setOpaque: NO ] setFlipped:YES ]; [ WindowID setContentView: ViewID ]; // text.... NXSetRect(&Frame1, 0, 0, 0, 0); NXSetRect(&Frame2, 350, 0, 0, 0); [ ViewID addSubview: CreateText(Frame1,Text1P) ]; [ ViewID addSubview: CreateText(Frame2,Text2P) ]; /// start... [ WindowID display ]; [ WindowID makeKeyAndOrderFront: nil ]; } void main(void) { NXApp = [Application new]; Start(); [NXApp run]; [NXApp free]; exit(0); } === >8 =========================== >8 =====================
Newsgroups: comp.sys.next.bugs From: yanik@planon.qc.ca (Yanik Crepeau) Subject: Re: webster Message-ID: <1994Apr12.011104.27393@planon.qc.ca> Sender: yanik@planon.qc.ca (Yanik Crepeau) References: <2o40gd$j20@newshost.lanl.gov> Date: Tue, 12 Apr 1994 01:11:04 GMT In article <2o40gd$j20@newshost.lanl.gov> wolf@lanl.gov (David R Wolf) writes: > here's a strange one: > > seems it's always possible to crash webster (NS 3.0) by searching for the > word "a". > > -- > ======================================================================= > David R. Wolf wolf@lanl.gov LANL, MS P940, 87545 (505) 667-3813 > ======================================================================= Not true. I have tried it and after a while (3 minutes) I have got all the words (Thesaurus only) begining with an "a". -- Yanik Crepeau The power of OOP used to stop when Programmer the programmer typed "make". With Planon Telexpertise NEXTSTEP and PDO, that is not true E-Mail: yanik@planon.qc.ca (NeXT) anymore!
From: kheit@gandalf.rutgers.edu (John Kheit) Newsgroups: comp.sys.next.sysadmin,comp.sys.next.bugs,comp.sys.next.hardware Subject: Re: EATA timeout report #2 Message-ID: <Apr.12.12.59.29.1994.3475@gandalf.rutgers.edu> Date: 12 Apr 94 16:59:30 GMT References: <2npsso$9p@marsu.tynet.sub.org> <2ns15b$iu3@tribune.usask.ca> <next2.765579196@info2.rus.uni-stuttgart.de> Followup-To: comp.sys.next.sysadmin Organization: Rutgers Univ., New Brunswick, N.J. next2@info2.rus.uni-stuttgart.de (Markus Wenzel) writes: >Just be happy about it. But I think more than 10 persons with similar >problems are enough to point out that something has to be done against it. >If your system were unstable, you'd know what I meant with 'unreliable'. >BTW: What devices did you connect to your 2022? Fast-SCSI-II harddisk drives? >My theory is that it only happens with new fast drives. No, I've had this happen with older drives too. It was particularly bad in 3.1 with a NICE motherboard and old DPT ROMS and 10m/sec transfer rates. I dont get catastrophic EATA's anymore, they just time out the system, re- synch, time out, re-synch etc... A hard reboot seems to clear out this cold. Anyway, if you have this promblem, somethings that seem to aleviate it are: - Get NS 3.2 - Reconfigure the DPT properly, and use only 5m/sec transfers also maybe push the smart rom from c800 to d800 (im not sure about the exact address space, so dont quote me on it). - Get the newest DPT ROMS for your card. BTW, cache seems to screw with opticals in a really horrible way (talkin big 16megs+ cache). Later, and good luck John
From: Jim_Mann@transarc.com Newsgroups: pgh.food,comp.sys.next.bugs Subject: Re: Green Card Lottery- Final One? Date: Tue, 12 Apr 1994 14:07:15 -0400 Organization: Carnegie Mellon, Pittsburgh, PA Message-ID: <4heiBHmSMUE3RkApUN@transarc.com> References: <2odkhb$3go@herald.indirect.com> In-Reply-To: <2odkhb$3go@herald.indirect.com> What does this have to do with food, NeXT bugs, the C language, or any of the hundreds of other groups to which it was posted. There really are better ways to get out this information than to blast it out to every news group in existence. ****************************************************************** Jim Mann jmann@transarc.com Transarc Corporation The Gulf Tower, 707 Grant Street Pittsburgh, PA 15219 (412) 338-4442
From: Michael Witbrock <mjw+@cs.cmu.edu> Newsgroups: comp.sys.next.bugs,pgh.food Subject: Re: Green Card Lottery- Final One? Date: Tue, 12 Apr 1994 18:03:35 -0400 Organization: Carnegie Mellon, Pittsburgh, PA Message-ID: <Ehelera00hsBRMeV4b@cs.cmu.edu> References: <2odkhb$3go@herald.indirect.com> In-Reply-To: <2odkhb$3go@herald.indirect.com> You don't need their info. Ask your foreign advisor instead. The procedure is extremely simple and the advice of lawyers is unlikely to be helpful.
From: eps@futon.SFSU.EDU (Eric P. Scott) Newsgroups: comp.sys.next.bugs Subject: mktime() incompatibility with SVR4? Date: 13 Apr 1994 09:36:52 GMT Organization: San Francisco State University Message-ID: <2ogefk$p6k@nic-nac.CSU.net> Software that "worked just fine" on NEXTSTEP 2.1 and 3.2 produced incorrect results when ported to Solaris and HP/UX. These systems require .tm_isdst be set to -1 to solicit "figure out what the correct timezone is" behavior (taking 0 or 1 as meaning "the programmer knows better"). Comparing the man pages, I infer that NEXTSTEP's implementation ignores .tm_isdst; setting it explicitly to -1 doesn't seem to hurt anything and makes the SysV-derived systems happy. -=EPS=- (who should have been asleep hours ago)
Newsgroups: comp.sys.next.bugs From: ralf@rafa.in-berlin.de (Ralf Neumann) Subject: Send me pictures for /LocalLibrary/Images/People Message-ID: <1994Apr13.072051.7421@rafa.in-berlin.de> Sender: ralf@rafa.in-berlin.de Organization: DRN Date: Wed, 13 Apr 1994 07:20:51 GMT Hallo, After I installed the faces.tar.Z from cs.orst. I am missing nearly everybody from the next newsgroup. I will be very happy to see who is writing so please send me small pictures by NeXT-email. But take care that they are really really small (hope not to get to much flames but sorry what's multimedia without medias ?-) and from the size not bigger than 64x64 pixels. Please remember: give it the correct name i.e.: fuuu@grgrgrg.bla.com must be fuu.grgrgrg.bla.com.tiff Thanks a lot in advance Ralf If I get enough together I will put them on an ftp-server and post where to get them or send them to Carl Edman who is moderating the faces library since a long time. %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% *| Ralf Neumann | ralf@rafa.in-berlin.de | Try NeXTmail please! |* *| ************ | Voice +49 30 / 321 78 84 | Fax +49 30 / 321 28 68|* %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Newsgroups: comp.sys.next.bugs From: cuilla@il.us.swissbank.com (Chris Cuilla) Subject: Problems with the Intel TokenExpress, Epson NX & NEXTSTEP? Message-ID: <1994Apr13.184407.21320@il.us.swissbank.com> Keywords: TokenExpress Epson NEXTSTEP Sender: root@il.us.swissbank.com (Operator) Organization: Swiss Bank Corporation CM&T Division Date: Wed, 13 Apr 1994 18:44:07 GMT Are there any known compatibility problems with the Intel TokenExpress Token-ring card, the Epson NX PC and NEXTSTEP? Work arounds? Problems using the IBM Token-ring card? Is that one supported yet? -- Chris Cuilla --------------------------------------------------------------------- NEXTSTEP Developer ccuilla@cuilla.plexare.com ---------------------------------------------------------------------
From: robert@steffi.demon.co.uk (Robert Nicholson) Newsgroups: comp.sys.next.bugs Subject: Re: Send me pictures for /LocalLibrary/Images/People Date: 13 Apr 1994 19:21:30 +0100 Organization: me organised, that's a joke. Message-ID: <2ohd7a$27i@steffi.demon.co.uk> References: <1994Apr13.072051.7421@rafa.in-berlin.de> Oh and one other thing. Post in the appropriate groups. For this you would probably find comp.sys.next.misc adequate not comp.sys.next.* Some FTP sites have People under the Images heirarchy. -- "Kill files are for pacifists" (ASCII for text only messages)
From: luomat@alleg.EDU (Timothy J. Luoma) Newsgroups: comp.sys.next.bugs,comp.sys.next.software Subject: Re: Funny NewsGrazer bug Date: 13 Apr 1994 17:51:36 -0500 Organization: UTexas Mail-to-News Gateway Sender: daemon@cs.utexas.edu Message-ID: <9404132243.AA16984@alleg.EDU> In comp.sys.next.software article <1994Apr7.182744.1689@rna.indiv.nluug.nl> you wrote: When I have an open mail window in Mail.app (a message I am composing) and I press Reply in NG, I get a new empty compose window and all my stuff in the orignal compose window has been replaced by the stuff from NG. Only the formatting (bold etc) is still intact. I've lost a bunch of "being composed" messages this way. There is a workaround (although not a great one) which has worked for me. Before you leave Mail.app, miniaturize the Compose window. Then when you click on reply it will start a new window and not kill the old. Tim -- Timothy J. Luoma Email: luomat@alleg.edu NeXT Mail Welcomed MIME Not Box 931 Allegheny College Meadville, PA 16335 USA
Newsgroups: comp.sys.next.bugs From: yoda@cis.uni-muenchen.de (Marc Guenther) Subject: Re: BackSpace bug? Sender: news@informatik.uni-muenchen.de (News System) Message-ID: <Co9IM4.8EH@informatik.uni-muenchen.de> Date: Thu, 14 Apr 1994 18:42:04 GMT References: <Cnn53x.8q@dsinc!flash> Organization: Institut fuer Informatik der Universitaet Muenchen In article <Cnn53x.8q@dsinc!flash> flash!jon@myxa.com writes: > > I think this happens when either you or root has a null password. Both your own > password and the root password will unlock the screen, in addition to your > Backspace password. Just make sure both accounts have a password. > "It's not a bug, it's a feature" BackSpace should warn the user, if one of the passwords is nil ! -- Marc Guenther ------------------------------------------------------------------------- Centrum fuer Informations | Wagmuellerstr. 23 | Phone: +49 89 211 0670 und Sprachverarbeitung | 80538 M"unchen | Fax: +49 89 211 0674 University of Munich | Germany | yoda@cis.uni-muenchen.de -------------------------------------------------------------------------
Newsgroups: comp.sys.next.bugs From: peter@corsica (Peter Eisch) Subject: New ThumbWheel broken Message-ID: <Co996t.380@news.cis.umn.edu> Sender: news@news.cis.umn.edu (Usenet News Administration) Organization: University of Minnesota Hospital and Clinic, Laboratory Information Services Date: Thu, 14 Apr 1994 15:21:20 GMT The new NeXT Mini-Example "ThumbWheel" is broken. This bug does not appear in the previous version. (The fix is at the bottom.) If you're using it, here is what to look for: TWCellDraw.m:35-63 - drawInside:(const NXRect *)theFrame inView:view { ... code omitted ... // Draw the background if([self isLinear]) { ... code omitted ... } // otherwise if([self isRadial]) else { NXSize imageSize; [image getSize:&imageSize]; > ^^^^^ what if image is nil? > if(NOTEQUAL(imageSize.width,NX_WIDTH(theFrame)) || NOTEQUAL(imageSize.height,NX_HEIGHT(theFrame))) [self generateImage:theFrame]; [image composite:NX_COPY toPoint:&theFrame->origin]; } If 'image' is nil, then imageSize will be initialized. If the real problem that 'image' isn't initialized properly, then I guess that should be fixed too. As my program was running, imageSize would be: (gdb) p imageSize $11 = { width = -NaN(0xffffffff), height = NaN(0x7fff0000) } which would cause the NOTEQUAL (because of ABS()) to evaluate to NaN and then the statement evaluates to FALSE, so the instance variable 'image' never gets to be an image until the wheel is click/dragged... The fix is to initialize imageSize to a value, perhaps 0's. So line 56 should be: ... NXSize imageSize = {0.0, 0.0}; ... peter PS. Thanks for the support of this cool object Jeff! -- peter@tahiti.umhc.umn.edu
Newsgroups: comp.sys.next.bugs From: glen@instep.wimsey.bc.ca Subject: printing a DBTableView Message-ID: <1994Apr12.223430.6239@instep.wimsey.bc.ca> Keywords: DBTableView print Sender: glen@instep.wimsey.bc.ca Organization: InStep Mobile Communications Inc. Date: Tue, 12 Apr 1994 22:34:30 GMT I find that when I print a DBTableView, centered and right justified columns print with the fist character in the center or at the right of the field, whith the remaining data spilling off to the right. A previous suggestion to use Eric Seymour's TablePrinter Palette didn't help. Anyone else have this problem? Is there a fix or a work around? -- Glen Biagioni Software Developer glen@instep.wimsey.bc.ca (small NeXTmail accepted) 604 872-7116 604 872-7125 fax
From: william@moomin.berkeley.edu (William E. Grosso) Newsgroups: comp.sys.next.bugs Subject: free () ?? Date: 15 Apr 1994 22:09:54 GMT Organization: University of California, Berkeley Message-ID: <2on3bi$128@agate.berkeley.edu> I seem to be getting some weird behaviour with "free". Basically, I'm mallocing and freeing a fair amount of memory and occasionally get the error Memory access exception on address 0x3f800000 on the line free(temporary) where temporary is pointer to 0x2b2c40 (and has previously been malloced). Eliminating the free() makes the code run perfectly (albeit somewhat hoggishly). All my pointers are correct. What gives ? This is not in KBNS (I don't think). Although, it is suspiciously similar to one of the indexing kit bugs. Bill Grosso
From: telesoft@delphi.com Newsgroups: comp.sys.next.bugs Subject: Problems getting NS3.2 to recognize DOS diskette Date: Fri, 15 Apr 94 22:26:56 -0500 Organization: Delphi (info@delphi.com email, 800-695-4005 voice) Message-ID: <hkyuiZ4.telesoft@delphi.com> I'm having trouble getting NeXTStep 3.2 to recognize any DOS diskettes. Has anyone else experienced similiar problems. This use to work in 3.1.
Control: cancel <2odkhb$3go@herald.indirect.com> Newsgroups: comp.sys.next.bugs,pgh.food From: alex@hal.rhein-main.de (Alexander Lehmann) Subject: REPOST: cmsg cancel <2odkhb$3go@herald.indirect.com> Message-ID: <CoCux1.KB@hal.rhein-main.de> References: <SuperBot.1718@megalith.miami.fl.us> Date: Sat, 16 Apr 1994 14:00:36 GMT Due to a format error in the Date: line, this message could not be processed by cnews. SuperBot (SuperBot@megalith.miami.fl.us) wrote: : cancel <2odkhb$3go@herald.indirect.com> : from nike@indirect.com (Laurence Canter) : in newsgroup(s) comp.sys.next.bugs,pgh.food : : This article canceled bye an automated AREXX Script
From: wjadams@hubcap.clemson.edu (William J Adams) Newsgroups: comp.sys.next.bugs Subject: InstantTeX prob on ns/fip ver 3.2 Date: 17 Apr 94 22:52:24 GMT Organization: Clemson University Message-ID: <wjadams.766623144@hubcap> I'm running NeXTSTEP 3.2 / fip and just got down the "latest version" that I've been able to find of Instant TeX, and installed it. It claims, in the installation package that it supports white hardware. Unfortunately it wouldn't completely install, the program called after installation (TeXViewremover.app) is only for blackhardware. I tried to get around that, and it did install, but my computer said the programs had been damaged. Anybody have any idea? Thanks, Bill D. Hippie Cat ps. You can mail me directly at wjadams@hubcap.clemson.edu.
From: Charles William Swiger <infidel+@CMU.EDU> Newsgroups: comp.sys.next.bugs Subject: Re: free () ?? Date: Sun, 17 Apr 1994 19:25:23 -0400 Organization: Carnegie Mellon, Pittsburgh, PA Message-ID: <EhgQJX600iUvA5sX5K@andrew.cmu.edu> In-Reply-To: <2on3bi$128@agate.berkeley.edu> Excerpts from netnews.comp.sys.next.bugs: 15-Apr-94 free () ?? by William E. Grosso@moomin > I seem to be getting some weird behaviour with "free". > Basically, I'm mallocing and freeing a fair amount of > memory and occasionally get the error > > Memory access exception on address 0x3f800000 [munch] Perhaps you have corrupted the internal data structures that malloc maintains. Try using the malloc_debug() call (read "man malloc") with a high debugging level, and see whether this can detect any problems. Also take a look at the NXMallocCheck() call. -Chuck Charles William Swiger - WhiteLight Systems | "All the world's a stage, and" --------------------------------------------+ "we are merely players...." AMS & normal mail: infidel@cmu.edu | NeXTmail: chuck@cswiger.slip.andrew.cmu.edu | "Semper ubi sub ubi."
Newsgroups: comp.sys.next.bugs From: Zacharias J. Beckman <zac@dolphin.com> Subject: NewsGrazer Lockups! Help! Message-ID: <1994Apr17.191612.3025@dolphin.com> Sender: zac@dolphin.com Organization: Dolphin Technologies Inc. Distribution: usa Date: Sun, 17 Apr 1994 19:16:12 GMT Has anyone out there experienced a problem where NewsGrazer 3.0 (75) works fine for about 10 or 15 minutes, then goes into an infinite spinning disk lockup? I've tried the obvious things, such as deleting my .newsrc and checking preferences, etc. The next try, I suppose, will be to download the binary again and hope that fixes it... 'til then, I just use NG until it locks, then kill it, then start again... -- Zacharias J. Beckman - Dolphin Technologies Inc. - zac@dolphin.com - NeXTMAIL! To be "matter of fact" about the world is to blunder into fantasy.... and dull fantasy at that, as the real world is strange and wonderful. --- R. A. Heinlen Those opinions I express herein are my own, I'm fairly sure. --- Z. J. Beckman
Newsgroups: comp.sys.next.bugs,comp.sys.next.programmer From: rfi@jeans.fokus.gmd.de (Robert Fischer) Subject: PDO: No bug, but a feature!? Message-ID: <1994Apr18.130923.21215@fokus.gmd.de> Sender: news@fokus.gmd.de (News system) Organization: GMD-Fokus Date: Mon, 18 Apr 1994 13:09:23 GMT Hi, I do some programming with PDOs and ran into trouble with NXConnection's "registerRoot:withName:". The connection was established correctly but no messages came through. The death of the connection was also reported correctly! My trial-and-error lasted for several hours when I tried to change the name, which I used to register root. That was the solution: I've used the name of the application to register, but with a leading lower letter. I think that the application name is reserved by another port, regardless of the case of the letters. There was no word about this in the documentation but 2 minutes of thinking about it should have solved my problem! Bye, Robert. -- ------------------------------------------------------------------------ Robert Fischer @ GMD-Fokus -------- __o ------- _`\<,_ fischer@fokus.gmd.de ------- (*)/ (*) ## NeXT-Mail welcome ## ------------------------------------------------------------------------
From: fliu@solo.eng.uci.edu (Feng Liu) Newsgroups: comp.sys.next.bugs Subject: NSI 3.2 panic with PAS studio and Adaptec 1542b Date: 18 Apr 1994 14:58:58 GMT Organization: University of California, Irvine Message-ID: <2ou77i$i5b@news.service.uci.edu> My NS/I 3.2 with an Adaptec 1542b had been running without any problem for several months. A week ago, I installed a PAS studio 16 sound card and a NEC 500 3x internal CDROM. Ever since then, the system started to crash rather randomly. It boots alright (sometime with an error message: ttysccc1 receive error 900 or ttysccc1 receive erorr 901, but it runs alright), after a while it all of a sudden panics, and then I had to reboot the machine. The time between booting and panic is rather random, sometimes a few minutes, in this case most probably I have Backspace running and there is not a CD-ROM in the CDROM Drive. Other times, it takes hours, in this case most of the time I did put a CD ROM in the CDROM Drive and I probably did not run backspace. Reading Nextanswers on the PAS 16 spectrum driver, I found it says that on some systems one can not play sound with over 22kHz rate, otherwise system may freeze or panic if one has and Adaptec card. But I am not playing any sound, in fact at times I did not even install the sound card driver. Any help or clue is appreciated. Thank you. Feng Liu Department of Mechanical and Aerospace Engineering UC, Irvine e-mail: fliu@uci.edu
Newsgroups: comp.sys.next.bugs From: nijenhui@sun005.research.ptt.nl (Wim Nijenhuis) Subject: Aaargh, can't eject disk Message-ID: <1994Apr18.163948.27487@news.research.ptt.nl> Sender: nijenhui@sun005 (Wim Nijenhuis) Organization: PTT Research, Groningen, The Netherlands Date: Mon, 18 Apr 1994 16:39:48 GMT Hi all, I'm not a regular user of the Nextworkstation. I inserted an ordinairy dos format disk but there was no disk icon in the filemanager. Also the eject option is disabled. I'm using NeXTStep 3.0. Is this a known bug? What's should I do? The console mentioned something about: "Mounted floppy disk at /3F f91{@ F" Is there another option than using the screwdriver :-) Thanks, Wim e-mail: r.w.nijenhuis@research.ptt.nl
From: sanguish@digifix.com Newsgroups: comp.sys.next.announce,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.bugs,comp.sys.next.programmer,comp.sys.next.marketplace Subject: Quick Guide to NEXTSTEP information on the Internet Date: 18 Apr 1994 15:25:40 -0400 Organization: Next Announcements Message-ID: <2oumrk$mij@digifix.digifix.com> This post is made weekly, to help 'point' users to more NEXTSTEP information Topics include: NEXTSTEP World Wide Web Product Information Server online NEXTSTEP Product Information Mail Server online comp.sys.next newsgroups related newsgroups comp.sys.next newsgroups mailing list ftp sites NeXTanswers Much of this information is also available using the World Wide Web, <http://digifix.digifix.com/index.html> NEXTSTEP World Wide Web Product Information Server online --------------------------------------------------------- A product directory built around the World Wide Web system, this will allow full multimedia announcements by NEXTSTEP developers, as well as the ability to browse the available products for NEXTSTEP. This service is online now, and can be reached at http://digifix.digifix.com/ it can be reached using OmniWeb (available from ftp.omnigroup.com) or Mosaic. The entries are coming in quite quickly, and I'll be moving the entire NEXTSTEP Third Party Catalog contents in as soon as they become available. NEXTSTEP Product Information Mail Server online --------------------------------------------------------- The NEXTSTEP Product Information Mail Server is now available for product literature and pricing from NEXTSTEP developers. You can get information on using the mail server at ns-products@digifix.com Suggestions or comments can be directed to me at sanguish@digifix.com comp.sys.next.* newsgroups -------------------------- Comp.Sys.Next.Advocacy This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. Comp.Sys.Next.Announce Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) The NEXTSTEP FAQs are posted here monthly as well. This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Comp.Sys.Next.Bugs A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT- specific groups as well. Comp.Sys.Next.Hardware Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. Comp.Sys.Next.Marketplace NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. Comp.Sys.Next.Misc For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! Comp.Sys.Next.Programmer Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. The NEXTSTEP programmer FAQs are posted here. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. Comp.Sys.Next.Software This is a place to talk about [third party] software products that run on NEXTSTEP systems. Comp.Sys.Next.Sysadmin Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. related Newsgroups ------------------ Comp.Soft-Sys.Nextstep Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. Comp.Lang.Objective-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. Comp.Object Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original Comp.Sys.Next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News ------------------------------------------- Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep-d next-advocacy-d next-announce-d next-bugs-d next-hardware-d next-marketplace-d next-misc-d next-programmer-d next-software-d next-sysadmin-d (For a full description, send mail saying LISTS to <digestif@antigone.com>). The subscription syntax is essentially the same as LISTSERV's. To subscribe, send a message to <digestif@antigone.com> saying: SUB Listname YourName Example: SUB next-hardware-d John Doe The ftp sites ------------- cs.orst.edu: The main site for North American submissions nova.cc.purdue.edu: Lots of older stuff, but very short on disk space ftp.informatik.uni-muenchen.de: In Germany. terra.stack.urc.tue.nl (Dutch NEXTSTEP User Group) and cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp.next.com: See the below ftp.next.com and NextAnswers@next.com ------------------------------------- From the document 1000_Help from ftp.next.com Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, or you can transfer them by anonymous ftp. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to NeXTanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO <your-address> If you have any problem with the system or suggestions for improvement, please send mail to NeXTanswers-request@NeXT.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to NeXTanswers-request@NeXT.com. Written by: Eric P. Scott eps@toaster.SFSU.EDU and Scott Anguish sanguish@digifix.com Additions from: Greg Anderson (Greg_Anderson@afs.com) and Michael Pizolato (Michael_Pizolato@afs.com)
From: luomat@alleg.EDU (Timothy J. Luoma) Newsgroups: comp.sys.next.bugs Subject: Re: Aaargh, can't eject disk Date: 18 Apr 1994 18:18:05 -0500 Organization: UTexas Mail-to-News Gateway Sender: daemon@cs.utexas.edu Message-ID: <9404182317.AA07063@alleg.EDU> Hi all, I'm not a regular user of the Nextworkstation. I inserted an ordinairy dos format disk but there was no disk icon in the filemanager. Also the eject option is disabled. I'm using NeXTStep 3.0. Is this a known bug? What's should I do? Is there another option than using the screwdriver :-) Well, you could upgrade to 3.1 or 3.2.... but if you were looking for a slightly easier solution, there is a DiskEjectFix patch on the archive (which I sent to the poster because I was having a nice day) as well as a less than perfect solution of a binary to eject from the commandline, which I also sent the poster.... I'm not sure that the binary would work, but the patch worked great when we were on 3.0. TjL --- Timothy J. Luoma Email: luomat@alleg.edu (Shell=tcsh) Workstation Environment using NeXTSTEP 3.1 Motorola MAIL: NeXT YES / MIME Mail NO No Root access, no super-user access
Newsgroups: comp.sys.next.bugs From: csaldanh@mae.carleton.ca (Chris Saldanha) Subject: Re: Aaargh, can't eject disk Message-ID: <CoHoHv.E08@cunews.carleton.ca> Sender: news@cunews.carleton.ca (News Administrator) Organization: Carleton University References: <1994Apr18.163948.27487@news.research.ptt.nl> Date: Tue, 19 Apr 1994 04:29:55 GMT Wim Nijenhuis (nijenhui@sun005.research.ptt.nl) wrote: : I inserted an ordinairy dos format : disk but there was no disk icon in the filemanager. Also the eject option is : disabled. I'm using NeXTStep 3.0. Is this a known bug? What's should I do? : The console mentioned something about: "Mounted floppy disk at /3F f91{@ F" : Is there another option than using the screwdriver :-) I hope you'd use a pin, and not take the case off... ;-) When the console says something like that, go up to the root level of the WorkSpace fileviewer, and look for a directory with the name given on the console (i.e. "3F f91{@ F" in this case). Click on that directory, and voila! the contents of your floppy. Also when you click on that directory, the Eject menu option will become available again, allowing you to eject the disk. I find this happens mostly with 720K (Double density) disks. --Chris Chris Saldanha -------------------------------------- Carleton University |"The eternal silence of these infinite| chris@computerActive.on.ca (NeXTMail) | spaces terrifies me." -Blaise Pascal| csaldanh@mae.carleton.ca (NeXT/MIME) ------------------:-o-----------------
From: robert@amo.mit.edu(Robert Lutwak) Newsgroups: comp.sys.next.bugs Subject: Re: Aaargh, can't eject disk Date: 19 Apr 1994 12:52:04 GMT Organization: Massachvsetts Institvte of Technology Message-ID: <2p0k5k$qfa@senator-bedfellow.MIT.EDU> References: <9404182317.AA07063@alleg.EDU> In article <9404182317.AA07063@alleg.EDU> luomat@alleg.EDU (Timothy J. Luoma) writes: > Well, you could upgrade to 3.1 or 3.2.... but if you were looking for > a slightly easier solution, there is a DiskEjectFix patch on the > archive (which I sent to the poster because I was having a nice day) > as well as a less than perfect solution of a binary to eject from the > commandline, which I also sent the poster.... I'm not sure that the > binary would work, but the patch worked great when we were on 3.0. Hmmm... While we're on the subject, why is it that I still get the following message on my console (even though we're up to 3.2)? Console Output: Software Version 3.2 (Thunder5S) probing for DOS NS3.0 disk eject bug workaround. Robert -- Robert Lutwak robert@amo.mit.edu MIT Atomic Resonance and Spectroscopy Laboratory --- Everything I say via this computer belongs to MIT. --- ---> NeXTmail always welcome <---
From: izumi@pinoko.berkeley.edu (Izumi Ohzawa) Newsgroups: comp.sys.next.bugs,comp.sys.next.software Subject: Dig Librarian (3.2) pushes search word to pasteboard for every search Date: 19 Apr 1994 17:10:58 GMT Organization: University of California, Berkeley Distribution: world Message-ID: <2p13b2$q63@agate.berkeley.edu> It looks like every time you press the search button in DL, it pushes the key words onto the pasteboard. This must be a bug as I can't see why that would be useful. This probably also has to do with why Igor.app that comes with AFS WriteUp gets launched on the first press of the search button in DL. -- Izumi Ohzawa [ $@Bg_78^=;(J ] USMail: University of California, 360 Minor Hall, Berkeley, CA 94720 Telephone: (510) 642-6440 Fax: (510) 642-3323 Internet: izumi@pinoko.berkeley.edu (NeXT and MIME mails OK)
Newsgroups: comp.sys.next.bugs,comp.sys.next.software From: davis@sonata.cc.purdue.edu (Robert Davis) Subject: Re: Dig Librarian (3.2) pushes search word to pasteboard for every search Sender: news@mozo.cc.purdue.edu (USENET News) Message-ID: <CoIqM3.ouo@mozo.cc.purdue.edu> Date: Tue, 19 Apr 1994 18:13:15 GMT References: <2p13b2$q63@agate.berkeley.edu> Organization: Purdue University, West Lafayette, IN In article <2p13b2$q63@agate.berkeley.edu>, Izumi Ohzawa <izumi@pinoko.berkeley.edu> wrote: > >It looks like every time you press the search button in DL, it >pushes the key words onto the pasteboard. This must be a bug >as I can't see why that would be useful. > Hmm. That doesn't seem to be happening here. I wish that DL would place searches onto the Find pboard, however. It does this for searches from the Find Panel (wihin a single document) but not for searches done on targets. Rob -- | Robert Davis davis@sonata.cc.purdue.edu | "Look up, Hannah. The clouds are lifting." NeXT Mail accepted --
Newsgroups: comp.sys.next.sysadmin,comp.sys.next.bugs From: cbarton@DaBoyz.dgcp.doc.ca (Casey Barton) Subject: Re: Configuring Serial line to allow 8-N-1 dial-in ? Message-ID: <1994Apr19.170944.1687@clark.dgim.doc.ca> Sender: news@clark.dgim.doc.ca (#Usenet News) Organization: Communications Canada References: <2osffr$b5o@senator-bedfellow.MIT.EDU> Date: Tue, 19 Apr 94 17:09:44 GMT Robert Lutwak writes > I've tried all possible combinations of ap:np:p8, > etc. in gettytab, but whatever I try, I get gibberish (or worse) > on my vt100 terminal (plugged right into the port) unless I set > the terminal to 7-e-1. > > Has anybody else succeeded or is this a bug? I've been trying to do exactly this for the last 3 days. I thought it was just me. (This is my first try at UNIX admin...) I just called NeXT, and the techdroid told me the there were no documented bugs with the serial communications. I've tried several different parity settings, none of which will get NeXT to talk to the modem in anything other than 7E1. (7O1 doesn't work either.) I'm starting to think this is a bug -- but I'd think this would have been a very visible one -- 8N1 dialin seems to be the most common, nowadays. So, does *anyone* out there have a serial device successfully running at 8N1 on an NSFIP 3.2 system? If not, I'll report back to NeXT myself...
Newsgroups: comp.sys.next.sysadmin,comp.sys.next.bugs From: kramer@fragile.termfrost.org (Mike Andrews) Subject: Re: Configuring Serial line to allow 8-N-1 dial-in ? Message-ID: <CoJ207.5Ls@fragile.termfrost.org> Organization: Terminal Frost, Springfield OH References: <2osffr$b5o@senator-bedfellow.MIT.EDU> <1994Apr19.170944.1687@clark.dgim.doc.ca> Date: Tue, 19 Apr 1994 22:19:18 GMT cbarton@DaBoyz.dgcp.doc.ca (Casey Barton) writes: >Robert Lutwak writes >> I've tried all possible combinations of ap:np:p8, >> etc. in gettytab, but whatever I try, I get gibberish (or worse) >> on my vt100 terminal (plugged right into the port) unless I set >> the terminal to 7-e-1. >> >> Has anybody else succeeded or is this a bug? > I've been trying to do exactly this for the last 3 days. I thought it >was just me. (This is my first try at UNIX admin...) I just called NeXT, >and the techdroid told me the there were no documented bugs with the >serial communications. > I've tried several different parity settings, none of which will get >NeXT to talk to the modem in anything other than 7E1. (7O1 doesn't work >either.) I'm starting to think this is a bug -- but I'd think this would >have been a very visible one -- 8N1 dialin seems to be the most common, >nowadays. Put :p8:zp at the front of the Default entry in gettytab, and it should start behaving.... zp is zero parity, and p8 is 8-bit, thus N81. Seems to work for me, anyway... While we're on the subject of gettytab tweaking, is there any way I can get the telnet server to close the connection at the login: prompt using Control-D (like most Unixes)? For some reason, it's using Control-Z now. -- Mike Andrews "This guy's pretty bizarre, Gus." root@fragile.termfrost.org [NeXTmail OK] - Primus kramer@wittenberg.edu (school) kramer@mik.uky.edu (hometown)
Newsgroups: comp.sys.next.bugs From: sngmedia@world.std.com Subject: MARKET SURVEY Message-ID: <CoJICL.Mx2@world.std.com> Organization: The World Public Access UNIX, Brookline, MA Date: Wed, 20 Apr 1994 04:12:20 GMT We are a small, dedicated group of people desperately trying to get a video production company off the ground. Recently, we secured enough financial backing to produce several videos on the subject of personal computing. Because we don't have the financial resources to perform a traditional market analysis, we considered several alternatives and concluded that we should consult the people of the Internet, who are technically proficient and tend to be enthusiastic personal computer users. We would like you to indicate which three of the below described videos would appeal to you most. Please note that each video synopsis is numbered. When you are ready to send your response, please specify the numbers corresponding to your three video selections, and use commas to separate the numbers. For example, if you choose selections four, eight, and eleven, the body of your letter would simply be: 4,8,11 and nothing more. When you send back your response, please specify "MARKET SURVEY" in the subject field. Then, direct your response back to sngmedia@world.std.com. Please realize that this is not an advertisement, or solicitation. We simply would like to know your response. You will not be contacted further. Thank you in advance for your help and cooperation. Tom and Randi Fecker sngmedia@world.std.com. Please select three videos that interest you most: 1. The Internet Pilot's Guide A plain English guide to getting around on the Internet by veteran users. How to use Internet resources to send/receive e-mail, and to search for documents and files relating to a particular subject. How to use LISTSERVs, GOPHER, ARCHIE, WWW and WAIS. How to use FTP and Telnet to access remote computers for file transfers, remote logins, etc. Also, learning how to observe Internet's "netiquette." And how to access multimedia Mosaic bulletin boards. 2. How to Make Money at Home With Your PC I -- Successful PC-Based Businesses Meet five successful entrepreneurs who started their own home-based businesses using a PC. Includes businesses that provide typing services, mailing list management, horoscopes, bookkeeping, and even a software developer who does not know how to program -- he hires other people to implement his software designs. Includes ideas for two dozen home-based businesses you can start using your PC. 3. How to Make Money at Home With Your PC II: Portfolio Management. Whether you manage your own portfolio for investment gains, or you take on clients for a consulting fee, a PC can give you powerful tools for portfolio management. Meet three successful home-based portfolio managers who use their PCs to maximize profits. They'll inspire you, and share some tricks of the trade that can help you land clients and get started. We'll also show you how to determine which software applications will meet your needs, and which online services will satisfy your requirements. 4. How to Make Money at Home With Your PC III: Telecommuting Learn how to use your PC to telecommute to companies across America that are looking for specialized consultants with skills like bookkeeping, software design and programming, editorial and script writing, technical proofreading, graphic design, copywriting, and more. Includes interviews with three successful telecommuters who use their PCs, faxes, modems and phones to serve distant clients, and information about a telecommuting referral service that could help you find employers. 5. How to Upgrade Your Hard Drive, Step-By-Step From choosing your new hard drive, through ordering, installing and testing it. This plain English video shows you how to do the job correctly, one simple step at a time. Using industry-wide standards, we'll show you what all hard drives have in common. Includes how to determine if your controller will support a particular hard drive. How to remove the old drive, and install, format and test the new drive. Includes a discussion on the differences between the various standards. 6. How to Design Your Own PC You don't have to be an engineer to design and build the PC of your dreams. This plain English video shows you how to figure out the PC design that is best for you, how to specify components, how to make sure they'll work together, and where to buy them. You'll end up with a top quality system that will save you money. 7. How to Build Your Own PC Once you've designed your PC, we'll show you how to build it. The actual process will take you only a few hours. Using an easy-to- understand method, we'll show you how to inspect, install and test components. Includes tips and tricks from computer production experts. The technical skills can be easily mastered by almost anyone, and you probably already own most of the tools you would need. 8. How to Increase Your Computer's Memory This plain English video shows you how to determine whether your computer memory can be increased, and how to do the job correctly, one step at a time. You'll learn about industry-wide standards for memory, how to configure additional RAM and cache, how and where to buy RAM chips, and three ways to eliminate low-quality RAM chips. Covers all phases of the process from opening your computer, to testing your memory. Includes discussions on how to ensure your DOS set-up is able to access all available memory, and how to use various memory management software applications. 9. How to Use MS-Windows 3.1 This powerful graphical user interface can help you work smarter and faster, but the manual and the online tutorial that come with Windows leave many questions unanswered. This plain English, step- by-step video will show you how to install Windows on your computer and set it up to get optimum performance. 10. How to Find a Job in the Information Age A PC can give you an incredible advantage when you're searching for a new job -- or even a new career. But you have to know just how it can help you. In this video, an experienced employment counselor will show you how to tap the power in your PC to find job leads, create a winning resume and customized cover letters, tap into databases and find bulletin boards that will lead you to job openings, and use online services to research potential employers. 11. How to Install a Sound Card in Your Computer Here's how to add incredible stereo sound to your computer with step-by-step help. In plain English, you'll learn how to determine if your computer can support a sound card, how and where to buy a high-quality sound card. How to open your computer, and install and test the sound card. 12. How to Install a CD-ROM Drive in Your Computer Using simple tools, this plain English video shows you how to install a CD-ROM Drive in your computer. You'll learn how to make sure your computer can support a CD-ROM drive -- and what to do if it can't. Covers internal vs. external drives, how and where to buy a high quality CD-ROM drive, what you need to know about differing industry standards, preparing the drive bay, testing and trouble-shooting. Covers SCSI and IDE. 13. How to Fix the Most Common Computer Problems Your computer serviceman may not want you to know this, but all you need is the know-how you'll get from this video, simple tools, and easily-obtainable diagnostic software -- and you can fix most common problems you'll ever encounter with a PC. 14. What to Do When a Virus Strikes Your Computer Viruses can come from almost anywhere: a colleague or friend's disks, a network, a bulletin board, even commercial software. If you ignore the first warning signs, a virus can wipe out your data and permanently damage your computer's memory. In plain English, this video will tell you how to scan disks for viruses, how to check downloaded files from bulletin boards, how to set up a virus prevention program for your home or office computer, and how and where to buy the best anti-virus software. We'll also cover the pros and cons of the antivirus software in DOS 6.X and Windows 3.X, how to use antivirus software, and more. 15. How Your PC Works: Inside the Case Here's a down-to-earth explanation of how your PC actually processes information, and what really goes on inside the case. You'll get a guided tour of the insides of a PC, learn about how the various components work and how they communicate with each other, and get a clear explanation of technical terms. A must for anyone who wants to really understand how to program, use and repair a PC. 16. How to Create Config.Sys, Autoexec.Bat and Batch Files These basic files can make it much easier to use your computer -- or cause incredible headaches if they are not written properly for your particular software and peripherals. Now you don't have to be at the mercy of murky tech manuals, because we'll show you how to create files that work for your system -- step-by-step, in plain English. You'll learn how to write, modify and test Autoexec.Bat and Config.Sys files; and how to create batch files. 17. How to Add a Modem or Faxmodem to Your Computer Here's the easy way to add a modem or faxmodem to your computer, with step-by-step guidance from this plain English video. You'll learn how to determine if your computer can support a modem or faxmodem, and what to do if it can't, how to choose and buy the best modem or faxmodem, how to open your computer, and install the modem or faxmodem, how to test it, how to quickly eliminate common problems, and how to set your modem or faxmodem correctly. 18. How to Make Money at Home With Your Computer The information age is opening up incredible new opportunities for PC owners to make undreamed of money, using skills and knowledge you may already have! Here's inside information on the ten most promising telecommuting jobs and 12 small businesses you can run right from your home, using your PC. Includes profiles of PC owners who are actually running PC-based home businesses. 19. The Super-Highway Roadmap This is your guide to where to go and what to see. You can make incredible contacts and gather powerful, valuable information on the Internet, but the problem is that most people can't begin to imagine the potential of something that seems so abstract. This plain English video will introduce you to the Internet, and make these opportunities concrete. Includes interviews with 7 people who did the impossible by gathering information and making contacts on the Internet. 20. How to Upgrade and Repair Your PC I: Basic This is the video your repairman doesn't want you to know about! Since the components of most PCs are highly modular, PC repair is easier than you think. Just pinpoint the problem component, unplug it, remove a few screws, replace it, and presto! You're in business again. This step-by-step video shows you how to pinpoint problems and replace your PC's components, using ordinary household tools. 21. How to Upgrade and Repair Your PC II: Multimedia Here's how to save big money on a PC with all the latest multimedia peripherals. You learn how to determine if your PC can be upgraded, how to upgrade your video card and bus, and how to add a CD-ROM drive, sound card, video accelerator, and more. Presented in plain English. The procedures you'll learn require ordinary household tools -- nothing fancy! 22. Plain English Guide to DOS 6+. The powerful sub-programs buried deep within DOS 6.0 and higher can help you work smarter and faster, but the manual and the online tutorial that come with DOS leave many questions unanswered. This plain English, step-by-step video will show you how to install DOS on your computer and set it up to get optimum performance. In addition to DOS commands, you'll learn how to use the shell, defragmentation, scan and antivirus programs that come with DOS. 23. Home Financial Management on a PC. Your computer can help you create and manage a budget, keep track of your credit card accounts, handle your tax deductions, and reconcile your bank accounts. But that's not all! You can also determine whether you should pay down your mortgage, finance a new car or pay cash, buy or rent your home, and how much you'll need for retirement. The financial information your computer can give you might mean the difference between just getting by and a very comfortable lifestyle -- if you ask the right questions and use your PC to develop a financial strategy. 24. The Online Bulletin Board Guide Bulletin boards can be the on-ramps to the Information Super Highway -- if you know how to access and use them. This step-by- step guide shows you how to find bulletin boards, set-up your modem, log on, find out what they have to offer, find bulletin board users who share your interests, search for information, and upload and download files. Thank you.
From: lemson@ux1.cso.uiuc.edu (David Lemson) Newsgroups: comp.sys.next.sysadmin,comp.sys.next.bugs Subject: PCNFSD problems (was Re: pop on 110/tcp) Date: 20 Apr 1994 05:01:08 GMT Organization: University of Illinois at Urbana Message-ID: <2p2cuk$aks@vixen.cso.uiuc.edu> References: <smh-170494034702@smh-ppc.netserv.com> <smh-170494140720@smh-ppc.netserv.com> <1994Apr18.012802.930@cyantic.com> mark@cyantic.com (Mark T. Dornfeld) writes: >Could this be the same type of problem that is preventing Sun's PCNFSD from >doing authentication on NS? >I have not been able to get PCNFSD to do authentication. I recently had a problem, I was meaning to make out a BugNeXT on it. I was using rpc.pcnfsd as shipped in 3.2. I have a 3-level netinfo hierarchy and on a client, I was unable to have rpc.pcnfsd authenticate a user that was listed in the local domain passwd map on that local machine. I removed the user from there, put it in the top level domain, and suddenly the rpc.pcnfsd would allow it. Can anyone else confirm this behavior? This is on black hardware. -- David Lemson University of Illinois Computing & Comm Services Office System Administrator Internet : lemson@uiuc.edu UUCP :...!uiucuxc!uiucux1!lemson NeXTMail & MIME accepted BITNET : LEMSON@UIUCVMD
From: robert@amo.mit.edu(Robert Lutwak) Newsgroups: comp.sys.next.bugs Subject: Re: Configuring Serial line to allow 8-N-1 dial-in ? Date: 20 Apr 1994 19:37:55 GMT Organization: Massachvsetts Institvte of Technology Message-ID: <2p40aj$24r@senator-bedfellow.MIT.EDU> References: <CoJ207.5Ls@fragile.termfrost.org> Many thanks to all of those who've offered suggestions to help get 8-N-1 to work on a serial line. I have tried all of the suggestions, and permutations thereof and still have little or no success. One additional tidbit: When you set gettytab for something like p8:zp, the login banner appears properly, and the login name is read into getty properly. However, the getty then fires up something like: login -p username Which asks for a password and then hangs waiting for input until I kill it. Some permutations of zp,p8, and ap cause the "password:" prompt to be garbled as well. Apparently, getty knows how to handle 8-n-1, but once it gives the tty to another process, it fails. I've tried gettytab entries like: :lo=/bin/echo: and they fail as well. Again, thanks for the help so far, but the solution still evades me. Robert -- Robert Lutwak robert@amo.mit.edu MIT Atomic Resonance and Spectroscopy Laboratory --- Everything I say via this computer belongs to MIT. --- ---> NeXTmail always welcome <---
From: croehrig@celegans.psych.ubc.ca (Chris Roehrig) Newsgroups: comp.sys.next.bugs Subject: /usr/etc/disk fails to scan for alternate superblocks on NS/Intel Date: 20 Apr 1994 23:45:23 GMT Organization: Computer Science, University of B.C., Vancouver, B.C., Canada Message-ID: <2p4eqj$r0i@cs.ubc.ca> /usr/etc/disk reports no alternative superblocks on my NEXTSTEP partition under NS/Intel (3.2, SCSI, 512 bytes/sector). I've tried two NSI boxes and the both fail to report any blocks. Workaround: If you get a BAD SUPER BLOCK: TRASHED VALUES IN SUPER BLOCK when you start up, you have to specify an alternate superblock for fsck. The fsck man page suggests block 16, and fsck does seem to proceed with this alternate, but my problem didn't go away. I tried alternate block 32 and that seems to work. (My black NeXT with 1024 bytes/sector reports the first alternate superblock at block 32. The document "Fsck - The UNIX File System Check Program" by M.K. McKusick and T.J. Kowalski, part of the UNIX system manager manuals, also suggests using 32.) Note: when it says to reboot without sync, don't use "reboot". Just reset your computer. -- Chris Roehrig (croehrig@celegans.psych.ubc.ca) Invertebrate Learning Group, University of British Columbia, Canada
From: eps@futon.SFSU.EDU (Eric P. Scott) Newsgroups: comp.sys.next.bugs Subject: Re: Configuring Serial line to allow 8-N-1 dial-in ? Date: 21 Apr 1994 02:56:47 GMT Organization: San Francisco State University Message-ID: <2p4q1f$22p@nic-nac.CSU.net> References: <CoJ207.5Ls@fragile.termfrost.org> <2p40aj$24r@senator-bedfellow.MIT.EDU> In article <2p40aj$24r@senator-bedfellow.MIT.EDU> robert@amo.mit.edu writes: >When you set gettytab for something like p8:zp, the login >banner appears properly, and the login name is read into >getty properly. However, the getty then fires up something >like: >login -p username >Which asks for a password and then hangs waiting for >input until I kill it. Some permutations of zp,p8, and >ap cause the "password:" prompt to be garbled as well. You need zp (which technically supersedes ap), and I doubt p8 is useful except in combination with rw (and it might have unpleasant side effects). >Apparently, getty knows how to handle 8-n-1, but once >it gives the tty to another process, it fails. Did you try my f2#012601200330 suggestion? Did it make any difference? > I've tried >gettytab entries like: >:lo=/bin/echo: >and they fail as well. Either you're mistaken, or there's a new-with-NS3.x bug here. There was a known bug in NS 2.x specific to /bin/login* that I believe to be present as well in 3.2, although I haven't tested it yet. The symptom was a garbled Password prompt (and garbled output from then on until stty pass8out was entered). I never saw a case where things hung, or input failed to work properly. Using ANY other program for lo= besides /bin/login also worked without problems. *My quick-and-dirty fix involved changing /bin/login as follows: #ifndef ORIGINAL ioctlval = 0; (void)ioctl(0, TIOCLSET, &ioctlval); #else ioctlval = LNOFLSH|LPENDIN|LNOHANG|LFLUSHO|LTOSTOP|LMDMBUF; (void)ioctl(0, TIOCLBIC, &ioctlval); ioctlval = 0; #endif Again, I'm not making any claims of completeness, correctness or applicability to NS 3.2 at this time. NeXT has since "overhauled" the tty driver, and it's quite possible something really is seriously broken now. I'll perform my own investigation and expect to follow up within the next few days. *sigh* -=EPS=-
Newsgroups: comp.sys.next.bugs From: thomas@gamelan.sh.sub.de (Thomas Funke) Subject: Re: free () ?? Message-ID: <1994Apr19.093631.1279@gamelan.sh.sub.de> Sender: thomas@gamelan.sh.sub.de (thomas) Organization: Disorganization References: <2on3bi$128@agate.berkeley.edu> Date: Tue, 19 Apr 1994 09:36:31 GMT In article <2on3bi$128@agate.berkeley.edu> william@moomin.berkeley.edu (William E. Grosso) writes: > > > I seem to be getting some weird behaviour with "free". > Basically, I'm mallocing and freeing a fair amount of > memory and occasionally get the error > > Memory access exception on address 0x3f800000 > > on the line > > free(temporary) Probably your code has destroyed some of the data used by the malloc-management (linked list etc.). Don't blame NeXT for everything, actually bugs like this are frequent in C. -- ------------------------------------------------------------------ Thomas Funke thomas@gamelan.sh.sub.de Brooks's Law: Adding manpower to a late software project makes it later ------------------------------------------------------------------
From: kpfleger@Xenon.Stanford.EDU (Karl Robert Pfleger) Newsgroups: comp.sys.next.hardware,comp.sys.next.bugs,comp.sys.next.misc Subject: kernel panic exception (6,3,1), NS/I on GW2K P5-66 Date: 21 Apr 1994 21:00:40 GMT Organization: Computer Science Department, Stanford University. Message-ID: <2p6pho$6ac@Times.Stanford.EDU> I finally got my system completely assembled, configured, partitioned, and loaded with NEXTSTEP 3.2 for Intel. Yeah! Mostly it seems to work great, except... Very occasionally (once every 5-7 boots roughly), I get a kernel panic which crashes the boot process and I have to reboot, which usually works. When it does boot correctly, most everything seems to work fine. Does anyone have any clue what's going on? More importantly, is this really bad, or just a minor booting hassle? I've got a matter of days to decide whether to ship the Gatewauy back before my 30 day money-back period is up. Should I be worried? The exact text I see when booting wih diagnostic messages is the following: ----- <after registering various devices> rootdev 600, howto 0 panic: (Cpu 0) vfs_mount root: cannot mount root panic: NeXT Mach 3.2: Mon Oct 18 22:08:07 PDT 1993; root (rcbuilder): mk-149.30.15.obj~2/RC_i386/RELEASE_I386 kernel panic exception (6,3,1) Waiting for remote debugger connection. (Type 'c' to continue or 'r' to reboot) ----- My exact configuration: Gateway 2000 P5-66 Best Buy, Pentium 66MHz, Intel 2 PCI / 3 ISA / 1 interleaved motherboard, 32MB RAM Seagate ST11200N 1GB SCSI HD Adaptec 1542CF ISA SCSI controller NEC 3xi internal triple-speed SCSI CD-ROM (model CDR-510) Miro Crystal 32S 4MB PCI video card Intel EtherExpress16 ISA network card MS Windows Sound System ISA sound card running NEXTSTEP 3.2 for Intel. SCSI setup: physical layout: Adaptec 1542CF connected to HD by one SCSI cable with a SCSI connector in the middle of the cable connected to the CD-ROM drive. There's about 11 1/2 - 11 3/4 " of cable between the CD-ROM and the HD, and about twice that between the CD-ROM and the controller. logical settings: SCSI ID terminated Adaptec 1542CF 7 yes (IRQ: 11, DMA channel: 6) Seagate ST11200N 0 yes NEC 3xi 1 no All the SCSI setup settings on the Adaptec card are identical to the NextAnswers document on it (memory address, etc.). I can't seem to find any pattern of when the panic happens and when it doesn't. It has happenend and not happened on soft reboots, hardware reset button reboots, and power-up boots in seemingly all possible contexts. I had decided to gamble and see if NS/I worked on a computer significantly cheaper than any of the equivalent officially supported systems, and it seems to run great on this Gateway. (A friend of mine who owns a great NeXTStation color turbo says my system feels great to him.) I had hoped that buying only completely known compatible and robust peripherals would insure that compatibility was a very all-or-nothing issue: if I could get it to install correctly, it would work, otherwise it wouldn't work at all. (I figured it would mostly depend on the only unsupported factor--the motherboard.) I think it is very strange that booting from power-up gives non-deterministic behavior. Any help much appreciated. -Karl P.S. I intend to post a long summary of configuration info, general information about buying an Intel computer to run NEXTSTEP, advice, impressions, thanx, etc. as soon as I'm sure I'm keeping the system. Until then, I can provide specific information on request. I'd like this to become a "listed system" in NextAnswers because I think it is a great system.
From: cgeyer@aol.com (Cgeyer) Newsgroups: comp.sys.next.bugs Subject: Re: Help:: Adaptec 1542 with new bios. Date: 21 Apr 1994 15:37:02 -0400 Organization: America Online, Inc. (1-800-827-6364) Sender: news@search01.news.aol.com Message-ID: <2p6kku$4kn@search01.news.aol.com> The Adaptec 1542CF bios is buggy according to NeXTAnswers 1542 (I think, not positive). I have a 1542CF card and had the same problem. I called up NeXT technical support and asked for their updated driver. The instructions for installing it are in that NeXTAnswer that I mentioned. I did not have to pay the $30/15min charge for requesting the driver. They send it first class and I should be getting it soon (I think I called them up Monday). For now I am stuck with a gig hard drive wasting 400mb of space. Christopher Geyer
From: cgeyer@aol.com (Cgeyer) Newsgroups: comp.sys.next.bugs Subject: CD-ROM and SoftPC Date: 21 Apr 1994 15:46:02 -0400 Organization: America Online, Inc. (1-800-827-6364) Sender: news@search01.news.aol.com Message-ID: <2p6l5q$4nl@search01.news.aol.com> Does anyone have any problems accessing a CD-ROM disk from SoftPC? I start up SoftPC and I put in an ISO9660 CD-ROM that is supposed to be used under DOS. I select 'Attach CD-ROM' and then type in 'USE_CD' and try to access F:. It then says that it is not an ISO or Sierra disk or something similar, abort, retry, etc. Christopher Geyer
From: kpfleger@Xenon.Stanford.EDU (Karl Robert Pfleger) Newsgroups: comp.sys.next.hardware,comp.sys.next.bugs,comp.sys.next.software,comp.soft-sys.nextstep,comp.sys.next.misc,su.computers.next Subject: robustness testing before performance tuning Date: 21 Apr 1994 23:12:03 GMT Organization: Computer Science Department, Stanford University. Distribution: inet Message-ID: <2p7183$903@Times.Stanford.EDU> An important question and some (I think) useful ideas and philosophy follow. And at the end I issue two challenges... Basic question: Is there any software, or are there any things to check by hand, to test out a new NEXTSTEP system when things appear to work correctly in order to determine with high confidence that everything is in fact working properly and there aren't any hidden incompatibility problems? Background: With all the complaints about incompatibility in the NS/I world, I posted a while ago a question about whether NS/I compatibility was a black and white issue or whether there were shades of gray. That is, if you go out and buy a non-supported system, does successful NS installation and brief use of all system components indicate that there is a good likelihood that everything works fine and is completely compatible? Specifically, it is easy to determine the supported components for NS/I and stick to those. The real potential for problems is with the motherboard. How likely is it that a motherboard can be 80% compatible, or 90% compatible, in ways that aren't ilkely to be noticed in the first 30 days of ownership? It certainly seems to be the case that certain peripherals (e.g., ProAudioStudio16 ISA sound card) seem to cause intermittant problems (freeze up and crash the OS). [Philosophizing & soap-box mode on] In general, it seems when people buy a new sophisticated computer system (NS or other), the first thing they do once it is set up (in addition to playing games or gawking at multimedia type stuff) is to start doing performance benchmarking. (At least serious computer junkies seem to.) This makes some sense since presumably they always could have bought cheaper systems if they were willing to suffer from worse performance of some kind. So benchmarking is a way to check that they got their money's worth. In addition, after an initial set of performance evaluations (and after bragging about these on the net and to their computer-junky-inclined friends :-) ) they can try to improve their systmes scores by fiddling with various things. This has some utility since presumably it will improve the quality of their system. But still, it seems to me they have missed something, namely robustness testing. Things often go wrong with computer systems and they should try to make whatever is going to go wrong actually do so as soon as possible. This is admittedly, a significantly less important issue in worlds where there are few or no compatibility issues, since it applies mostly to compatibility problems (since actual hardwar failures are tested for more extensively before shipping and also are covered under warranty for a reasonably long period of time). But in the NS/Intel world this is especially important. Furthermore, I would claim that compatibility problmes will be increasingly important in the future as more and more hardware is capable of running more and more different software and operating systems. Thus, this will be an issue for more and more people. Having gotten no responses to my black & white vs. shades of gray question, I went ahead and bought a Gateway Pentium to run NS/I, even though it was not a certified or listed NextAnswers system, having heard over the net of one person successfully installing NS on the Intel motherboard used in these systems. I installed NS and everything seems to work, for the most part, great (and fast too! :-) ). So what should I do now, in the roughly 1 week I have left before my 30 days are up to put my new system through its paces? Challenge (1): Send me the hardest tasks you think a NEXTSTEP system has to face that you think are most likely to cause problems to show up if they exist. Challenge (2): I think someone who has access to a lot of different PC hardware should design a piece of software and/or suite of tests which would with high confidence (I'd be impressed by 75%) and in only a few hours expose problems which still exist after NEXTSTEP installation. Or do you actually believe that both challenges are a waste of time because you have confidence that compatibility issues are black or white. Care to back up the opinion? -Karl
From: hbott@esu.edu (Howard B Ott) Newsgroups: comp.sys.next.hardware,comp.sys.next.bugs,comp.sys.next.misc Subject: Re: kernel panic exception (6,3,1), NS/I on GW2K P5-66 Followup-To: comp.sys.next.hardware,comp.sys.next.bugs,comp.sys.next.misc Date: 22 Apr 1994 01:10:57 GMT Organization: East Stroudsburg University, Pennsylvania Message-ID: <2p7871$j7i@jake.esu.edu> References: <2p6pho$6ac@Times.Stanford.EDU> Karl Robert Pfleger (kpfleger@Xenon.Stanford.EDU) wrote: : I finally got my system completely assembled, configured, partitioned, : and loaded with NEXTSTEP 3.2 for Intel. Yeah! Mostly it seems to work : great, except... : Very occasionally (once every 5-7 boots roughly), I get a kernel panic : which crashes the boot process and I have to reboot, which usually : works. When it does boot correctly, most everything seems to work fine. : ----- : <after registering various devices> : rootdev 600, howto 0 : panic: (Cpu 0) vfs_mount root: cannot mount root : panic: NeXT Mach 3.2: Mon Oct 18 22:08:07 PDT 1993; root (rcbuilder): mk-149.30.15.obj~2/RC_i386/RELEASE_I386 : kernel panic exception (6,3,1) : Waiting for remote debugger connection. : (Type 'c' to continue or 'r' to reboot) : ----- I have the same problem, my config is DELL XPS 66Mhz 486 ATI ULTRA pro VLB Adaptec 1542C ISA SCSI controller Intel EtherExpress16 ISA network card NEC CDR74 Fujitsu 520 MB No sound card running NEXTSTEP 3.2 for Intel. Berk Ott
Newsgroups: comp.sys.next.bugs From: thompson@netcom.com (Eric Thompson) Subject: Dig. Librarian can't search on regular expressions? Message-ID: <thompsonCoMwB7.8vL@netcom.com> Organization: Saperstein, et al. Date: Fri, 22 Apr 1994 00:06:42 GMT Pardon me of this is old news or if I'm doing something wrong.. but every time I choose "Expression" (as opposed to "Word") searching in DL, the damn program crashes. No matter what I enter. This is under NS/I 3.2. thanks for any help, Eric
From: silbar@mpx2.lampf.lanl.gov (Dick Silbar) Newsgroups: comp.sys.next.hardware,comp.sys.next.bugs,comp.sys.next.misc Subject: Re: kernel panic exception (6,3,1), NS/I on GW2K P5-66 Date: 21 Apr 1994 21:37 MST Organization: LAMPF Data Analysis Center, Los Alamos, New Mexico Distribution: world Message-ID: <21APR199421370648@mpx2.lampf.lanl.gov> References: <2p6pho$6ac@Times.Stanford.EDU> News-Software: VAX/VMS VNEWS 1.41 In article <2p6pho$6ac@Times.Stanford.EDU>, kpfleger@Xenon.Stanford.EDU (Karl Robert Pfleger) writes... >I finally got my system completely assembled, configured, partitioned, Go get the newer Adaptec drivers from ftp.next.com.
From: af@iaka.biomath.jussieu.fr (Alain Fauconnet) Newsgroups: comp.sys.next.sysadmin,comp.sys.next.bugs Subject: Netinfo clients don't bind to clone server Date: 22 Apr 94 11:31:58 GMT Organization: Universites Paris VI/Paris VII - France Distribution: world Message-ID: <af.767014318@iaka> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Hello, I have recently tried to make our NeXTs more reliable by adding a clone server to our (simple, two-level) Netinfo network of Nextstations. I carefully followed the instructions in the "System and Network Administration" book, and it doesn't seem to work at all. When the usual server crashes, all the NeXTstations more or less hang, and even rebooting them doesn't seem to make them re-bind to the clone server. They complain at boot about not finding a server. The clone server does run a "netinfod network" daemon, I have checked that. I have read the pages 250-251 of the doc explaining the binding process and I'm stuck. Can anyone help ? Master and clone servers running 3.0 (3.1 is just not reliable enough and I don't have 3.2), clients are a mix of 3.0 and 3.1. The two servers (beryl and vercoquin) seem correctly configured: > niutil -read / /machines/beryl bootparams: ip_address: 134.157.72.76 serves: beryl/local ./network <------------- system_type: NeXTstation Color name: beryl netgroups: trusted > niutil -read / /machines/vercoquin bootfile: mach bootparams: ip_address: 134.157.72.84 en_address: 0:0:f:1:4f:d6 serves: vercoquin/local ./network <------------- system_type: NeXTstation Color name: vercoquin netgroups: trusted The manual states that "the ip_address property for broadcasthost is 255.255.255.255, so the bind request is broadcast to all the hosts on the local network". I'm not a TCP/IP expert, but it seems to me that on a local net using subnetting (netmask=255.255.255.128, broadcast=x.y.z.127) like ours, a packet to 255.255.255.255 will go to the default router and not be sent as an broadcast Ethernet frame, will it ? Anyway the binding to the master server in normal conditions does occur, so I guess that's a bad track. Thanks for any hints, _Alain_ -- Alain FAUCONNET Ingenieur systeme - System Manager AP-HP/SIM Public Health Research Labs 91 bld de l'Hopital 75013 PARIS FRANCE Mail: af@biomath.jussieu.fr (*no* NeXTmail !) Tel: (+33) 1-40-77-96-19 Fax: (+33) 1-45-86-80-68
From: jweiss@casbah.acns.nwu.edu (Jerry Weiss) Newsgroups: comp.sys.next.hardware,comp.sys.next.bugs,comp.sys.next.software,comp.soft-sys.nextstep,comp.sys.next.misc,su.computers.next Subject: Re: robustness testing before performance tuning Date: 22 Apr 1994 12:54:48 GMT Organization: Northwestern University, Evanston IL USA Distribution: inet Message-ID: <2p8heo$l74@news.acns.nwu.edu> References: <2p7183$903@Times.Stanford.EDU> In article <2p7183$903@Times.Stanford.EDU>, Karl Robert Pfleger <kpfleger@Xenon.Stanford.EDU> wrote: >An important question and some (I think) useful ideas and philosophy follow. >And at the end I issue two challenges... > >Basic question: Is there any software, or are there any things to check >by hand, to test out a new NEXTSTEP system when things appear to work >correctly in order to determine with high confidence that everything is in >fact working properly and there aren't any hidden incompatibility problems? > > Background deleted > >Challenge (1): Send me the hardest tasks you think a NEXTSTEP system has >to face that you think are most likely to cause problems to show up if >they exist. > >Challenge (2): I think someone who has access to a lot of different >PC hardware should design a piece of software and/or suite of tests >which would with high confidence (I'd be impressed by 75%) and in only a >few hours expose problems which still exist after NEXTSTEP installation. > >Or do you actually believe that both challenges are a waste of time >because you have confidence that compatibility issues are black or white. >Care to back up the opinion? > I'd have to say that the hardest task is simply getting NEXTSTEP itself to run. There's no application in general that is going to exercise the basic hardware of the computer, memory, video interfaces, disk I/O, interrupts and dma more than NS itself. Then of course you have specific interfaces and/or hardware that NS itself won't directly exercise in the course of everyday operation, but these tend to rather specific to the mission the user has intended (ie:peripherals and applications related to them). The difficulty you encounter at this point is whether you are testing the hardware or NS itself with any particular package. NS requires/demands good performance of its components. Hardware flaws or firmware bugs in cards or peripherals or masked by BIOS code that escaped notice in lesser Operatings Systems (msdos, windoze,...) can interfer with NS operation. Bugs within NS (the few that exist ;-) can make it difficult to debug certain types of problems. I suspect we could probably come up with a general list of Apps to run and some additional test code to use as a system exerciser, but you are not going to find those problems that make you want to tear your hair out until you can't run the application YOU need. -- Jerry S. Weiss j-weiss@nwu.edu Dept. Medicine, Northwestern Univ. Medical School, Chicago, Illinois %SYSTEM-S-PHALOKTARG, Phasers Locked on Target, Ready to Fire
Newsgroups: comp.sys.next.sysadmin,comp.sys.next.bugs From: richardm@cd.com (Richard Masoner) Subject: Re: Configuring Serial line to allow 8-N-1 dial-in ? Message-ID: <Coo1K1.2F1@cd.com> Organization: Central Data Corp., Champaign, IL References: <2osffr$b5o@senator-bedfellow.MIT.EDU> <1994Apr19.170944.1687@clark.dgim.doc.ca> Date: Fri, 22 Apr 1994 14:57:36 GMT In article <1994Apr19.170944.1687@clark.dgim.doc.ca> cbarton@DaBoyz.dgcp.doc.ca (Casey Barton) writes: > I've tried several different parity settings, none of which will get >NeXT to talk to the modem in anything other than 7E1. (7O1 doesn't work >either.) I'm starting to think this is a bug -- but I'd think this would >have been a very visible one -- 8N1 dialin seems to be the most common, This has been my experience too, and in fact NeXT's tty layer doesn't seem to have any flags to support changing the number of data bits. When using the POSIX interface, some of the flags are just completely ignored at the driver level! When writing a driver for Central Data's tty hardware, I had to apply some creative interpretation on some of the meanings of the flags to get these things to work. -- Richard F. Masoner ------------------------------------------------ "A soft answer turns away wrath." -Proverbs 15:1
Newsgroups: comp.sys.next.bugs From: paul@seer.demon.co.uk (Paul Lynch) Subject: Re: /usr/etc/disk fails to scan for alternate superblocks on NS/Intel Message-ID: <1994Apr22.054335.11819@seer.demon.co.uk> Sender: news@seer.demon.co.uk Organization: P & L Systems References: <2p4eqj$r0i@cs.ubc.ca> Date: Fri, 22 Apr 1994 05:43:35 GMT In article <2p4eqj$r0i@cs.ubc.ca> croehrig@celegans.psych.ubc.ca (Chris Roehrig) writes: > Note: when it says to reboot without sync, don't use "reboot". Just > reset your computer. 'reboot -n' is a better option. Paul -- Paul Lynch P & L Systems (NeXTmail) paul@seer.demon.co.uk Tel: (0494)671501 9 Stable Lane, Seer Green, Fax: (0494)680228 Bucks, HP9 2YT, UK
From: robert@steffi.demon.co.uk (Robert Nicholson) Newsgroups: comp.sys.next.bugs Subject: Re: Dig. Librarian can't search on regular expressions? Date: 22 Apr 1994 17:54:09 +0100 Organization: me organised, that's a joke. Message-ID: <2p8vfh$6ou@steffi.demon.co.uk> References: <thompsonCoMwB7.8vL@netcom.com> thompson@netcom.com (Eric Thompson) wrote in comp.sys.next.bugs >Pardon me of this is old news or if I'm doing something wrong.. but every >time I choose "Expression" (as opposed to "Word") searching in DL, the damn >program crashes. No matter what I enter. This is under NS/I 3.2. > >thanks for any help, >Eric Well I just did this and it didn't crash... It didn't find anything either... What does Librarian see as an "Expression"? -- "Kill files are for pacifists" (ASCII for text only messages)
Control: cancel <CoJICL.Mx2@world.std.com> Newsgroups: comp.sys.next.bugs From: sngmedia@world.std.com Subject: cmsg cancel <CoJICL.Mx2@world.std.com> Organization: The World Public Access UNIX, Brookline, MA Date: Wed, 20 Apr 1994 04:12:20 GMT Message-ID: <cancel.CoJICL.Mx2@world.std.com> MARKET SURVEY
From: sulkanen@pecos.msfc.nasa.gov (Martin E. Sulkanen) Newsgroups: comp.sys.next.bugs Subject: Re: Aaargh, can't eject disk Date: 22 Apr 1994 22:22:32 GMT Organization: NASA/MSFC Distribution: world Message-ID: <2p9in8$ibb@avdms8.msfc.nasa.gov> References: <CoHoHv.E08@cunews.carleton.ca> In article <CoHoHv.E08@cunews.carleton.ca> csaldanh@mae.carleton.ca (Chris Saldanha) writes: > Wim Nijenhuis (nijenhui@sun005.research.ptt.nl) wrote: > : I inserted an ordinairy dos format > : disk but there was no disk icon in the filemanager. Also the eject option is > : disabled. I'm using NeXTStep 3.0. Is this a known bug? What's should I do? > : The console mentioned something about: "Mounted floppy disk at /3F f91{@ F" > > : Is there another option than using the screwdriver :-) > > I hope you'd use a pin, and not take the case off... ;-) > > When the console says something like that, go up to the root level of the > WorkSpace fileviewer, and look for a directory with the name given on the > console (i.e. "3F f91{@ F" in this case). Click on that directory, and > voila! the contents of your floppy. Also when you click on that directory, > the Eject menu option will become available again, allowing you to eject > the disk. Well, I've tried that and it the eject menu option is still grayed out... -- /================================================================\ | Martin Sulkanen + internet: sulkanen@ssl.msfc.nasa.gov | | ES 65 + Information & opinions offered here are | | NASA/MSFC + mine and not that of NASA | \================================================================/
Newsgroups: comp.sys.next.bugs From: csaldanh@mae.carleton.ca (Chris Saldanha) Subject: Re: Dig. Librarian can't search on regular expressions? Message-ID: <Coq05D.DK4@cunews.carleton.ca> Sender: news@cunews.carleton.ca (News Administrator) Organization: Carleton University References: <thompsonCoMwB7.8vL@netcom.com> <2p8vfh$6ou@steffi.demon.co.uk> Date: Sat, 23 Apr 1994 16:22:25 GMT Robert Nicholson (robert@steffi.demon.co.uk) wrote: : thompson@netcom.com (Eric Thompson) wrote in comp.sys.next.bugs : >Pardon me of this is old news or if I'm doing something wrong.. but every : >time I choose "Expression" (as opposed to "Word") searching in DL, the damn : >program crashes. No matter what I enter. This is under NS/I 3.2. : Well I just did this and it didn't crash... : It didn't find anything either... : What does Librarian see as an "Expression"? On page 177 of the user's guide it states: EXPRESSION: To search for UNIX regular expressions, which are described in the UNIX manual page for ed. You know, this is thrice I've found something useful in the User's Guide. What a concept! And I thought I knew everything! --Chris Chris Saldanha -------------------------------------- Carleton University |"The eternal silence of these infinite| chris@computerActive.on.ca (NeXTMail) | spaces terrifies me." -Blaise Pascal| csaldanh@mae.carleton.ca (NeXT/MIME) ------------------:-o-----------------
Newsgroups: comp.sys.next.bugs From: thompson@netcom.com (Eric Thompson) Subject: Re: Dig. Librarian can't search on regular expressions? Message-ID: <thompsonCoqCJr.LGB@netcom.com> Organization: Saperstein, et al. References: <thompsonCoMwB7.8vL@netcom.com> <2p8vfh$6ou@steffi.demon.co.uk> <Coq05D.DK4@cunews.carleton.ca> Date: Sat, 23 Apr 1994 20:50:14 GMT I tried regular expressions, irregular expressions, and even some foul, gutteral expressions :-) DL simply quits, _every single time_. I _just_ tried searching the MAN PAGES on my turbo color, and after searching for about 30 seconds for expression [Cc][Aa][Tt], IT QUIT TOO. What can I say.. I'm less than favorably impressed. Eric thompson@netcom.com Chris Saldanha <csaldanh@mae.carleton.ca> wrote: >On page 177 of the user's guide it states: >EXPRESSION: To search for UNIX regular expressions, which are described >in the UNIX manual page for ed. > >You know, this is thrice I've found something useful in the User's Guide. >What a concept! And I thought I knew everything!
From: kpfleger@Xenon.Stanford.EDU (Karl Robert Pfleger) Newsgroups: comp.sys.next.hardware,comp.sys.next.bugs,comp.sys.next.misc Subject: Re: kernel panic exception (6,3,1), NS/I on GW2K P5-66 Date: 24 Apr 1994 01:10:55 GMT Organization: Computer Science Department, Stanford University. Message-ID: <2pcguv$5eb@Times.Stanford.EDU> References: <2p6pho$6ac@Times.Stanford.EDU> <2p7871$j7i@jake.esu.edu> In article <2p7871$j7i@jake.esu.edu>, Howard B Ott <hbott@esu.edu> wrote: >Karl Robert Pfleger (kpfleger@Xenon.Stanford.EDU) wrote: >: Very occasionally (once every 5-7 boots roughly), I get a kernel panic >: which crashes the boot process and I have to reboot, which usually >: works. When it does boot correctly, most everything seems to work fine. >: kernel panic exception (6,3,1) >: Waiting for remote debugger connection. >: (Type 'c' to continue or 'r' to reboot) >I have the same problem, my config is > DELL XPS 66Mhz 486 > ATI ULTRA pro VLB > Adaptec 1542C ISA SCSI controller > Intel EtherExpress16 ISA network card > NEC CDR74 > Fujitsu 520 MB Consensus seems to be that it is the Adaptec card. One person said that switching to the DPT ISA controller made the problem go away. A number of people said they get the problem on EVERY boot. A couple people said to try upgrading to the new release of the Adaptec driver, available from NeXTAnswers. I had thought that I had installed that, but I'm not really sure, so I'll install it again to make sure. Good luck everyone. (Further data on the problem: every time I change the configuration a lot, like play with fdisk to change around the non-NEXTSTEP partitions, the next boot panics, but the one after that is okay.) -Karl
From: pas@phantom.com (Paul Shetler) Newsgroups: comp.sys.next.hardware,comp.sys.next.bugs,comp.sys.next.misc Subject: Re: kernel panic exception (6,3,1), NS/I on GW2K P5-66 Date: 23 Apr 1994 23:19:50 GMT Organization: [MindVox] / Phantom Access Technologies / (+1 800-MindVox) Message-ID: <2pcaem$8nd@dockmaster.phantom.com> References: <2p6pho$6ac@Times.Stanford.EDU> <2p7871$j7i@jake.esu.edu> Howard B Ott (hbott@esu.edu) wrote: : Karl Robert Pfleger (kpfleger@Xenon.Stanford.EDU) wrote: : : I finally got my system completely assembled, configured, partitioned, : : and loaded with NEXTSTEP 3.2 for Intel. Yeah! Mostly it seems to work : : great, except... : : Very occasionally (once every 5-7 boots roughly), I get a kernel panic : : which crashes the boot process and I have to reboot, which usually : : works. When it does boot correctly, most everything seems to work fine. : : ----- : : <after registering various devices> : : rootdev 600, howto 0 : : panic: (Cpu 0) vfs_mount root: cannot mount root : : panic: NeXT Mach 3.2: Mon Oct 18 22:08:07 PDT 1993; root (rcbuilder): mk-149.30.15.obj~2/RC_i386/RELEASE_I386 : : kernel panic exception (6,3,1) : : Waiting for remote debugger connection. : : (Type 'c' to continue or 'r' to reboot) : : ----- : I have the same problem, my config is : DELL XPS 66Mhz 486 : ATI ULTRA pro VLB : Adaptec 1542C ISA SCSI controller : Intel EtherExpress16 ISA network card : NEC CDR74 : Fujitsu 520 MB : No sound card : running NEXTSTEP 3.2 for Intel. I also get this error, config is: Compaq Deskpro XE 560 16Mb RAM, localbus Qvision Fujitsu 1.2 GB drive Adaptec 1542c PAS Basic Chinon CDROM NS/FIP 3.2 pas
Newsgroups: comp.sys.next.bugs From: antoine@arrakis.osd.ulaval.ca (Antoine Gautier) Subject: Re: Aaargh, can't eject disk Message-ID: <Cos284.C2x@athena.ulaval.ca> Sender: news@athena.ulaval.ca Organization: Universite Laval References: <2p9in8$ibb@avdms8.msfc.nasa.gov> Date: Sun, 24 Apr 1994 19:02:27 GMT > > Wim Nijenhuis (nijenhui@sun005.research.ptt.nl) wrote: > > : I inserted an ordinairy dos format > > : disk but there was no disk icon in the filemanager. Also the eject option is > > : disabled. I'm using NeXTStep 3.0. Is this a known bug? Yes. you need diskejectfix, a daemon available from NeXT. It is probably on their ftp site (ftp.NeXT.com). --- +-------------------------------------------------------+ | Antoine Gautier | | Departement OSD, FSA | | Universite Laval, Quebec, PQ | | Internet: Antoine@arrakis.osd.ulaval.ca | +-------------------------------------------------------+
Newsgroups: comp.sys.next.hardware,comp.sys.next.bugs From: yong@iitmax.iit.edu (Yong Yoo) Subject: About Adaptec 1542B driver Message-ID: <1994Apr25.012721.11944@iitmax.iit.edu> Organization: Illinois Institute of Technology, Chicago Date: Mon, 25 Apr 94 01:27:21 GMT I've read someone ordering a 1542cf driver for NS3.2 that will recognize drive bigger than a gig from NeXT. I've been waiting for very long time and I bet I am not the only one who is either. My question is... DID NeXT released a driver that will recognize drive that is bigger than a gig for Adaptec 1542B,C,CF (whatever) or just CF? If this is the driver that will recognize drive bigger than a gig, do I need a New bios from Adaptec to support the new driver from NeXT? ( I have 1542B and I've ordered the bios from adaptec to support the drive that is bigger than a gig.) Another question is... Are other SCSI controllers have same type of problem? I am specially interested on DPT card, since it allowes to have cache. thanks in advance Yong Yoo
From: robert@steffi.demon.co.uk (Robert Nicholson) Newsgroups: comp.sys.next.bugs,comp.unix.programmer Subject: stripping links? Date: 25 Apr 1994 13:40:30 +0100 Organization: me organised, that's a joke. Message-ID: <2pgdnu$q9@steffi.demon.co.uk> This is probably a unix related issue... Why does strip not use lstat to determine if it's stripping via a link? Today I stripped a binary, however I stripped it through links. "strip" turned the links into stripped binaries. So I had two copies of the stripped binary instead of a link pointing to a stripped binary. Is this the desired behaviour? -- "Kill files are for pacifists" (ASCII for text only messages)
From: statman@stat.ufl.edu (Charles D. Kincaid) Newsgroups: comp.sys.next.bugs Subject: Re: Dig. Librarian can't search on regular expressions? Date: 25 Apr 1994 13:58:12 GMT Organization: University of Florida Distribution: world Message-ID: <2pgi9kINNq8c@no-names.nerdc.ufl.edu> References: <thompsonCoqCJr.LGB@netcom.com> In article <thompsonCoqCJr.LGB@netcom.com> thompson@netcom.com (Eric Thompson) writes: > I tried regular expressions, irregular expressions, and even some foul, > gutteral expressions :-) DL simply quits, _every single time_. I _just_ > tried searching the MAN PAGES on my turbo color, and after searching for > about 30 seconds for expression [Cc][Aa][Tt], IT QUIT TOO. > > What can I say.. I'm less than favorably impressed. > Here's what it's doing to me. I tried the expression [Nn][Ee][Xx][Tt] on one of my cleaned up NG newsfolders. It seemed not to use the index that's there, but went through all of the files. Then it returned nothing. I know the expression (that's a correct expression, right?) is there. When I did a search on the word NeXT it had no problem finding all of them. I'm using NS3.2 on a non-turbo station. It did quit on me a couple of times, but that was because the wn2rtf.service was messed up somehow. Even after I fixed the pathnames in wn2rtf.service it still quit with this error >Filter error ID:108 - Error during conversion. in my console. After I removed the service completely it didn't crash. I hope that some of this helps. -- Sincerely, charles d. kincaid -------------------------------------------------------------------- Dept. of Statistics 'Damn fine coffee...and hot, too!' Univ. of Florida Pres: G-ville NeXT Users Group
From: sanguish@digifix.com Newsgroups: comp.sys.next.announce,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.bugs,comp.sys.next.programmer,comp.sys.next.marketplace Subject: Quick Guide to NEXTSTEP information on the Internet Date: 25 Apr 1994 14:02:38 -0400 Organization: Next Announcements Message-ID: <2ph0ju$e0j@digifix.digifix.com> This post is made weekly, to help 'point' users to more NEXTSTEP information Topics include: NEXTSTEP World Wide Web Product Information Server online NEXTSTEP Product Information Mail Server online comp.sys.next newsgroups related newsgroups comp.sys.next newsgroups mailing list ftp sites NeXTanswers Much of this information is also available using the World Wide Web, <http://digifix.digifix.com/index.html> NEXTSTEP World Wide Web Product Information Server online --------------------------------------------------------- A product directory built around the World Wide Web system, this will allow full multimedia announcements by NEXTSTEP developers, as well as the ability to browse the available products for NEXTSTEP. This service is online now, and can be reached at http://digifix.digifix.com/ it can be reached using OmniWeb (available from ftp.omnigroup.com) or Mosaic. The entries are coming in quite quickly, and I'll be moving the entire NEXTSTEP Third Party Catalog contents in as soon as they become available. NEXTSTEP Product Information Mail Server online --------------------------------------------------------- The NEXTSTEP Product Information Mail Server is now available for product literature and pricing from NEXTSTEP developers. You can get information on using the mail server at ns-products@digifix.com Suggestions or comments can be directed to me at sanguish@digifix.com comp.sys.next.* newsgroups -------------------------- Comp.Sys.Next.Advocacy This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. Comp.Sys.Next.Announce Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) The NEXTSTEP FAQs are posted here monthly as well. This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Comp.Sys.Next.Bugs A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT- specific groups as well. Comp.Sys.Next.Hardware Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. Comp.Sys.Next.Marketplace NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. Comp.Sys.Next.Misc For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! Comp.Sys.Next.Programmer Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. The NEXTSTEP programmer FAQs are posted here. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. Comp.Sys.Next.Software This is a place to talk about [third party] software products that run on NEXTSTEP systems. Comp.Sys.Next.Sysadmin Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. related Newsgroups ------------------ Comp.Soft-Sys.Nextstep Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. Comp.Lang.Objective-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. Comp.Object Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original Comp.Sys.Next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News ------------------------------------------- Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep-d next-advocacy-d next-announce-d next-bugs-d next-hardware-d next-marketplace-d next-misc-d next-programmer-d next-software-d next-sysadmin-d (For a full description, send mail saying LISTS to <digestif@antigone.com>). The subscription syntax is essentially the same as LISTSERV's. To subscribe, send a message to <digestif@antigone.com> saying: SUB Listname YourName Example: SUB next-hardware-d John Doe The ftp sites ------------- cs.orst.edu: The main site for North American submissions nova.cc.purdue.edu: Lots of older stuff, but very short on disk space ftp.informatik.uni-muenchen.de: In Germany. terra.stack.urc.tue.nl (Dutch NEXTSTEP User Group) and cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp.next.com: See the below ftp.next.com and NextAnswers@next.com ------------------------------------- From the document 1000_Help from ftp.next.com Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, or you can transfer them by anonymous ftp. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to NeXTanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO <your-address> If you have any problem with the system or suggestions for improvement, please send mail to NeXTanswers-request@NeXT.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to NeXTanswers-request@NeXT.com. Written by: Eric P. Scott eps@toaster.SFSU.EDU and Scott Anguish sanguish@digifix.com Additions from: Greg Anderson (Greg_Anderson@afs.com) and Michael Pizolato (Michael_Pizolato@afs.com)
From: straffin@Xenon.Stanford.EDU (Ethan Douglas Straffin) Newsgroups: comp.sys.next.hardware,comp.sys.next.bugs Subject: Re: About Adaptec 1542B driver Date: 25 Apr 1994 23:16:59 GMT Organization: Computer Science Department, Stanford University. Message-ID: <2phj1b$pii@Times.Stanford.EDU> References: <1994Apr25.012721.11944@iitmax.iit.edu> In article <1994Apr25.012721.11944@iitmax.iit.edu>, Yong Yoo <yong@iitmax.iit.edu> wrote: >I've read someone ordering a 1542cf driver for NS3.2 that will recognize >drive bigger than a gig from NeXT. I've been waiting for very >long time and I bet I am not the only one who is either. My question >is... DID NeXT released a driver that will recognize drive that is bigger >than a gig for Adaptec 1542B,C,CF (whatever) or just CF? If this is the >driver that will recognize drive bigger than a gig, do I need a New bios >from Adaptec to support the new driver from NeXT? >( I have 1542B and I've ordered the bios from adaptec to support >the drive that is bigger than a gig.) Here's the lowdown on 154x support: 1542B 1542C 1542CF 3.1 2 3 1 3.2 2 3 2 3.2+update 3 3 3 where: 1 = not supported 2 = supported for < 1GB drives only 3 = supported for all drives The update has been available from NeXTanswers since February (documents 1520 and 1521); I believe NeXT technical support will also send it to you on a floppy for free. If you need to use the update as a loadable boot driver (i.e. to install), you'll probably need NeXTanswer 1541 as well. The 1542B requires that BIOS/firmware upgrade you mention to support greater-than-1GB drives, even with the new driver. >Another question is... Are other SCSI controllers have same type of problem? >I am specially interested on DPT card, since it allowes to have cache. The DPT cards aren't susceptible to these problems regarding Extended BIOS Translation; however they generally require recent firmware to work. Might want to check NeXTanswers. Ethan
From: otto@tukki.jyu.fi (Otto J. Makela) Newsgroups: comp.sys.next.advocacy,comp.sys.next.bugs,comp.sys.next.misc Subject: No free documentation on the fax interface? Followup-To: comp.sys.next.bugs Date: 26 Apr 1994 15:43:01 +0300 Organization: University of Jyvaskyla, Finland Sender: otto@network.cc.jyu.fi Message-ID: <OTTO.94Apr26154255@tukki.jyu.fi> When I emailed a guy at NeXT touted as a "fax specialist" and asked... >I'd like to try to roll my own NeXT fax driver to support a bit more >reasonable/modern set of fax modems. Can you give me an idea what is >involved in this type of thing? He forwarded the query to tech support, which wrote me back... >NeXT's Developer Support Program provides direct, individual support >for software development efforts through telephone, fax, or e-mail. >Developer support is available through an annual Developer Hotline >Subscription, which provides two (2) designated developer contacts >unlimited hotline support at a SRP of $5000/year, subject to the >registered developer discount. Up to two additional named contacts >can be added for a SRP fee of $500/contact/year (subject to >discount), for a total of four (4) named contacts. To better provide >for a smaller organization's support needs, developer support can be >purchased as needed at a rate of $200/hour, one hour minimum (subject >to discount). [More salesdrivel removed] I read this to mean that free documentation on the fax interface does not exist. Correct? They can go and f**k themselves with a cream cheese dildo if they think I'm going to pay that kind of money (AND pay for calling them in the US) to just get a set of interface specs. They are certainly trying to get rid of "hobbyist" type programmers. -- /* * * Otto J. Makela <otto@jyu.fi> * * * * * * * * * * * * * * * * * * */ /* Phone: +358 41 613 847, BBS: +358 4#################################################################### Xref: informatik.uni-muenchen.de comp.sys.next.software:9993 comp.sys.next.bugs:1665 Path: informatik.uni-muenchen.de!lrz-muenchen.de!fauern!news.th-darmstadt.de!terra.wiwi.uni-frankfurt.de!zeus.rbi.informatik.uni-frankfurt.de!news.dfn.de!scsing.switch.ch!swidir.switch.ch!univ-lyon1.fr!ghost.dsi.unimi.it!batcomputer!cornell!travelers.mail.cornell.edu!news.tc.cornell.edu!news.cac.psu.edu!news.pop.psu.edu!ctc.com!newsfeed.pitt.edu!gatech!swrinde!sdd.hp.com!vixen.cso.uiuc.edu!usenet From: cplager@uiuc.edu (Charles Plager) Newsgroups: comp.sys.next.software,comp.sys.next.bugs Subject: X windows Date: 26 Apr 1994 20:21:18 GMT Organization: University of Illinois at Urbana Lines: 10 Message-ID: <2pjt3u$qd1@vixen.cso.uiuc.edu> NNTP-Posting-Host: hobbes.physics.uiuc.edu I am trying to X windows running on my color NeXTstation (running 3.0 with 16 meg ram). I have installed mouseX following (I think) the instructions. The problem I am having is that when I run X, the computer changes backgrounds, waits 20 seconds, and quits. If anybody has any ideas on what could be causing the trouble (I did chmod on /dev/vid0) or has another X program that will (relatively) easily switch back and forth between X and NeXTStep, all advice is would be appreciated. Thanks, Charles Plager cplager@uiuc.edu
From: Mark Crispin <mrc@Tomobiki-Cho.CAC.Washington.EDU> Newsgroups: comp.sys.next.bugs Subject: Re: No free documentation on the fax interface? Date: Tue, 26 Apr 1994 15:42:33 -0700 Organization: University of Washington Message-ID: <MS-C.767400153.1103527590.mrc@Tomobiki-Cho.CAC.Washington.EDU> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII To: Gregor Hoffleit <flight@mathi.uni-heidelberg.de> In-Reply-To: <2pjo3h$cj7@sun0.urz.uni-heidelberg.de> Gregor Hoffleit writes: > So if this is the case, wouldn't it be possible to release this interface > information at least 'on demand', even on a non-profit base ? In a word, NO. NeXT doesn't give a damn about hobbyist programmers. Never have, never will. NeXT doesn't give a damn about individual users. Never have, never will. If you want NeXT to talk to you, send them $5000. Then, instead of telling you ``send us $5000'', they will tell you ``this information is not available for public consumption.''
From: phyd@interaccess.com (Brian Leake) Newsgroups: comp.sys.next.bugs Subject: Re: No free documentation on the fax interface? Date: 26 Apr 1994 21:39:31 -0500 Organization: InterAccess Co., Chicago's Best Internet Provider Message-ID: <2pkj93$b8r@home.interaccess.com> References: <OTTO.94Apr26154255@tukki.jyu.fi> In article <OTTO.94Apr26154255@tukki.jyu.fi>, Otto J. Makela <otto@tukki.jyu.fi> wrote: >When I emailed a guy at NeXT touted as a "fax specialist" and asked... >>I'd like to try to roll my own NeXT fax driver to support a bit more >>reasonable/modern set of fax modems. Can you give me an idea what is >>involved in this type of thing? > >He forwarded the query to tech support, which wrote me back... Hold on, obviously this wasn't the right guy then, was it? Who told you to go to this guy? If you have problems in getting through to people, then ask Darren Smith for the right person. Mail him at NeXT. He's pointed me in the right direction before. Mail Steve_Jobs@next.com if you are getting pissed at service. >>NeXT's Developer Support Program provides direct, individual support >>for software development efforts through telephone, fax, or e-mail. << stuff cut about standard tech-support/developer contract >> This is very standard developer support spiel, nothing unusual here, you'll find it with all the workstation and O/S companies. Getting this kind of support is not cheap, it isn't cheap to give either. >I read this to mean that free documentation on the fax interface >does not exist. Correct? > Not necessarily, you just need to get through to the right person. >They can go and f**k themselves with a cream cheese dildo if they >think I'm going to pay that kind of money (AND pay for calling them Well, I think that's a little strong. >in the US) to just get a set of interface specs. They are certainly >trying to get rid of "hobbyist" type programmers. It's not NeXT's fault you don't live in the US. They don't profit in any way from your long-distance call. Use e-mail, it's always the best way. Go back to you original contact, and tell him you didn't get the right answer. >-- > /* * * Otto J. Makela <otto@jyu.fi> * * * * * * * * * * * * * * * * * * */ > /* Phone: +358 41 613 847, BBS: +358 41 211 562 (V.32bis/USR-HST,24h/d) */ > /* Mail: Kauppak. 1B18/FIN-40100 Jyvaskyla/Finland, ICBM: 62.14N 25.44E */ >/* * * Computers Rule 01001111 01001011 * * * * * * * * * * * * * * * * */ -- _____________________________________________________________________ Image Art Take the next "No problems... Only Solutions" Brian Leake step with NEXTSTEP phyd@interaccess.com
Newsgroups: comp.sys.next.bugs From: mgoedel@muaddib.isar.muc.de (Maximilian Goedel) Subject: Re: About Adaptec 1542B driver Message-ID: <Covwpy.2qL@muaddib.isar.de> Sender: mgoedel@muaddib.isar.de (Michael Maximilian Goedel) Organization: Michael Maximilian Goedel References: <1994Apr25.012721.11944@iitmax.iit.edu> Date: Tue, 26 Apr 1994 20:53:58 GMT In article <1994Apr25.012721.11944@iitmax.iit.edu> yong@iitmax.iit.edu (Yong Yoo) writes: | I've read someone ordering a 1542cf driver for NS3.2 that will recognize | drive bigger than a gig from NeXT. I've been waiting for very | long time and I bet I am not the only one who is either. My question | is... DID NeXT released a driver that will recognize drive that is bigger | than a gig for Adaptec 1542B,C,CF (whatever) or just CF? If this is the | driver that will recognize drive bigger than a gig, do I need a New bios | from Adaptec to support the new driver from NeXT? | ( I have 1542B and I've ordered the bios from adaptec to support | the drive that is bigger than a gig.) | Another question is... Are other SCSI controllers have same type of problem? | I am specially interested on DPT card, since it allowes to have cache. | | | thanks in advance | | Yong Yoo Well I think, the adaptec 1542B has generell problems with drives bigger than 1 GIG. But I heard from somebody, that you can change the EPROM on the 1543B to a new one. Nice Greetings from Munich -- Name: Michael Maximilian Goedel NeXT: mgoedel@muaddib.isar.muc.de LINUX: max@funman.boss.sub.org SGI: f11cs1@rz.unibw-muenchen.de
Newsgroups: comp.sys.next.bugs From: mgoedel@muaddib.isar.muc.de (Maximilian Goedel) Subject: Re: Aaargh, can't eject disk Message-ID: <Covwvo.2rC@muaddib.isar.de> Sender: mgoedel@muaddib.isar.de (Michael Maximilian Goedel) Organization: Michael Maximilian Goedel References: <Cos284.C2x@athena.ulaval.ca> Date: Tue, 26 Apr 1994 20:57:23 GMT In article <Cos284.C2x@athena.ulaval.ca> antoine@arrakis.osd.ulaval.ca (Antoine Gautier) writes: | > > Wim Nijenhuis (nijenhui@sun005.research.ptt.nl) wrote: | > > : I inserted an ordinairy dos format | > > : disk but there was no disk icon in the filemanager. Also the eject option | is | > > : disabled. I'm using NeXTStep 3.0. Is this a known bug? | | Yes. you need diskejectfix, a daemon available from NeXT. It is probably on | their ftp site (ftp.NeXT.com). | | --- | +-------------------------------------------------------+ | | Antoine Gautier | | | Departement OSD, FSA | | | Universite Laval, Quebec, PQ | | | Internet: Antoine@arrakis.osd.ulaval.ca | | +-------------------------------------------------------+ In addition to that, ejecting a bad disk works with the command localhost# disk -e /dev/rfd0b. Because I do also have a fix for that problem. But there are several other lakes, your floppydisk can fall in :-) where the fix seem not working for me. Nice Greetings from Munich -- Name: Michael Maximilian Goedel NeXT: mgoedel@muaddib.isar.muc.de LINUX: max@funman.boss.sub.org SGI: f11cs1@rz.unibw-muenchen.de
Newsgroups: comp.sys.next.bugs From: mgoedel@muaddib.isar.muc.de (Maximilian Goedel) Subject: Re: Configuring Serial line to allow 8-N-1 dial-in ? Message-ID: <Covx43.2t1@muaddib.isar.de> Sender: mgoedel@muaddib.isar.de (Michael Maximilian Goedel) Organization: Michael Maximilian Goedel References: <Coo1K1.2F1@cd.com> Date: Tue, 26 Apr 1994 21:02:27 GMT In article <Coo1K1.2F1@cd.com> richardm@cd.com (Richard Masoner) writes: | In article <1994Apr19.170944.1687@clark.dgim.doc.ca> cbarton@DaBoyz.dgcp.doc.ca (Casey Barton) writes: | > I've tried several different parity settings, none of which will get | >NeXT to talk to the modem in anything other than 7E1. (7O1 doesn't work | >either.) I'm starting to think this is a bug -- but I'd think this would | >have been a very visible one -- 8N1 dialin seems to be the most common, | | This has been my experience too, and in fact NeXT's tty layer | doesn't seem to have any flags to support changing the number | of data bits. When using the POSIX interface, some of the flags | are just completely ignored at the driver level! When writing a | driver for Central Data's tty hardware, I had to apply some | creative interpretation on some of the meanings of the flags to | get these things to work. | -- | Richard F. Masoner | ------------------------------------------------ | "A soft answer turns away wrath." -Proverbs 15:1 Sorry, but never had problems with that: here's the example: ######### /etc/gettytab ############### # default:\ :p8:zp:im=\r\n\r\nNeXT Mach (%h) (%t)\r\n\r\r\n\r:\ :hn=Muaddib:tt=vt100:er=\177:kl=^U:in=^C:lm=\r\n%h login\72 :sp#9600: ... h|callin:\ :zp:p8:sp#9600:im=\r\n\r\nNeXT Mach(%t)\r\n\r\n\r: ############## /etc/ttys ########################## ... ttyfb "/usr/etc/getty std.9600" unknown off secure ttyda "/usr/etc/getty callin" vt100 on Nice Greetings from Munich - Germany -- Name: Michael Maximilian Goedel NeXT: mgoedel@muaddib.isar.muc.de LINUX: max@funman.boss.sub.org SGI: f11cs1@rz.unibw-muenchen.de
From: af@iaka.biomath.jussieu.fr (Alain Fauconnet) Newsgroups: comp.sys.next.sysadmin,comp.sys.next.bugs Subject: Re: Netinfo clients don't bind to clone server Date: 27 Apr 94 13:40:51 GMT Organization: Universites Paris VI/Paris VII - France Distribution: world Message-ID: <af.767454051@iaka> References: <af.767014318@iaka> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit af@iaka.biomath.jussieu.fr (Alain Fauconnet) writes: >I have recently tried to make our NeXTs more reliable by adding a >clone server to our (simple, two-level) Netinfo network of >Nextstations. >I carefully followed the instructions in the "System and Network >Administration" book, and it doesn't seem to work at all. When the >usual server crashes, all the NeXTstations more or less hang, and >even rebooting them doesn't seem to make them re-bind to the clone >server. They complain at boot about not finding a server. (...) Okay, since I've had a few queries about this, I thought I'd post a followup since I've made some progress. Hint #1: have a Solaris 2.x system around to debug net problems ;-), "snoop" is the nicest tool I've seen coming with a Unix box Hint #2: make sure your would-be backup Netinfo server has explicit HOSTNAME= and INETADDR= specifications (*not* -AUTOMATIC-) in /etc/hostconfig so that it knows its own identity without the need for any other machine being up Hint #3: make sure it also has NETMASTER=-YES- in the same file, so that it starts bootpd and rpc.bootparamd at bootup. If not, it won't reply to client's queries. With that, I've been able to have the backup (i.e. clone) Netinfo server survive a crash of the main server and have rebooted clients bind to the backup server. I still haven't fully tested dynamic rebinding, since I have hung NFS mounts to handle first. Someone suggested adding entries for the main and clone hosts as well as "broadcasthost" in /etc/hosts. It didn't seem to change anything. >The manual states that "the ip_address property for broadcasthost >is 255.255.255.255, so the bind request is broadcast to all the >hosts on the local network". I'm not a TCP/IP expert, but it >seems to me that on a local net using subnetting >(netmask=255.255.255.128, broadcast=x.y.z.127) like ours, a >packet to 255.255.255.255 will go to the default router and not >be sent as an broadcast Ethernet frame, will it ? This is not relevant. I checked that the IP broadcast is actually sent a a broadcast Ethernet frame. _Alain_ -- Alain FAUCONNET Ingenieur systeme - System Manager AP-HP/SIM Public Health Research Labs 91 bld de l'Hopital 75013 PARIS FRANCE Mail: af@biomath.jussieu.fr (*no* NeXTmail !) Tel: (+33) 1-40-77-96-19 Fax: (+33) 1-45-86-80-68
Newsgroups: comp.sys.next.bugs From: js@euler.hnv.icem.de(Juergen Sell) Subject: Re: No free documentation on the fax interface? Message-ID: <Cox5A2.C0@euler.hnv.icem.de> Sender: js@euler.hnv.icem.de (Juergen Sell) Organization: Ink Unknown References: <OTTO.94Apr26154255@tukki.jyu.fi> Date: Wed, 27 Apr 1994 12:56:26 GMT In article <OTTO.94Apr26154255@tukki.jyu.fi> otto@tukki.jyu.fi (Otto J. Makela) writes: > [...] > I read this to mean that free documentation on the fax interface > does not exist. Correct? > > They can go and f**k themselves with a cream cheese dildo if they > think I'm going to pay that kind of money (AND pay for calling them > in the US) to just get a set of interface specs. They are certainly > trying to get rid of "hobbyist" type programmers. Well, I do not neccessarily agree with your expression of an emotional outburst bt I do find it a very inconvenient situation also to say the least. I once did want to implement my own set of drivers for a very rare faxmachine with serial interface I happen to own. If you get your hands on anything - I am still interested. For starters you might take a look at jolly's am (answering-machine) which does some of the interesting stuff. Misses the outgoing capabilities, though. am is on the archives, good luck. Juergen --- Juergen Sell E-Mail js@icem.de ( NeXTMail ok ) Deisterstr. 18 Fax ++49-511-440617 BRD 30449 Hannover Fon ++49-511-440688 == What time do we live in when all the word 'revolution' makes you think of == is a new generation of soap powder ?
Newsgroups: comp.sys.next.bugs From: mdw@sitar.jazz.att.com (Mark Wuest) Subject: Re: No free documentation on the fax interface? In-Reply-To: flight@mathi.uni-heidelberg.de's message of 26 Apr 1994 18:55:45 GMT Message-ID: <MDW.94Apr27141933@laura.jazz.att.com> Sender: news@nntpa.cb.att.com (Netnews Administration) Organization: AT&T Bell Labs, Red Hill References: <OTTO.94Apr26154255@tukki.jyu.fi> <2pjo3h$cj7@sun0.urz.uni-heidelberg.de> Date: Wed, 27 Apr 1994 18:19:33 GMT In article <2pjo3h$cj7@sun0.urz.uni-heidelberg.de> flight@mathi.uni-heidelberg.de (Gregor Hoffleit) writes: >Otto J. Makela (otto@tukki.jyu.fi) wrote: >: When I emailed a guy at NeXT touted as a "fax specialist" and asked... >: >I'd like to try to roll my own NeXT fax driver to support a bit more >: >reasonable/modern set of fax modems. Can you give me an idea what is >: >involved in this type of thing? > >: He forwarded the query to tech support, which wrote me back... >: >NeXT's Developer Support Program provides direct, individual support <Please Send Money diatribe deleted> >: I read this to mean that free documentation on the fax interface >: does not exist. Correct? Correct. It is therefore UNDOCUMENTED, since very few are going to PAY to look at an API. >I can only second this. It's a pity that NeXT is so concerned about >parts of their OS - hardware interfaces. The fax driver is just one >example for that. >In some magazines NEXTSTEP was listed as 'lacking FAX support'! Funny, >if there's a FAX button on every print panel... NeXTSTEP lacks fax support. There are a few faxmodems supported by NXFax. Mine isn't, for example (it's a DataPort from that small fly-by-night outfit, AT&T). Geez. NeXT doesn't support any AT&T faxmodems? Nope. Let's say it again (as if it will do any good): NEXTSTEP is lacking fax support. If you want to fax, buy NXFax bundled with a modem. I personally use my DoveFax for faxxing and my DataPort for data (but it means I have to unplug my DeskJet). I *did* get source to a generic fax driver that shows the NS side of things, but not from NeXT. Now all I need is time to munge on it so I can write my own... Mark -- -- Mark Wuest mdw@sitar.jazz.att.com (NeXT Mail Ok)
From: gaia@wam.umd.edu (L. Anathea Brooks) Newsgroups: comp.sys.next.bugs Subject: chunk files Date: 28 Apr 1994 01:50:43 GMT Organization: University of Maryland, College Park Message-ID: <2pn4pk$6a2@cville-srv.wam.umd.edu> Does anyone else have the great troubles I have with .chunk files? I'm on Intel hw, and when I get home with a file divided between a number of disks (with .chunk extension), 9 out of ten times something goes wrong. Often it's a "File mover died unexpectedly". Sometimes the file system doesn't find chunk number two, for ex. Anyone? Thanks Robert de Lucca Johns Hopkins
From: swift@acs3.nntp-read.bu.edu (Matthew Swift) Newsgroups: comp.sys.next.bugs Subject: Re: Aaargh, can't eject disk Date: 28 Apr 94 00:14:02 Organization: Boston University Information Technology Boston, MA USA Message-ID: <SWIFT.94Apr28001402@acs3.nntp-read.bu.edu> References: <Cos284.C2x@athena.ulaval.ca> <Covwvo.2rC@muaddib.isar.de> In-reply-to: mgoedel@muaddib.isar.muc.de's message of Tue, 26 Apr 1994 20:57:23 GMT Hi, I've had consistent problems with removable disks using NS3.0; I'll review them, and tell you how I get along now. My system has a CD and SCSI floppy, and an optical, and these problems happen with all three, though with the read-only CD least of all, I think. I tried the fix demon, and it only made things worse for me. Using it, a newly-inserted disk would show up from the unix prompt, but never be recognized by the workspace manager, eg, never open a new browser if that's how you had it set in Preferences, etc. If you can eject a disk from the WSM, then that's great, but sometimes it's just grayed out. So go to a csh. If you brute-force eject a disk (with disk -e) without "umount" it, you can run into system-crashing problems. I think this only happens if you put in another disk, but that's why you ejected the last one, right? The problem then is that "umount" won't do its thing if it thinks the deivce is "busy" and ther is the real problem with NS 3.0: devices sometimes get stuck in busy mode. In this case, compile and use the short C program below, which conveniently umounts and ejects a disk if possible. Unfortunately, you need root access to invoke umount. If you don't have it, the way around it is to get someone who does to install the file properly, as indicated in my friend's comment. Ask someone else to find out more, I don't know this security stuff. Sometimes the handy little prog won't work anyway, saying the disk is busy. Here are a list of things to do if that happens: 1) make sure it's not legitimately busy. Check all your apps to see if they have any file from it that's open. Make sure all your shells have no stopped jobs that are using a file, and make sure the shell itself is not cd'd to the disk! (That last one took me a while.) 2) Go to the WSM and CMD-U)pdate viewers. Then try again. 3) If it's still busy, wait five minutes. I'm serious, it's like magic, sometimes the stupid drives will unbusy themselves of their own accord. 4) Log in and log out. We're getting desperate at this point. 5) If that doesn't work, and trying again on the fresh login doesn't work, then just brute force it with disk -e, but I would log in and log out again after doing this, or you might be in trouble. 6) syrup of ipecac in the drive door? Here is the program you can compile, strip, and use if you have root access; I use it with this alias: alias fject su -f -c 'feject !*' /* feject.c : written 10/15/92 by Garth Zeglin Feject calls /etc/umount and /usr/etc/disk to unmount and eject a floppy. It is written as a C program so it can run with root priveleges when the setuid mode bit (mask 4000) is set. The only possible option is a different name for the mount directory. The default is /UntitledDisk. */ #include <stdio.h> #include <sys/wait.h> #define FLOPPYDEVICE "/dev/rsd1a" char name[] = "/UntitledDisk"; char *mntname; union wait status; int waitval; int mntid; main(int argc, char **argv) { int c; if (argc > 1) mntname = argv[1]; else mntname = name; mntid = fork(); if (mntid == 0) { /* if this is the child process */ execl("/etc/umount", "umount", mntname, NULL); /* run the unmount */ /* if it returns, there has been an error */ fprintf(stderr, "\nCannot run /etc/umount.\n"); exit(1); } if (mntid == -1) { /* fork error */ fprintf(stderr, "\nUnable to fork.\n"); exit(1); } /* wait for unmount process to complete */ waitval = wait(&status); if (waitval == -1) { /* no children? */ fprintf(stderr, "\nwait() error: reports no child processes.\n"); exit(1); } if (status.w_status & 0xff != 0) { fprintf(stderr, "\nUnmount failed.\n"); exit(1); } if (((status.w_status >> 8) & 0xff) > 0) { fprintf(stderr, "\n/etc/umount reported non-zero exit status.\n"); exit(1); } /* otherwise, run eject */ execl("/usr/etc/disk","disk","-e",FLOPPYDEVICE,NULL); }
From: swift@acs3.nntp-read.bu.edu (Matthew Swift) Newsgroups: comp.sys.next.bugs Subject: Re: Dig. Librarian can't search on regular expressions? Date: 28 Apr 94 00:23:41 Organization: Boston University Information Technology Boston, MA USA Distribution: world Message-ID: <SWIFT.94Apr28002341@acs3.nntp-read.bu.edu> References: <thompsonCoqCJr.LGB@netcom.com> <2pgi9kINNq8c@no-names.nerdc.ufl.edu> In-reply-to: statman@stat.ufl.edu's message of 25 Apr 1994 13:58:12 GMT The indexing library is so buggy in 3.0 that I use it as a last resort; find and GNU's greps do pretty well. 3.2 was supposed to have fixed all this up. I was even going to buy 3.2 for that reason, until problems like these showed themselves in 3.2 also. (anyone at NeXT listening?) In 3.0, Librarian crashes, and on %90 of indexed directories, it seems to search the whole tree instead of using the index. I have found no way around this. Using the indexing routines from the command line produces the same results. I recommend cursing, waiting for the next NS release, and teaching yourself to use 'find' 'XXXtoascii' and GNU's 'grep's. 'xargs' is a little-known and very useful utility, also.
Newsgroups: comp.sys.next.bugs From: garylang@netcom.com (Gary Lang) Subject: Re: No free documentation on the fax interface? Message-ID: <garylangCoyJpo.6qB@netcom.com> Organization: NETCOM On-line Communication Services (408 241-9760 guest) References: <OTTO.94Apr26154255@tukki.jyu.fi> Date: Thu, 28 Apr 1994 07:05:47 GMT Otto J. Makela (otto@tukki.jyu.fi) wrote: : When I emailed a guy at NeXT touted as a "fax specialist" and asked... : >I'd like to try to roll my own NeXT fax driver to support a bit more : >reasonable/modern set of fax modems. Can you give me an idea what is : >involved in this type of thing? : They can go and f**k themselves with a cream cheese dildo if they : think I'm going to pay that kind of money (AND pay for calling them : in the US) to just get a set of interface specs. They are certainly : trying to get rid of "hobbyist" type programmers. : -- I hope so. Supprting everyone who calls in and asks for technical and developer support for free is not a very fruitful way to make money. -- Gary Lang Email:garylang@netcom.COM
From: don@darth.byu.edu (Don Yacktman) Newsgroups: comp.sys.next.bugs Subject: Re: chunk files Date: 28 Apr 1994 05:25:00 GMT Distribution: world Message-ID: <2pnhbc$mf1@bones.et.byu.edu> References: <2pn4pk$6a2@cville-srv.wam.umd.edu> L. Anathea Brooks writes ) Does anyone else have the great troubles I have ) with .chunk files? I'm on Intel hw, and when I get ) home with a file divided between a number of disks ) (with .chunk extension), 9 out of ten times something ) goes wrong. Often it's a "File mover died unexpectedly". ) Sometimes the file system doesn't find chunk number ) two, for ex. Chunk them yourself from the terminal. Do NOT use the Workspace for this! Chunking is seriously broken and you'll have to unchunk it by hand half the time anyway it seems. To chunk by hand, check out the split(1) command...works like a charm, and almost as easy as the WorkSpace...but it actually works reliably unlike the WorkSpace. Especially for large files, do NOT chunk them! The restoration typically requires that you have 3-4X the space of the chunked file available in VM, meaning that on a single disk system the unchunking process will typically fill the disk until no space is left to write the restored file to disk. Bad news. As an example, I recently aided a friend in un-chunking a 70 Meg file that was spread across 30 disks. Trying from WorkSpace made the swapfile grow to OVER 180 MEG before the system died. (The file, by hand, was unpackable in about 80 meg with some creative work on repacking into smaller files as I unpacked it.) It seems that WorkSpace tries to keep the following in memory all at once: (1) the original images off the disks, (2) the re-assembled .Z file, (3) the unpacked file. And then, with all that taking up swap space it tries to write out to disk. So for the 70 Meg example above, I would have needed to have at least 200 Meg free if it were a single drive system, or 130 Meg on the swap device and 70 Meg on the target drive. (Those are actually lowball numbers, since there's something even more braindead going on. like an extra 70 meg copy in the swapfile somewhere, since 180 meg swap wasn't enough when I tried the second scenario. Aargh!) I guess whomever wrote the WorkSpace chunking feature hasn't ever heard of /tmp and such. (Actually /tmp might not be the best solution either, but it's better than growing the swapfile without bounds.) Until chunking works reasonably though, I'd avoid it like the plague. Sorry, but after having to deal with frustrations like this, it is a bit of a sore spot with me. !@*$)(!@?!&$!! braindead implementations. -- Later, -Don Yacktman Don_Yacktman@byu.edu
From: jeff@motorrad.Berkeley.EDU (Jeff Louie) Newsgroups: comp.sys.next.bugs,comp.sys.next.sysadmin Subject: IDE problem Date: 28 Apr 1994 06:11:26 GMT Organization: University of California, Berkeley Message-ID: <2pnk2e$9u5@agate.berkeley.edu> I'm trying to set up NextStep 3.2 to run on a 486 machine, but am having problems with it. The machine has a 540MB IDE drive, and the CD ROM drive is hooked up to an Adaptec 1542C SCSI controller. It gets to as far as formatting/partitioning the disk, but while it is laying out the basic boot files, it comes up with the following error message about 30% of the way through: Ide: ideFlushIntMsgs:StrayInterrupt :ideWriteMultiple Ide:softResetIde diagError = 0x1 IDE Drive 0 using MultiSector Mode This message is repeated 2 or 3 times with slight changes, but usually when it's about done with loading the boot files. Later, as it is trying to boot up with what it has, it comes up only in read-only mode, and doesn't get much further than the initial settings panel. Does anyone know what the error message is trying to tell me? It there a possible conflict with the IDE controller and the SCSI controller? Or is the disk drive suspect? Any advice would be greatly appreciated. Jeff Louie (jeff@ieor-next1.Berkeley.EDU) <-- please reply to this address UC Berkeley-IEOR
From: jds@aplexus.jhuapl.edu (John D Stanhope) Newsgroups: comp.sys.next.bugs Subject: Re: chunk files Date: 28 Apr 1994 12:23:08 GMT Organization: Johns Hopkins U. Applied Physics Lab Distribution: world Message-ID: <2po9rc$571@aplcomm.jhuapl.edu> References: <2pn4pk$6a2@cville-srv.wam.umd.edu> In article <2pn4pk$6a2@cville-srv.wam.umd.edu>, gaia@wam.umd.edu (L. Anathea Brooks) writes: |> Does anyone else have the great troubles I have |> with .chunk files? I'm on Intel hw, and when I get |> home with a file divided between a number of disks |> (with .chunk extension), 9 out of ten times something |> goes wrong. Often it's a "File mover died unexpectedly". |> Sometimes the file system doesn't find chunk number |> two, for ex. |> I would like to second that. But my problem occurs when saving files larger than 1MB to floppy. It starts off well enough but when I put in the second disk it complains about not seeing chunk 0. I have tried this a dozen or so time and it has only worked once. I just wish I remebered what I did to get it to work. -- ------------------------------------------------------------------------------- Otto "Apes don't read philoshpy" disclaimer Wanda "Sure they do, they just don't understand it" -------------------------------------------------------------------------------
From: robert@steffi.demon.co.uk (Robert Nicholson) Newsgroups: comp.sys.next.bugs Subject: /bin/strip is an ugly beast.... Date: 28 Apr 1994 11:05:31 +0100 Organization: me organised, that's a joke. Message-ID: <2po1pb$23f@steffi.demon.co.uk> Why does strip remove suid persmissions from executables? I'd advise against "strip /usr/local/bin/*" this just toasted my MTA. -- "Kill files are for pacifists" (ASCII for text only messages)
From: hvillega@roxette (Hugo Villegas Roji) Newsgroups: comp.sys.next.bugs Subject: Agfa-NeXT via ethernet? Date: 28 Apr 1994 15:00:37 GMT Organization: ITESM, Campus Monterrey Message-ID: <2poj2l$1cr1@mtecv2.mty.itesm.mx> Hi! I istalled NeXTSTEP 3.2 in our server and we found the PrintManager does not have the possibility to conect and Agfa Accuset imagesetter via Ethernet wi th apple share protocols. With 3.0 it was working well but I can't install 3.0 a gain. Anybody out there knows how can I conect this imagesetter to our NeXT comp uters net? I'll apreciate any help Thank you in advance ................................................................. Hugo Ernesto Villegas Roji . Nexit Diseno Grafico . . Ricardo Margain Zozaya 210-5 / CP 66267 . Valle de Santa Engracia / Garza Garcia / Nuevo Leon / Mexico . tels. 52 (8) 335 19 14 / 335 19 15 / 335 08 87 / 356 30 53 fax . .................................................................
Newsgroups: comp.sys.next.bugs From: ctm@ardi.com (Clifford T. Matthews) Subject: Re: /bin/strip is an ugly beast.... Message-ID: <CozDEq.FBE@cobra.cs.unm.edu> Sender: news@cobra.cs.unm.edu Organization: ARDI References: <2po1pb$23f@steffi.demon.co.uk> Date: Thu, 28 Apr 1994 17:47:16 GMT In article <2po1pb$23f@steffi.demon.co.uk> robert@steffi.demon.co.uk (Robert Nicholson) writes: >Why does strip remove suid persmissions from executables? > >I'd advise against "strip /usr/local/bin/*" this just toasted my MTA. I mentioned various problems with setuid bits a long time ago. I sent info to NeXT with their BugNeXT application and I sent a message to this mailing list. I no longer have copies of my posts, or even incoming mail from back then, but I do have a copy of some outgoing e-mail I sent to a kind NeXT employee who responded to my comments in comp.sys.next.bugs (his reply was that POSIX compliance caused the ruckus): The entire body of my e-mail, sent 22 Jul 93 02:06:51 MDT was: Thanks for the reply. Even though there is I say more about this issue in the rest of my letter I appreciate the reply more than you can know and don't care if you ignore the rest of what I say. I certainly don't want to dump on you because you were kind enough to send me e-mail. I sent something with the Bug NeXT program a while back (while 3.1 was still in beta) and didn't get anything back. At minimum, the man page for write(2) should be updated, but that begs the question of commands like strip. Does POSIX stipulate that when a root users strips a setuid root file that the setuid bit gets cleared? If not, three cheers for POSIX and strip should be fixed. If so, well I've never been a big fan of POSIX, but that too should be documented. I sincerely hope that being POSIX compliant is a net win for NeXT. I wish you guys the best. [upon rereading, I realize the second sentence doesn't parse-- oops] I show the letter above both to show that I've had individual correspondence with someone from NeXT and that I sent in a bug report to NeXT about it. Even if they didn't fix the offending behaviour, they could have at least updated their manual pages. I just checked on 3.2, and they are still broken. Oh yes, we weren't allowed to participate in the 3.2 beta program. In fact, we're not even registered developers anymore, and the only person in NeXT developer relations who took us seriously has been shifted to another part of NeXT. Believe it or not, I still wish NeXT all the best. I am just realistic and recognize that they don't want bug reports from us, or products developed for their OS. March '94 was ARDI's first profitable month in a year, and April is better than March. We owe it all to Executor sales ... ... Executor/DOS, that is. --Cliff ctm@ardi.com
From: grio@next.com (Dan Grillo) Newsgroups: comp.sys.next.bugs Subject: Re: /bin/strip is an ugly beast.... Date: 28 Apr 1994 19:58:03 GMT Organization: Technical Support, NeXT Computer, Inc. Message-ID: <2pp4gb$oa3@rosie.next.com> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Versions: dmail 1.9/makemail 2.2a robert@steffi.demon.co.uk writes in comp.sys.next.bugs: > Why does strip remove suid persmissions from executables? > I'd advise against "strip /usr/local/bin/*" this just toasted my MTA. Was this the problem? From the write(2) man page: If the real user is not the super-user, then write clears the set-user-id bit on a file. This prevents penetration of system security by a user who "captures" a writable set- user-id file owned by the super-user. --Dan
Newsgroups: comp.sys.next.bugs Subject: Re: /bin/strip is an ugly beast.... Message-ID: <Cp09Fq.H30@cobra.cs.unm.edu> From: ctm@ardi.com (Clifford T. Matthews) Date: Fri, 29 Apr 1994 05:19:05 GMT Sender: news@cobra.cs.unm.edu References: <2pp4gb$oa3@rosie.next.com> Organization: ARDI In article <2pp4gb$oa3@rosie.next.com> grio@next.com (Dan Grillo) writes: > >robert@steffi.demon.co.uk writes in comp.sys.next.bugs: >> Why does strip remove suid persmissions from executables? >> I'd advise against "strip /usr/local/bin/*" this just toasted my MTA. > >Was this the problem? > >From the write(2) man page: > > If the real user is not the super-user, then write clears > the set-user-id bit on a file. This prevents penetration of > system security by a user who "captures" a writable set- > user-id file owned by the super-user. > > --Dan No. That is not the problem. Note the "If the real user is not the super-user" part. See below: bash# 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: 25 MHz Processor active: 0 System type: 1 Board revision: 0x0 Primary memory available: 20.00 megabytes. Default processor set: 62 tasks, 96 threads, 1 processors Load average: 0.00, Mach factor: 0.99 bash# whoami root bash# gcc -o foo foo.c bash# chmod u+s foo bash# ls -l foo -rwsrwxr-x 1 root wheel 48944 Apr 28 23:06 foo bash# strip foo bash# ls -l foo -rwxrwxr-x 1 root wheel 16384 Apr 28 23:06 foo BTW, the problem is in fact with the write system call not behaving as it is documented. See below: bash# cat foo2.c #include <sys/file.h> main() { int fd; fd = open("dummyfile", O_WRONLY); write(fd, "", 1); close(fd); } bash# ls -l dummyfile -rws--x--x 1 root wheel 0 Apr 28 23:09 dummyfile bash# ./foo2 bash# ls -l dummyfile -rwx--x--x 1 root wheel 1 Apr 28 23:11 dummyfile I reported this bug back when *3.1* was in beta. I mentioned both the discrepancy between the write documentation and the write behaviour (which bit me since I had code that depended on the setuid bit not getting gratuitously changed when a root process did a write -- can you say "race condition"?) and I also mentioned the strip problem sometime, although I don't recall whether or not it was a separate bug report. I think it was, because I remember being bitten by the strip problem also, although it's quite easy to be bitten even when you know a bug already exists. I received an electronic confirmation that my bug report had been submitted, but I received no comments from a human until after I posted on comp.sys.next.bugs, and then the person who commented was a concerned NeXT employee (NeXT is filled with them), but not someone whose job it is to fix bugs (this space intentionally left blank). --Cliff ctm@ardi.com
From: robert@steffi.demon.co.uk (Robert Nicholson) Newsgroups: comp.sys.next.bugs Subject: Re: /bin/strip is an ugly beast.... Date: 29 Apr 1994 07:27:40 GMT Organization: me organized? That's a joke! Message-ID: <ROBERT.94Apr29082740@steffi.demon.co.uk> References: <2pp4gb$oa3@rosie.next.com> To: grio@next.com (Dan Grillo) In-reply-to: grio@next.com's message of 28 Apr 1994 19:58:03 GMT In article <2pp4gb$oa3@rosie.next.com> grio@next.com (Dan Grillo) writes: robert@steffi.demon.co.uk writes in comp.sys.next.bugs: > Why does strip remove suid persmissions from executables? > I'd advise against "strip /usr/local/bin/*" this just toasted my MTA. Was this the problem? From the write(2) man page: If the real user is not the super-user, then write clears the set-user-id bit on a file. This prevents penetration of system security by a user who "captures" a writable set- user-id file owned by the super-user. --Dan Really? root:/usr/local/bin>ls -al smail 8:26 -r-sr-xr-x 1 root 196608 Apr 28 10:56 smail root:/usr/local/bin>strip smail 8:26 strip: input object file stripped: smail root:/usr/local/bin>ls -al smail 8:26 -r-xr-xr-x 1 root 196608 Apr 29 08:26 smail -- "Emacs isn't pretty. It's functional!" (ASCII for text only messages)
Newsgroups: comp.sys.next.bugs From: andrew@stone.com (Andrew Stone) Subject: Re: chunk files Message-ID: <1994Apr29.025617.3134@stone.com> Sender: andrew@stone.com Organization: Stone Design Corp References: <2pn4pk$6a2@cville-srv.wam.umd.edu> Date: Fri, 29 Apr 1994 02:56:17 GMT In article <2pn4pk$6a2@cville-srv.wam.umd.edu> gaia@wam.umd.edu (L. Anathea Brooks) writes: > Does anyone else have the great troubles I have > with .chunk files? I'm on Intel hw, and when I get > home with a file divided between a number of disks > (with .chunk extension), 9 out of ten times something > goes wrong. Often it's a "File mover died unexpectedly". > Sometimes the file system doesn't find chunk number > two, for ex. > > Anyone? > > Thanks > > Robert de Lucca > Johns Hopkins > Are you using 3.less than 2? I tried mailing to you earlier via Sys admin mail, but it bounced, and this might be of general use to the net. Try this failsafe sneaker-net via uuecode technique: # tar & compreess up the directories in question: gnutar cf directory.gtar directory gzip -9 directory.gtar # uuencode it: uuencode directory.gtar.z directory.gtar.z > directory.gtar.z.uu # split it into maximal size for 1.4 meg floppies, DOS formatted mkdir pieces cd pieces split -22900 ../directory.gtar.z.uu pieces. # this produces pieces.aa, pieces.ab, ... # copy them onto DOS formatted disks # sneaker on home with floppies # copy them one by one into /tmp cat pieces.a? > directory.gtar.z.uu uudecode directory.gtar.z.uu gunzip directory.gtar.z | gnutar xf - # do a sum on the directory.gtar at both ends to be extra certain andrew -- ||<<->>||<<==>>||<<++>>||<<?>|<+>>||<<-->>||<<==>>||<<+>>|| !! Andrew Stone | (505) 345-4800 !! !! andrew@stone.com | Stone Design Corp !! ||<<->>||<<==>>||<<++>>||<<?>|<+>>||<<-->>||<<==>>||<<+>>||
Newsgroups: comp.sys.next.bugs,comp.sys.next.programmer From: karl@trapac.com (Karl Kraft) Subject: Re: /bin/strip is an ugly beast.... Message-ID: <Cp14F1.1AI@trapac.com> Sender: karl@trapac.com (Karl Kraft) Organization: Trans Pacific Container Service Corporation References: <2po1pb$23f@steffi.demon.co.uk> Date: Fri, 29 Apr 1994 16:28:12 GMT In article <2pgdnu$q9@steffi.demon.co.uk> robert@steffi.demon.co.uk (Robert Nicholson) writes: > Why does strip not use lstat to determine if it's stripping via a link? In article <2po1pb$23f@steffi.demon.co.uk> robert@steffi.demon.co.uk (Robert Nicholson) writes: > Why does strip remove suid persmissions from executables? > Well, I'll join these two threads together, because the answer is the same for both. Let us say you are going to strip a binary called "mycli". When you strip this program, strip moves the file to "mycli.strip" It then reads this file to create a NEW executable called "mycli". This begs the question, "why not strip in place?" Because strip can do more than just kill the macho section that contains debugging symbols. It actually needs to write a new file. You can do it in place. So unless you wanted strip to be a suid program, (BAD IDEA!), it couldn't put suid bits on the new executable. Finally, strip shouldn't be a program you run on random installed binaries. It should be used as part of the "make install" process. It's a part of the build process, not a general tool. Since it's not a general tool don't expect it to behave like one. You can't just strip a binary and expect it to still work. Many programs need to not be stripped to support runtime linking. -- Karl Kraft Karl_Kraft@trapac.com Karl_Kraft@ensuing.com [My opinions are my own]
From: dillon@apollo.west.oic.com (Matthew Dillon) Newsgroups: comp.sys.next.bugs,comp.sys.next.programmer Subject: Re: /bin/strip is an ugly beast.... Date: 29 Apr 1994 14:50:20 -0700 Organization: Obvious Implementations Corp Distribution: world Message-ID: <2prves$8ml@apollo.west.oic.com> References: <2po1pb$23f@steffi.demon.co.uk> <Cp14F1.1AI@trapac.com> In article <Cp14F1.1AI@trapac.com> karl@trapac.com writes: :same for both. : :Let us say you are going to strip a binary called "mycli". When you :strip this program, strip moves the file to "mycli.strip" It then :reads this file to create a NEW executable called "mycli". : :This begs the question, "why not strip in place?" : :Because strip can do more than just kill the macho section that :contains debugging symbols. It actually needs to write a new file. :You can do it in place. : :So unless you wanted strip to be a suid program, (BAD IDEA!), it :couldn't put suid bits on the new executable. Nonsense, strip doesn't have to be suid in order to set the suid bit when the user running strip owns the file in question! :You can't just strip a binary and expect it to still work. Many :programs need to not be stripped to support runtime linking. This has nothing to do with the incorrect behavior of strip, though. -Matt -- Matthew Dillon dillon@apollo.west.oic.com 1005 Apollo Way Incline Village, NV. 89451 ham: KC6LVW (no mail drop) USA Sandel-Avery Engineering (702)831-8000 [always include a portion of the original email in any response!]
From: sowa@amdew.llnl.gov (Erik C. Sowa) Newsgroups: comp.sys.next.bugs Subject: Re: chunk files Date: 29 Apr 1994 22:39:58 GMT Organization: LLNL Chemistry and Materials Science Message-ID: <SOWA.94Apr29153958@amdew.llnl.gov> References: <2pn4pk$6a2@cville-srv.wam.umd.edu> <1994Apr29.025617.3134@stone.com> In-reply-to: andrew@stone.com's message of Fri, 29 Apr 1994 02:56:17 GMT Or build the binary-file splitter bsplit and skip the uuencode/decode step. Here's one short source, courtesy of Nelson H. F. Beebe. /*********************************************************************** Split a binary file into pieces of a specified (or default) size. Usage: bsplit [[-nnn] [-] filename] A -nnn bytecount option sets the split piece size for subsequent files. Stdin is assumed if a bare - is given, or if no files are specified. Multiple filenames may be specified. No check is made to see whether the output files (named with the input file name and the suffixes "-001", "-002", etc.) exist. [29-Aug-1990] Nelson H. F. Beebe ***********************************************************************/ #include <stdio.h> #include <string.h> #ifdef __STDC__ #ifndef __GNUC__ #include <stdlib.h> #endif /* __GNUC__ */ #ifdef __COMPILER_KCC__ #define WB_FLAGS "wb8" #define RB_FLAGS "rb8" void bsplit(); #else /* NOT __COMPILER_KCC__ */ void bsplit(int size, char *name); #endif /* __COMPILER_KCC__ */ #else /* NOT __STDC__ */ void bsplit(); #endif /* __STDC__ */ #ifdef VMS #define RB_FLAGS "rb","rfm=fix","bls=512","mrs=512" #define WB_FLAGS "wb","rfm=fix","bls=512","mrs=512" #endif #ifndef RB_FLAGS #define RB_FLAGS "r" #endif #ifndef WB_FLAGS #define WB_FLAGS "w" #endif #ifdef ABS #undef ABS #endif #define ABS(n) ( ((n) < 0) ? -(n) : (n) ) #define ROUNDABSUP(n) ((ABS(n) + 511) & ~0x1ff) #define DEFAULT_SIZE 32768 int main(argc, argv) int argc; char *argv[]; { int size; int k; int nfiles; size = DEFAULT_SIZE; for (nfiles = 0, k = 1; k < argc; ++k) { if (strcmp(argv[k], "-") == 0) /* bsplit stdin */ { bsplit(ROUNDABSUP(size), (char *) NULL); nfiles++; } else if (*argv[k] == '-') /* expect -nnn */ { (void) sscanf(&argv[1][1], "%d", &size); if (size < 512) /* prevent nonsense values */ size = DEFAULT_SIZE; } else { bsplit(ROUNDABSUP(size), argv[k]); nfiles++; } } if (nfiles == 0) /* just process stdin */ bsplit(ROUNDABSUP(size), (char *) NULL); exit(0); return (0); /* keep compilers happy */ } void bsplit(size, inname) int size; char *inname; { register FILE *in; register FILE *out; char outname[512]; register int n; int npart; register int c; if (inname == (char *) NULL) { in = stdin; inname = "stdin"; } else in = fopen(inname, RB_FLAGS); if (in == (FILE *) NULL) { (void) fprintf(stderr, "?Cannot open file [%s]\n", inname); return; } (void) fprintf(stderr, "[%s]\n", inname); for (npart = 1; !feof(in); ++npart) { sprintf(outname, "%s-%03d", inname, npart); out = fopen(outname, WB_FLAGS); if (out == (FILE *) NULL) { (void) fprintf(stderr, "?Cannot open file [%s]\n", outname); return; } (void) fprintf(stderr, "\t[%s]", outname); for (n = 0; n < size; ++n) { c = getc(in); if ((c == EOF) && feof(in)) break; putc(c, out); } (void) fclose(out); (void) fprintf(stderr, " %d bytes [OK]\n", n); } } -- erik sowa (sowa@amdew.llnl.gov)
Newsgroups: comp.sys.next.bugs,comp.sys.next.admin From: mark@ve6mgs.ampr.ab.ca (Mark Gregory Salyzyn) Subject: Re: /bin/strip is an ugly beast ... a fix Message-ID: <Cp2uAH.AD1@ve6mgs.ampr.ab.ca> Organization: VE6MGS Gateway References: <2po1pb$23f@steffi.demon.co.uk> Date: Sat, 30 Apr 1994 14:44:39 GMT robert@steffi.demon.co.uk (Robert Nicholson) writes: >Why does strip remove suid persmissions from executables? > >I'd advise against "strip /usr/local/bin/*" this just toasted my MTA. A couple of choices, depending on how much attention to detail you prefer ... mv /bin/strip /bin/strip.NeXT mv <following shell script> /bin/strip chmod 755 /bin/strip - or - ln /bin/strip /bin/strip.NeXT (Or edit strip.fixed below) mv <following shell script> /usr/local/bin/strip.fixed chmod 755 /usr/local/bin/strip.fixed (and enter strip.fixed instead of strip) - or - Complain Ciao -- Mark Complain profusely, it is your democratic right to complain and not suffer! ---------------------- Go Ahead, Cut me! --------------------- #! /bin/sh # Name: strip # Description: replace the NeXT POSIX compliant strip with one that # serves my purpose more adequately. # Author: Mark Salyzyn mark@ve6mgs.ampr.ab.ca April 1994 # Bug reports and updates gleafully accepted # ARGS= FILES= while test $# -gt 0 ; do case $1 in -s|-d) ARGS="$ARGS $1 $2" ; shift ;; -s*) ARGS="$ARGS -s `echo X$1 | dd bs=3 skip=1 2>/dev/null`" ;; -d*) ARGS="$ARGS -d `echo X$1 | dd bs=3 skip=1 2>/dev/null`" ;; -A|-n|-u|-S|-X|-x) ARGS="$ARGS $1" ; shift ;; -) shift ; break ;; *) FILES="$FILES $1" ;; esac shift done # Remaining arguments are to be file names if test $# -gt 0 ; then FILES="$FILES $*" fi # Now process the arguments, one file at a time (to keep my wits) for i in $FILES ; do if test -d $i ; then echo "$i is a directory, will not strip" 2>&1 continue fi if test ! -f $i ; then echo "$i is not a regular file, will not strip (Try \`strip.NeXT $ARGS $i')" 2>&1 continue fi # Record the current mode flags MODE=`ls -l $i | dd bs=10 count=1 2>/dev/null` if test "X-" != "`echo X$MODE | dd bs=2 count=1 2>/dev/null`" ; then echo "$i is not a regular file, will not strip (Try \`strip.NeXT $ARGS $i')" 2>&1 continue fi # Run the original NeXT version of strip. strip.NeXT $ARGS $i # Now, interpret the various ls -l bits to restore the mode of the file MODE="`echo X$MODE | dd bs=2 skip=1 2>/dev/null`" FAIL=-f for j in u g o ; do BITS="`echo XXX$MODE | dd bs=3 skip=1 count=1 2>/dev/null`" MODE="`echo X$MODE | dd bs=4 skip=1 2>/dev/null`" case $BITS in r??) chmod $FAIL $j+r $i ; FAIL= ;; esac case $BITS in ?w?) chmod $FAIL $j+w $i ; FAIL= ;; esac # we take the simple route of *assuming* execute status for the # hidden execute bits where suid and sticky bit is involved ... case $BITS in ??x) chmod $FAIL $j+x $i ; FAIL= ;; ??s) chmod $FAIL $j+x,$j+s $i ; FAIL= ;; ??t) chmod $FAIL $j+x,$j+t $i ; FAIL= ;; esac done done
From: work@dannug.dk (Michael Hallin) Newsgroups: comp.sys.next.bugs Subject: Re: Agfa-NeXT via ethernet? Date: 30 Apr 1994 11:50:09 GMT Organization: Danish NeXT User Group Message-ID: <2ptglh$4ai@machthenext.dannug.dk> References: <2poj2l$1cr1@mtecv2.mty.itesm.mx> In article <2poj2l$1cr1@mtecv2.mty.itesm.mx> hvillega@roxette (Hugo Villegas Roji) writes: |>Hi! |> |> I istalled NeXTSTEP 3.2 in our server and we found the PrintManager does |> not have the possibility to conect and Agfa Accuset imagesetter via Ethernet wi |>th apple share protocols. With 3.0 it was working well but I can't install 3.0 a |>gain. Anybody out there knows how can I conect this imagesetter to our NeXT comp |>uters net? |> I'll apreciate any help |> |>Thank you in advance |> |>.................................................................. |> Hugo Ernesto Villegas Roji . |> Nexit Diseno Grafico . |> . Hi, I have succesfully connected a NeXT Turbo Color to a Fiery ColorServer over Ethernet, and the Fiery ColorServer uses AppleTalk/EtherTalk. We installed a software package on the NeXT called uShare/Partner from a company called IPT Technologies. Works great! Best regards Michael --- _____________________________________________ Michael Hallin Copenhagen, Denmark NeXTMail: work@dannug.dk NonNeXTMail: mh.xeroxvang@rxdk.xerox.com Voice: Int + 45 43 53 34 33 Fax: Int + 45 43 53 34 33 Work: Int + 45 43 43 43 03 Fax: Int + 45 43 43 59 70 _____________________________________________
From: swift@acs.nntp-read.bu.edu (Matthew Swift) Newsgroups: comp.sys.next.bugs Subject: Bug in 3.0 Admin. dox, SCSI Disks file. Date: 1 May 94 13:10:41 Organization: Boston University Information Technology Boston, MA USA Distribution: comp Message-ID: <SWIFT.94May1131041@acs.nntp-read.bu.edu> In the SCSI Disk section of the Peripherals chapter, in the section on "Making a Mount Entry for a Hard Disk," we are told to: ---begin quote 4. Open /etc/fstab and modify it to contain an entry for the new disk. Find the line: /dev/sd0a / 4.3 rw,noquota,noauto 0 1 Copy that line, and edit the new line to read: /dev/sdna /mount_point 4.3 rw,noquota,noauto 0 2 where n is the device number of the new disk and mount_point is the path of the mount point you've chosen. --------end THERE SHOULD BE NO "noauto"! The example is wrong because the whole point of editing /etc/fstab in the first place is so that the rc.standard script will mount the new hard disk with the command: mount -vat 4.3 and "noauto" explicitly prevents that. The "noauto" is there for the root/startup disk because by the time rc.standard comes along, it's already been mounted. Matt Swift
From: eagle@catt.ncsu.edu (Daniel C. L'Hommedieu) Newsgroups: comp.sys.next.bugs Subject: Console bug...still? Date: 1 May 1994 18:25:40 GMT Organization: North Carolina State University Message-ID: <2q0s74$ghb@taco.cc.ncsu.edu> I'm using NextStep 2.1 and I've noticed a bug with the console. I have no access to a higher version to check to see if it's still a bug, unfortumately. I noticed that the file /tmp/console.log is read-write by everyone. That means that, when somebody writes me (as in "write eagle") and it gets sent to tty-console, then anybody can read /tmp/console.log to see it. So, every time I log in, I now type "chmod 600 /tmp/console.log"... ARGH! Daniel -- Daniel "eagle" L'Hommedieu Daniel_LHommedieu@nest.catt.ncsu.edu eagle@nest.catt.ncsu.edu (NeXT Mail preferred)
From: luomat@alleg.EDU (Timothy J. Luoma) Newsgroups: comp.sys.next.bugs Subject: WordPerfect (one of several) Date: 2 May 1994 15:09:08 -0500 Organization: UTexas Mail-to-News Gateway Sender: daemon@cs.utexas.edu Message-ID: <9405022008.AA03415@alleg.EDU> Ok, this is getting annoying. I'll be working on a paper in WP, and suddenly I will get a panel: _________________________________________________________ | automatic document backup | | warning backup file deleted due to disk error | |_______________________________________________________| So I try to save the file. but of course I can't. I get a Save As panel (even though I saved the file before) I select the same name as before, get the "Replace?" panel, select yes and get the _________________________________________________________ | This file has been overwritten by someone else | | Rename Overwrite Cancel | |_______________________________________________________| panel. If I select overwrite I get "error, file not found" and then : ________________________________ | could not save the document, | | press any key to continue | | NULL NULL OK | |_______________________________| if I select rename I get the save as panel, but anything I try to name it sends back the panel saying "file not found" (which of course, should be true, because the file is being created, so it should not be found.... but it never does get saved. The only thing I can do at this point is kill WordPerfect (because if I quit then all the backups will be deleted) or move/rename the backup files by hand before they are deleted. If I want to save what I have, I have to cut/copy it, quit WordPerfect and start it again, then paste it (well, I could also select all and do /services/mail/selection, but that is a little extreme) So I don't expect that there is anything I can do about this, but it bugged me.... Please don't suggest I buy this or that wordprocessor... I don't have the space or the money. WP is all we have here.... although I may have to start using Edit.app more often... TjL ps -- that's the extent of my ASCII drawing ability too... 8^} --- Timothy J. Luoma Email: luomat@alleg.edu (Shell=tcsh) Workstation Environment using NeXTSTEP 3.1 Motorola MAIL: NeXT YES / MIME Mail NO
Newsgroups: comp.sys.next.bugs From: macrae@pandora.geo.ucalgary.ca (Andrew MacRae) Subject: install -d - not implemented? Message-ID: <May2.202135.47076@acs.ucalgary.ca> Date: Mon, 2 May 1994 20:21:35 GMT Organization: The University of Calgary, Alberta, Canada Hi: In NextStep v.3.1, /usr/bin/install is implemented without the "-d" (i.e. create a directory) option. Does anyone know of a workaround? (I could, of course, create the directories by hand, but I am trying not to extensively modify makefiles.) If anyone has a replacement, or has any other suggestions, please let me knw. -Andrew macrae@pandora.geo.ucalgary.ca or: macrae@geo.ucalgary.ca
Newsgroups: comp.sys.next.bugs From: karl@trapac.com (Karl Kraft) Subject: Re: /bin/strip is an ugly beast.... Message-ID: <Cp6u55.I1@trapac.com> Sender: karl@trapac.com (Karl Kraft) Organization: Trans Pacific Container Service Corporation References: <2prves$8ml@apollo.west.oic.com> Date: Mon, 2 May 1994 18:31:53 GMT In comp.sys.next.bugs article <2prves$8ml@apollo.west.oic.com> Matt wrote: > In article <Cp14F1.1AI@trapac.com> karl@trapac.com writes: > :same for both. > : > :Let us say you are going to strip a binary called "mycli". When you > :strip this program, strip moves the file to "mycli.strip" It then > :reads this file to create a NEW executable called "mycli". > : > : > :So unless you wanted strip to be a suid program, (BAD IDEA!), it > :couldn't put suid bits on the new executable. > Nonsense, strip doesn't have to be suid in order to set the suid > bit when the user running strip owns the file in question! > But as I said, strip creates a NEW file. A non-privileged program cannot put the suid bit on a file it creates. Also, if you modify a suid executable, the suid bit is supposed to be removed automatically. Although in previous versions of the NeXT OS, the suid bit was not removed. -- Karl Kraft Karl_Kraft@trapac.com Karl_Kraft@ensuing.com [My opinions are my own] -- Karl Kraft Karl_Kraft@trapac.com Karl_Kraft@ensuing.com [My opinions are my own]
From: sanguish@digifix.com Newsgroups: comp.sys.next.announce,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.bugs,comp.sys.next.programmer,comp.sys.next.marketplace Subject: Quick Guide to NEXTSTEP information on the Internet Date: 3 May 1994 14:46:29 -0400 Organization: Next Announcements Message-ID: <2q6665$r65@digifix.digifix.com> This post is made weekly, to help 'point' users to more NEXTSTEP information Topics include: NEXTSTEP World Wide Web Product Information Server online NEXTSTEP Product Information Mail Server online comp.sys.next newsgroups related newsgroups comp.sys.next newsgroups mailing list ftp sites NeXTanswers Much of this information is also available using the World Wide Web, <http://digifix.digifix.com/index.html> Stepwise NEXTSTEP/OpenStep WWW Information Server online --------------------------------------------------------- A product directory built around the World Wide Web system, this will allow full multimedia announcements by NEXTSTEP developers, as well as the ability to browse the available products for NEXTSTEP. This service is online now, and can be reached at http://digifix.digifix.com/ it can be reached using OmniWeb (available from ftp.omnigroup.com) or Mosaic. The entries are coming in quite quickly, and I'll be moving the entire NEXTSTEP Third Party Catalog contents in as soon as they become available. NEXTSTEP Product Information Mail Server online --------------------------------------------------------- The NEXTSTEP Product Information Mail Server is now available for product literature and pricing from NEXTSTEP developers. You can get information on using the mail server at ns-products@digifix.com Suggestions or comments can be directed to me at sanguish@digifix.com comp.sys.next.* newsgroups -------------------------- Comp.Sys.Next.Advocacy This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. Comp.Sys.Next.Announce Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) The NEXTSTEP FAQs are posted here monthly as well. This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Comp.Sys.Next.Bugs A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT- specific groups as well. Comp.Sys.Next.Hardware Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. Comp.Sys.Next.Marketplace NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. Comp.Sys.Next.Misc For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! Comp.Sys.Next.Programmer Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. The NEXTSTEP programmer FAQs are posted here. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. Comp.Sys.Next.Software This is a place to talk about [third party] software products that run on NEXTSTEP systems. Comp.Sys.Next.Sysadmin Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. related Newsgroups ------------------ Comp.Soft-Sys.Nextstep Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. Comp.Lang.Objective-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. Comp.Object Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original Comp.Sys.Next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News ------------------------------------------- Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep-d next-advocacy-d next-announce-d next-bugs-d next-hardware-d next-marketplace-d next-misc-d next-programmer-d next-software-d next-sysadmin-d (For a full description, send mail saying LISTS to <digestif@antigone.com>). The subscription syntax is essentially the same as LISTSERV's. To subscribe, send a message to <digestif@antigone.com> saying: SUB Listname YourName Example: SUB next-hardware-d John Doe The ftp sites ------------- cs.orst.edu: The main site for North American submissions nova.cc.purdue.edu: Lots of older stuff, but very short on disk space ftp.informatik.uni-muenchen.de: In Germany. terra.stack.urc.tue.nl (Dutch NEXTSTEP User Group) and cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp.next.com: See the below ftp.next.com and NextAnswers@next.com ------------------------------------- From the document 1000_Help from ftp.next.com Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, or you can transfer them by anonymous ftp. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to NeXTanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO <your-address> If you have any problem with the system or suggestions for improvement, please send mail to NeXTanswers-request@NeXT.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to NeXTanswers-request@NeXT.com. Written by: Eric P. Scott eps@toaster.SFSU.EDU and Scott Anguish sanguish@digifix.com Additions from: Greg Anderson (Greg_Anderson@afs.com) and Michael Pizolato (Michael_Pizolato@afs.com)
From: eps@futon.SFSU.EDU (Eric P. Scott) Newsgroups: comp.sys.next.bugs Subject: SECURITY: NFS Mount Vulnerability (with patch) [1/2] Date: 3 May 1994 20:05:11 GMT Organization: San Francisco State University Message-ID: <2q6apn$pue@nic-nac.CSU.net> A serious security vulnerability exists in NEXTSTEP 3.2 and previous releases (going at least as far back as 2.1) that affects networked multiuser systems. This package contains a description of the problem along with source code for a patch to neutralize it. N.B.: I regret that this could only be verified on Motorola machines; the Intel system that had been loaned to us went away at the end of April, and we have no plans to replace it. 1) uudecode this message to extract nfsmntfix.tar.Z size 6061 BSD sum 10770 md5 cb3e5d5598a43fa6de81aa14986580b1 2) zcat nfsmntfix.tar|tar xvopBf - View the README file. -=EPS=- ------- begin 644 nfsmntfix.tar.Z M'YV0;LS,:>.&CIDT>%X`6,BPH<.'$"-*G$BQ8D40&&_4J`$"`$:,,FAT_`@B MY$B2*&%@W&@CQHP;,5K&^&@#1HP:'BWJW,FSI\^?0(,*'4JTJ-&C2)$&'%CP M8$(I18(0:5(DJ=6B-&F(]/C1)%>06U&"B`%CA@T;*VNTI`&C+(V9&=O&R'FU MKMV[>//JW<NWKQ,C4T`T>5.G(`@C"$&8>2,'A),B6*A,H5($"H@9+F0H4!`$ MQ)PR<M(0GN.YS)@ZH>GD`8'F#9LR(,K@23.'#NDT;AQ#EDS9,F89(,*X(0," M3FC&(.24>1WF\QP%*,Z\P7T&A)@P8]8$IP.">>W@I,V$:=S<<9G(M<O`*>DB M1@H7&)-P#\.&S9L[I(7G43#F31LXK]&QW&J%&9>&'6F\=D89Q-7Q66-TO`$" M;@+*809HL-V1!AUH%"?'&W"`YED>Z;6A`(B@A4&':&ZP$-QP((PA7&S-);B: M&+#)!N!T`A(7(0AOB$%'&+AY5D>(<C@H(@H?OD%'"@I@-T89<\SAPF94H$%; M<=BM$<:",;Y1$)%ND#8'87),&289L"U&GF=I_/<:"$R\$08988@QYQ*@N;&< M`E.`9H>('*H8IAS*C6$;"!S"9ER0K[5Q)99H./JAGF6TH<-F08S1GQQD4,>H MA(V"0,8;I[51QI@KBJD8<FT05A`*,J3`PF8H?1%K871\`4(;8=P(&VYVO+$& M@T"ZP<:-JY4ZQY&@M:"D')-6$1!C=!2FXH`N<KCEEFXXR6B29<`'0A*Y89<M M?1OFX:)^($P;XXQ_3=&"`KL:YFJ$Z[UAQHNK(3A'&IB:FD:B$<J11PLME#%H M;J$BS%@:5`)YAY_$B;%?DW0P+!QQX7+GYY15CF<C"'>DEN></Y8J8WV@F9LE M;&(NJ\`<(8Z1!GW!*<HBHY6"<*P<?K+!:!C'DG9'I;FY3-]KC8$K;J,W0YOD M@Q.2]J.;:I8J0KC/ID&&"+_*RAV(K;IA[A5:LJPE::=2>3._0.?8W&H4?DA& M'3Z[ZF]R;YQ1!VR?=5R'V!X&+D<8;<SAXH:FOE%QR`H8Y_"JW#TK1FT;UI$V MD/\>]!IIRMUIW6H&(J@@LH^2[-RM3]\GZF>#+FZT;+2MZ$9UCYZQ>..C6D=X M'6+4WN.[,(I`QX=J"*=`:W(X1_:/.'97AAG<N1EOF0[FK'MUTUK)*1L<$G9& MAZ7Z7L:VWW%8.G=;_C@'B0("K^*08U3J(ZE!GV'?=6Q0`+HJY"?NH``W8V!# M'4*U.Q!,@0E)L$SX;!6;A\%)0$K;4(=HXYIM$<<^+XM2Z4!0P#LP9@WY40X( M[%`'-OAI<9B:U!2FY`:3O6%3"F#-&S@W)['X\(<(\IF)5L"P(C)L!3],(D:( M:,0BKD`!?,#(#Z:($3[TH`<1#),;1K8B!*GFBGQX3&0F4QD^*(")342B$G^( MQC1N1@2MX6$9R+:E,ZGJ5]C1DI^2A3(MY:]N$RI(DMC4IS!LB-.&&!0E M3U')2AA!@7RRYH83B,PTCS39LBJXJA<YJ5*-J8(3DH`%/.8/-V7PF`*@4(8R M#*&5R2F,&T35!'L1X0F!T5!](N>B,QRHDX7*GG+*<(?QP*9>I9%#[6[F+SH4 M4X5;TA";-AD&.Q")#2LK5PHP(H(@KJ@-=,R/;L;8&U/J$3814L#2TO#'4N5M M@2(Z`YGRX\@J@0=@VTM=@LJPH`8]:%(S"^3RX-F8,;RO8F$PF,04QD<X[G!% MKR%;?\:$&^K@ZPVA.@BR:C<P,9'F;Z*KF*NZR<YOLH!L'XL-'N"`+:T9D@X* MB)AI$K::TI&A!35K%O\RQ!@VD$%F3'/12+T9)Q'@,`$O*`,=QO""?'&G!;EI M@<@$`@*<`HX%9\(>"\(5ASHX*:%QA&@9=/"".:0)!"\@0VTVLS0,H=6L8\@: M('6$+63)5%&,6<U(P[I/E,(HK=^AC0)*=5>:-I2HX.SC#F$S40$5)(7HE%#U MGH6DS[#IIPH(*%2!!(?/_<BI'KK<8X.3'-,PACA_*Q5HS?"A-@C/HBG;D&.# ME]2E-G5(8F"8:`$YOQ(%APQMJ&AM%I>PPY:TJ)6[U,I68\)*<J=ZV0R>@'89 M3$""%G>U@20(GN"&KE5*`>\D)&O*$Z0AH9(XE`6-O*3TR,?!3YP#DQ-L@$6' M"HWJ9I>#$,4RR"$0.%2.?B6.<F(UJ.#L$H&A"6X-OU=!L:UJ2I/*;-"<@L.E M$,0@B=E20NMTI^CRB6C+<:"@"/72T#YL4<AT*FDTIH`"A49U_$36RZ"FW4GJ MTF@A$RAH+J1"#?4W7'(`EM%Z6[\3(4E%+'(1D%&F0?-$Y@2D0688/.<?L\4, M2WD($4:`=2R,4.]P/@4D$Y80F,*(5XR\J<QEV@,"%,!K/*<4D*)04P8H(0?- M9/1-9MH\P!#G]#T@L):;LK5@;MT3#N,YV[^<<!XJ0)D_8M+=X1IXO4$'[UCJ MJ=O!8N2@"*DJ>A,Z`Y"%,Q\858H-`&KO8"MU(\_E#T/$V6VL5/A(S.W,:)6Z MDXA"RE_T!>TW"I#"<M;W&>1E[$-=QAD;7-`A[0TF81U,G.L8<YNF!:UUC[P2 MHQDT.JU):`[!ZA8HRP#EX(B!,-Q1@1Y`\P85@(`*0X#""R+8`N5@LT="ZU.( M^\.F51LJ;J3),9M"!*-]O4TQ^PQ!)+-`&,^TIH5D<"X)6TF<6:/3D&Z(DF*( M^2O<>*YB6TN#X%2(;AW>(7C89I"+Q."Y"?TK#X0Y@0IY&,#:\&T-*Z]#=1@= M&7.7''(N"/J+B+/..<&\#C*'C:$V]*XY:*=1KEW:?AJU&F/&R)!R\&4#Q2!R M$)#@Z\';[40')MY$IF>6#?S1:6KC']!`6DYDLHV+#%E@,='L7Z4R8?2XDZC3 M5AN0N-G;<*O^'/&0AW3#KE$#$;W42H4'5F%0`W(^<YK4!*R%+\Q3@C:TWULU MUY)&@@-+Y5#J/'BK@9\D\8Q<AAKE&$;!R!&V=Y0^G%73;(]Q4E6HMK5)`S'& MV'CL,N2>9IV6QXFE52+8G-S$GTZW'=3TL2GJ=NB<@:4]:,KI*FT6&>@'OD`( M4R#""X;0A"IX!DU34D`B:>B<X)673(#D.17R#>(AHQ_D)38AQ`ULGY-#?;`2 MXA1U\QGG1WH?]2^/@BG``W6?P0:#4FWI83I_4TR<UT#,9T=E8'>QP08$^",4 M,F)^X@)0H@!&<&>[D6=K!@,ND@3"DTB7`P)'=P*[Y"?(\ED8E09FT"Q!TP1( M<SW[=%]W0APM0`1.@`5?,`1/T`10D`1,4`12\`50T81!,`5%\`4S\`4J\2-# M8`1,$`1',`634H*-X02!(0,N@`<NHB1;UH,D-`?`$4@1B%K_PF7E@B42LBK/ MHD*E(H`:)GHN&&MYQ"AQ<G&$0S^9(E?GYB0J]UMDP$RJ<GK5D4C'LADX"`); M4%7_\@)!YP(O8&%-@1!?8&^H`@)=L`-UDW%(Y2!R4%M,M09L\`6>DR!5E5": M&'2=*!`7YA2AN!RHD@`^D%8.\P)CYQJP(0,^8`(QH``'L1D_PHHOD"8N`$+T MX2+*`2#8(2JUR(E,UE][V$)&PWCHLU,*P'4UI##FX@2*"#2&<G3SDAMN!B-S M\'`^M4WV83K.,F)14Q#B6'FN1P>;E"1;-#N&Z%KE<0?+P0:F.".YQR`[(R"E ME8CO14).H@#9=SC*@5F;402A$5=0`!\SY"0P103JD6BJHB__,@3^`0<?UQ@S M1#'=50:`,B-&L#C=11O]X4!#XI#6\DO1TRX*$!-M`0)(X!K]%RP@$`0/,S@Q MF1LS*1PZ,P?]X2)#T!D@D`-O,0.:H0`\N!HSX"(QD`-6J0!],99D699F>984 MX8D8EA`\>"PAA99YD15AT15S"18G@1(NP19IT1)F00,R<`,BD1$U0`,W0!=P M>9B(F9B*N9B,V1!.$`14T0-JZ10*L(5,6`13T`,D@`*/217O,0:`4@1#0`5& M<)F9R6%D<(3^`2S#8264QQW64H^IB9)MP)IJY0*OJ0!/4)I-F)F;:9F]J0,N M,`8]X`)O`"4C``)'\`5,\`12$01"T(0]T`+$L9F=600I\`6RF4T*L)S-^9S1 M600]`"A2,`29V98^.">_R9N8N4V;286CR9Y3`"5)0(94$`1,P`1$D`12T`,O MH(HO((UL$*`$\P)#XP;::2>$]`+6"9G8N1E/LRD)``1TZ'6<Z:#9*8IQQ04* MD``)L(5=^(4](`)#N`1.^!A,4%53$01#@`1?4`54^(1!L(0JV@0LB@15]03` MT0)7\#0BL!EE(@,22J%LV*">N8L@!`(<ZJ$@ZH692:)$8*)2@*(J*J4H^@5& M$!544`50$1A#:*,MFJ-5U:/U\:,*8*38B:2HH@,6NINFZ9XH`)^D^:8=ZHK: M*82Y@:;;U`)&@YJJ69L?XYHS555U$&C*HJ!_:IN"JBA*VJ$)T`)I`(=#\I(6 MZIW.2030V81[*B$D``1M*I]0XJ'@529#4A]LJJ=J&E<DL`?U.1GXJ9_\V0<= MBANU07PM$%<MX%HVH!55)2$<4U75T58AAJH:ZG6L:I^ONI]2(*MG>JRNFI_* MV@<2V@9K$#&DT:D=RA_,X0826F^NU0+_0JR\J*H7>J3;63!Z:IQBV9CLVJ[N M^JX0,9D(<8M,L9;#":]`(9=W61)UR:_[VA4WT!)[Z1(V,)@RP!%C`0-_:0.& MB:\.^[`0&[%E^0+N)J]XP*;*,5PEY1B`43:\XFVAIUY8DU,@H`(OH``4"P(: MR4X@T)$.U!_UY2)3(),T^90W.04Y"1L[R5'MXB);.19@*1(FB[+N9@7LXB-H M\"$ZUR%X5DZ_T50V@`/:@0),(')H0`=H=P8U,)\E>[)$ZQF>0G:QT08X0@:7 MU8+WT69G@BAYL$U#JP`C<'S8`@(\<*`/8@<&VB=?<+>@\06J$2)6@@8^`+=R M2WIT:[>"DK=$L[?XN(L`D@?,-KAQ:P9LLFA&B(1*>)E/&(510856B(6$^Q]S M6[=]<K<O\"@1\K=4$KE?*QUFRSBY(7_6@6R$`R#,%DAK)@-0*[4AT+M=&[JC MQQVD2S2F"RRGY"=E9;N""[RCNSED4%8D\@)V(!"L&[>B:[@\X+S0.P=)A2CA M4KW+\1G,B[U^@@=TP"9XZU,_M;S6&[QT:[SGE+S+QKZ%*[R]);W42[_7:[_1 M"QI$\P;@.QPXN!FR04">,5",.KUG`@<!)Q!?H,`@,@<[H``%W"<"14(.[%0H M,%Q\PQT*7+(NDC_C`<+S8K9RX+<I,,$*<*",JTQ]"S^DZI1EH,(4@G!XP"!? ML!2Z8C8;C,`>3%4JH,!P$,)85[+@&,)!>,+<@2=#`B5[,*O_@@)"W`(^H,"P M*#8AT`,P8`(F$`+/\B`HD`+;I!S90C0J"P5.V`03G`!DC!JYH<,:+,0N<L2F MHB)AD,(*P*S%@C@6^P4510=2/!UDX,0Y]!'*X4OID8_P`R`J3!*5B`(Z#,$, M[`)6[%17!,=FLTU/G`!I`"`]H&Q0J3]AO,:1+!`13,D9;#8]X!0XC,F\,L&3 M.QS7XQB7FX1+V(2;6P12Z+E7"`,=&LAB@YRKDE$=JFQX<$``@L<)`,R#;"!- M@0(BD&)F8\.2^C0,P@5N(`)X3!*R2A+A"QO,G`(L?+=?@&BS-`8H,,Z"\@40 M.5QA``?EC"TH8`*T.JE38BL=<=NUH9%L/U@62N8@1!<)E$$`)<``,'S:%` M@`(CD`(6.YP)$`,NH!)Q"AA5L$TU\`*Z:Y7:/,%Z+,@87*^Z.#"BYB]F$,X* M\,0D<<BY(R(UW,D(B2L?\<BEO,!64LFJO,H(T<JIS"N:/*N>#,JO1@:CW*$U M?<HXS2N2V=,%`<LX*,N6>X2VK+E0J,N=6X6]_,M[/,AP.\P#G`#&C,QLH,SA M[,P&`<W2S"O4S";U;,UD@,T=7<@8(:NRNA,6^P)^2IN*BILS);$0H:]?X:^! M[15)9!8#"P,U<`,W\)<8$0,Q0`-8V;!^/=F47=F671%3F9\>VL=_[*&^[*I2 M0`4*T`_0```!\`$J```,8-HNT!`!``H0H-H`(`!!QQ"GO0(`$-L&`!52016V M_0$L```4```(@)YOZ=H?``/"#0`+\-!CL!"G/1<<```2D->K&:A\K2B_+0,# M0`G"'9N(JM?8_9J77=[F?=Z->=?@?2>).MY]7=F`31*$39?_2A(S,`.'O1%^ MB19C$0,!2P.2C=X"/N`$?I:9S02;C8N?B`=[*W+A8@9FX-D*``-```"D;=JH M+=L!T-JV#=NR3=L<CN&XK=N\/155`=W`O=S%W8/'_=O*/=S-K>#V^MP8+MW4 M;=V`VIK930?;W=W??:CL+=XZ3MX%7N1&?N1(GN1*ON1,WN1._N10'N52/N!S M<W,YMW.-YG/&QQV;.'1]](,QJ$)+1P=-]W25$G5H,'6L%APJ),)9)RI<5QU? M1P)A9T%CYV"-87:.)2IJYWR?]G;62"&.,UYUMT>I%31Z]QU]!RI_YT[#T6D, MU1S*.,+E86_$)BJ,]VJ/UQC`(GF-07FHT2XKA'DIPG6)M")4XGEB`GJ4Y;[Z M`8E`TB@0@G&`Q(^8TW%+)GO$1GN-6"H:N)"[-UVH<QQR`'P5.GR[Q'+P([K) M5S#,MW:>!AKYP0;25QS4]TA\CGUEH'T#HY/>!W[B1W[F=R:HD7[KAX<@)R'O M5R2E(KL'&F+DGB;X9RCZ%V:QXW]G#H#4W"@$N+:+\C<)&"DNE2D-^("2NCYR MB#)$PF`7V'8:^,W!\X$N'((C*(;CE&9ZIH+GTH(44V`Q.(,4%SRQDE$Z"!O& M_8/I%(0J6H12G;FX7-6[C-58&#Q-^H5A>&=E>(9I2(#!!QM!BA'UC/"@T_,` M=8?<HX<3EF'YX8<4`XCM-(B\-9"(^`:*2`;O8K:.J%1Z%(EI,(G@]2^7"*YH MM8GTFHN@6*RE>(H=^I]FY8QV&HM&TP*T2/9]K*&^"(QX.XQS8HS(J(QIP(P2 MXHS0**#4J![8I#,-E(TOL(V^UB;>R"7<*([D.!Z0BQ'HZ)#5Q8XRXHXI%8^$ M,8_=H:"\A8\"M8^MASG_*$L"Z5L%>9`)F1O`WI"PH1P0*5>44Y$'PR`1MK(< MZ9$P&Y(C27HEJ6@@0)LJ:5\M^6`P.;-,6;,V*2$XNRV&RI/=OA]`J1)#61_W M891(N2I*R?R'X?Q0^0922956Z1)9^;-=";1A.>7N/^!W??*O<>3Q3=^#W:\? MD9<J`0(L0;!^"9@?02,0I@#W_@J@`817UVFIB32$4)GDDV8J5]AI.(6F^&2: M>@".VVNOR5#)IO8VY&:*;G*`Z\DT"2?B9)R0DW)B3I<J4XDGZF2AKE-V.E>O MH3NA0/`DG<B3>>H!\@\VA,#>!*?D%*@2(,@*6O$G_P2@!!2!$@.*"T'))M#` MH"#@"(I0'8I(=1EQE:26U(?B0DYJ1)6H$Z7+:M2->E$QZ@O,*!;TI6Y4CMI1 M9(H-F*D@-:0J%!4<15:PYCVI+3BENB`=O%)9*@ALJ2[E!<-4"W@"8\I';88W M.`9.%0IP4SS00OE`.I4`[%28V2QZJJKTJ?!VW3H@HY(6&I`2YKC;E`&M(*2J M9I1J,UDJ&HB=>I77\52;"1&V)T<UJFJ5J6J!&"I5&:M6=9^$X+*:5?XL[N$J M7<6KK,JO:@'!JE(,*R<H"U<5+4Q6L6H0.JM:"*N6U;2J5@?C6@&!;)5`U@>W MZE#>"A/!PB-5K%`5#-2!3E!='<!QZ.3NFG,K</7/+MV_^E82`A9<Z']FP6`A M++*PL`@@.;R'^'!B52P91V%*B\9B5&EMM,@/JR$O2-;02EF\KV7Y/I`DLVB6 M4WI^.$GZ[2QIU[,$@U'Z2E;I=Z4LHY5(D);2.A\7#P4]K380M:96U3H?6(LZ M;"VWY;52%E0:`V(K4Y2MLX4C^H_:8@P*@R5NAO8UNA"7,DF"+:QV^*TLL[KT ME_L:7F4B<:DS%W;"JM&RJ%XXJ'+1LI9WRYP0S+MJG\N7\43LY1/Q%NIR$D0Q M<`VNE.6Z\`1!<#+S1PS0+L^@O'#72"R)O4OA#*VM*+RZ8E/)(ZA$?D'%_46W ?M-?]4F![\2CZQ?[EO0`8^_IFXTMXE:_S%8S45V!\`[RZ ` end
From: eps@futon.SFSU.EDU (Eric P. Scott) Newsgroups: comp.sys.next.bugs Subject: SECURITY: NFS Mount Vulnerability (with patch) [2/2] Date: 3 May 1994 20:06:06 GMT Organization: San Francisco State University Message-ID: <2q6are$pvo@nic-nac.CSU.net> This is a nfsmntfix_reloc for NEXTSTEP 3.2 for Motorola Processors (Lightning5S). If at all possible, you should compile the source instead (the general rule that you should avoid binaries from untrustable media ... especially something like this!). In no case should you use this version if it doesn't match the checksum information given below. 1) uudecode this message to extract nfsmntfix_reloc.Z size 1410 BSD sum 11631 md5 833bb6ec889a2804259655afdd8e4655 2) uncompress nfsmntfix_reloc -=EPS=- ------- begin 644 nfsmntfix_reloc.Z M'YV0_MKI<P:@H(&"``(@++"PX`!@"",J3$C@RQ<J1;!0B<BQ(P`0"`=P*P@2 MX8&&#B-:I%,&#QV/""UBU`BS)BZ'(PM.')(3`(2:,;^,>>-FSDN8,C-N!%KP M9D$D!0E00BB`:4V+8XS*2>/FC,>D-)E.+$@C:D^K58&.!6:12!`J0:R6M%KP M9%VJ""M^(1.&3AB@;=_&93HWXHB>:>ERM"AFSARF@>'*12NV<D)&3-Z$(5.& M#(@I9>38"<UQ[MS#)"/:!;":8530HD.#<!*F31F.F3=W_AQZM)S2$1,41*VX M9F[/0]ZT:1/	G'!8_OANV[M/""'H:7*UZSBALVFI$K9^[\L73/U$E'!'$" MX8:")`9PAYFDJ%\W8\J`L!+F-\+SO,7F'T(@"($0`O!E-U]'Z<FQ7VASI$%4 M=.%-UYMZ!"J!T$0DF+!@38EA4!`17"`D(@!%\%&0`%"@!``6&R(DP8<TUFB5 M$U84]((")+@!``)*J!$`"6<X$,8_2/[CCA*`F$$`'`&``<$++OA$)0`,['AD MDM;T^`\_3GCA1!TX%A1$/A$Q`$E$9G@`!#!&\#/"1[Q$!$0P1@"#1!XZ`;`E MDJ"`H<&>".%"*$)__L.("X`D2DB88Y8)P)EIKHF0&1R\&>><(-2)T)UY'AI` M!HE*`2F9<3AAQ!0@M/%&'6[0`8(9:>`!`E=&A<$&&WU)Z`8(1@21!!-%$!%" M04"@,$(*;I@Q1QNQTHJ'"V,D$(,+,("`PA2K5I$""#6\(,,+.915D*JLN@JK MK-+>:I^N;'2F``#HMOIJK+/6"@)GN/JUJ[PV!BSPP`07;/#!"">L\,(,-^SP MPQ!'+/'$%%=L\<489ZSQQAQW[/''((<L\L@DEVSRR2BGK/+*++?L\LLPQRSS MS#37;//-..>L\\X\]^SSST`'+?301!=M]-%()ZWTTDPW[?334$<M]=145VWU MU5AGK?767'?M]==@ARWVV&27;?;9:'?=[+/1UCI$$$PPD4`":T-+A[1?<)4& M'7/#H,`4<$E!A0)OQSUWW6WC\46$9[CQAAEF]*U`O\WE)T/:F&>N^>:<=^[Y MYZ"'+OKHI)=N^NFHIZ[ZZJRW[OKKL,<N^^RTUV[[[;CGKOONO/?N^^_`!R_\ M\,07;_SQR">O_/),N5)0`=`@I.*/T1=TU`'5`P#&\]FKL&+V,R:4/2H.`8&0 M)RNV6)"E`9A?D")1J0^`CP`8(#\:*[H/@!;E(Z1^^P@Q$`!8A!`>Q$\G+'C` M6#@2@!XLD(%$>*!$GJ!`CMA`)U^0X(;6H$&=S$&"!-!)'SJ8D$N0,`"HJ&!$ M"*&378!0)]-X84*P(<,`A*.&YZAA/"0(D83HXX$!^,D`"R!#`3"@B!PH8@N* M6`,)ND\`/G#BBHX@Q:](JS-?6-L7U!4K`"RN#G,@31:=M<5[O62,;+M;K?+F MAKUY48MV<-8;X/"8+\!A*]'R(N7P<YLOK"$T;EA<;[X@AC>\@0Y:"0,<[/@& M.9SQCW((9!A%8\?FI&$,;W26W?#&.,=!SHNV<94<\K#%YH3A#*'92QG,$(8Z ML.&,8JB#&;P82\A!Z(VUI*4LX7#*,M2QEF@(PQSP1T@SR*$,96!#&HSBQ3G< M(9>+>Z8LOV!,9"J3F=&LI5#BU9QKGK$HTISE%^A`!S2DX0Y]\2(YP8/.,Y(3 M#6_P2D&H62LL:I&+9YP#&,5X3S-F,HUXT]LWR1C'.<RQCG?DRMWT^"X^>A&2 MDAQD(0^9R$7"H9&/!*0@*<G+-F(2C9M<8R<?)\Y0-I*4Y.FE'%3)2E?"4I:Z MM*4<ZN@&:-:2EZC\I2R#.4Q:5C.9R\QG.)L93FH>$ZC8=*8VQ\#-[P3UC4J% /Z3C+><YT3I6=5GUG/!$" ` end
From: stu1d44@bnr.ca (Gregory Gee) Newsgroups: comp.sys.next.bugs Subject: Problem with DUPLICATE in menu. Date: 3 May 1994 21:43:08 GMT Organization: Bell-Northern Research Ltd. Distribution: world Message-ID: <2q6ghc$rg9@bmerha64.bnr.ca> I finally had the oportunity to upgrade to NS3.2 recently and am now finding my first bugs. The bug I found was when I selected a file in the file viewer and slected DUPLICATE in the menu. I was copying an app and the duplicate copy somehow got corrupted. When I try to invoke it, I get a message saying that it is damaged. Any suggestion? ******** Gregory Gee Phone : x3-9768 Dept. 1D32 Ottawa, Ontario, Canada BELL/NORTHERN RESEARCH Meriline Tower 7, Floor 4
Newsgroups: comp.sys.next.bugs From: alex@cs.umd.edu (Alex Blakemore) Subject: Re: /bin/strip is an ugly beast.... Message-ID: <Cp8uFK.6rq@genoa.com> Sender: alex@genoa.com (Alex Blakemore) Organization: Genoa Software Systems References: <Cp6u55.I1@trapac.com> Date: Tue, 3 May 1994 20:33:19 GMT Karl Kraft writes > A non-privileged program cannot put the suid bit on a file it creates. What gives you this idea? -rwxr-xr-x 1 root wheel 16384 Sep 1 1993 /bin/chmod ^^^^^ note chmod is not privileged, in that it is not suid yet it can still set the uid bit for a file. % ls -lg testfile -rwxr--r-- 1 alex genoa 0 May 3 16:29 testfile % chmod u+s testfile % ls -lg testfile -rwsr--r-- 1 alex genoa 0 May 3 16:29 testfile perhaps, you mean a non-privileged program cannot change the file owner to root? that's true enough. -- Alex Blakemore alex@cs.umd.edu NeXT mail accepted
Newsgroups: comp.sys.next.bugs From: rfi@jeans.fokus.gmd.de (Robert Fischer) Subject: Re: Problem with DUPLICATE in menu. Message-ID: <1994May4.070051.2356@fokus.gmd.de> Sender: news@fokus.gmd.de (News system) Organization: GMD-Fokus References: <2q6ghc$rg9@bmerha64.bnr.ca> Date: Wed, 4 May 1994 07:00:51 GMT In article <2q6ghc$rg9@bmerha64.bnr.ca> stu1d44@bnr.ca (Gregory Gee) writes: > > I finally had the oportunity to upgrade to NS3.2 recently and am > now finding my first bugs. The bug I found was when I selected a file > in the file viewer and slected DUPLICATE in the menu. I was copying > an app and the duplicate copy somehow got corrupted. When I try > to invoke it, I get a message saying that it is damaged. > > Any suggestion? > Hi, I think that this error occurs only with an app-wrapper, because: - Duplicate changes the folder name to 'CopyOf...' - The name of the inner files will not be changed - The app wrapper relies on the name of the folder in relation to the binary inside the wrapper: 'BlaBla.app/BlaBla'. Workaround: rename the binary inside the folder or the folder itself. Hope that helps, Robert. ------------------------------------------------------------------------ Robert Fischer @ GMD-Fokus -------- __o ------- _`\<,_ fischer@fokus.gmd.de ------- (*)/ (*) ## NeXT-Mail welcome ## ------------------------------------------------------------------------
From: stu1d44@bnr.ca (Victor Koa) Newsgroups: comp.sys.next.bugs Subject: cmsg cancel <2q6ghc$rg9@bmerha64.bnr.ca> Control: cancel <2q6ghc$rg9@bmerha64.bnr.ca> Date: 4 May 1994 13:27:53 GMT Organization: Bell-Northern Research Distribution: world Message-ID: <2q87sp$inm@bmerha64.bnr.ca> References: <2q6ghc$rg9@bmerha64.bnr.ca>
From: mcquill@next.duq.edu (Tod McQuillin) Newsgroups: comp.sys.next.bugs Subject: FAX problems Date: 4 May 1994 11:02:29 -0400 Organization: Duquesne University, Pittsburgh, PA USA Message-ID: <2q8de5$e3@keen.ccit.duq.edu> A problem has cropped up in the last couple of months that I have never encountered before. There are some destinations to which I attempt to send a fax which yield the following message in my fax log: Unsupported (local) and the fax does not go through. I've looked everywhere I can think off to find an interpretation of this message to no avail. Can you shed any light on this? BTW my fax modem is a DoveFax. Thanks in advance for any help you can provide.
From: anstine@uniblab.sas.upenn.edu (David R. Anstine) Newsgroups: comp.sys.next.bugs Subject: cmsg cancel <2q8dk8$ca2@netnews.upenn.edu> Control: cancel <2q8dk8$ca2@netnews.upenn.edu> Date: 4 May 1994 15:08:13 GMT Organization: University of Pennsylvania Message-ID: <2q8dot$ceu@netnews.upenn.edu> References: <2q6apn$pue@nic-nac.CSU.net> <2q8dk8$ca2@netnews.upenn.edu> Originator: anstine@uniblab.sas.upenn.edu <2q8dk8$ca2@netnews.upenn.edu> was cancelled from within rn. -- ____________________________________________________
From: gt6963c@prism.gatech.edu (John) Newsgroups: comp.sys.next.misc,comp.sys.next.bugs,comp.sys.next.sysadmin Subject: Problems wih CDPlayer.app Date: 4 May 1994 13:41:34 -0400 Organization: Georgia Institute of Technology Message-ID: <2q8moe$nc5@acme.gatech.edu> Well, at first I thought this might be related to the fact that I didn't have an audio cable between my cdrom and my PAS-basic..but that's not the case. For a while I couldn't get audio cd play out of any of my 3 OS's (Nextstep, Dos, OS/2). But, I've fixed it on OS/2 (it needed the audio cable), and Dos/windows (needed both the cable, and the scsi drivers weren't loading propperly)..but now i'm still having problems wiht Nextstep. I have a main user account (kzin) and some lesser project accounts (that aren't relevent), plus root. I have NS/Intel 3.2 on my Gateway 2000, 16mb ram, 400mb disk space for NS, an Adaptek 1542c, a Sony CDU561 cdrom, and a Pro-Audio 16 basic. Oh, and an internal modem that is using the MUX serial driver that was posted here. (The CD rom is id#2 on the scsi bus, and it's the only thing on the bus besides the host adaptor) For the whole time, I've had kzin set the public sound server option. Last night I figured maybe the problem was that root needed to set it as well. So I logged in as root, set the option. Then I inserted the audio cd.. the console showed it probing for the filesystem and mounting it.. and I got the cdplayer app..for the first time ever.. So I clicked to play the cd.. NO SOUND. I can hook up headphones to the drive and listen, but there's no sound comming out of my speakers (from the sound card). So I went back to my kzin account... put the cd in..and it started to probe for the cd , but after it mounted it as /cdaudio, it said: "cdaudio.util Cannot open CDPlayer" and ejected the cd. Same error I always get. I've tried having CDPlayer be setuid root, or not. I've tried a LOT of things..none of which seem to work. Also, even when root gets cdplayer.app to run, there is no such mount as /cdaudio (which is where the console informs me that it has placed the cd). So..what's the deal here? If it were a hardware problem, it wouldn't work on the other operating systems.. so it's got to be a software problem. Has anyone else gotten a setup like this to work propperly? (oh, in case it matters, I have deleted the "me" account, and its files). Thanks for any help, John -- John "Kzin" Rudd gt6963c@prism.getech.edu (ex-kzin@cc.gatech.edu) Ascii Mail only (no Nextmail) (ex-kzin@ucscb.ucsc.edu) ========Intel: Putting the backward in backward compatable.===============
Newsgroups: comp.sys.next.bugs From: macrae@pandora.geo.ucalgary.ca (Andrew MacRae) Subject: Re: install -d - not implemented? Message-ID: <May4.183136.60237@acs.ucalgary.ca> Date: Wed, 4 May 1994 18:31:36 GMT References: <May2.202135.47076@acs.ucalgary.ca> Organization: The University of Calgary, Alberta, Canada In article <May2.202135.47076@acs.ucalgary.ca> macrae@pandora.geo.ucalgary.ca (Andrew MacRae) writes: > Hi: > > In NextStep v.3.1, /usr/bin/install is implemented without the > "-d" (i.e. create a directory) option. Does anyone know of a >workaround? For the benefit of others... Matthew Seaman (Matthew.Seaman@dyson.ox.ac.uk) suggested the GNU "fileutils" package, which includes a replacement "install". It works great (it does not support all the strip options of NeXT's install, but I did not need them anyway). A suggestion to NeXT: either ditch the shell script in /usr/bin/install, and replace it with GNU install extended to support all the strip options; or fix the shell script to support all the standard install options. Thanks, -Andrew macrae@pandora.geo.ucalgary.ca or: macrae@geo.ucalgary.ca
From: tms@cfc.com (Todd M. Swan) Newsgroups: comp.sys.next.bugs Subject: Re: /bin/strip is an ugly beast.... Date: 4 May 94 09:55:01 Organization: Chrysler Financial, MIS, Center Line, MI Message-ID: <TMS.94May4095501@litespeed.cfc.com> References: <Cp6u55.I1@trapac.com> <Cp8uFK.6rq@genoa.com> In-reply-to: alex@cs.umd.edu's message of Tue, 3 May 1994 20:33:19 GMT In article <Cp8uFK.6rq@genoa.com>, alex@cs.umd.edu (Alex Blakemore) writes: >Karl Kraft writes >> A non-privileged program cannot put the suid bit on a file it creates. >What gives you this idea? >-rwxr-xr-x 1 root wheel 16384 Sep 1 1993 /bin/chmod >^^^^^ > note chmod is not privileged, in that it is not suid >yet it can still set the uid bit for a file. >% ls -lg testfile >-rwxr--r-- 1 alex genoa 0 May 3 16:29 testfile >% chmod u+s testfile >% ls -lg testfile >-rwsr--r-- 1 alex genoa 0 May 3 16:29 testfile >perhaps, you mean a non-privileged program cannot change the file >owner to root? that's true enough. Alex, I think the key phrase up there is "on a file it creates." In your example chmod didn't create testfile. Todd -- _____ __ __ ____ |_ _| \/ | ___| Todd M. Swan - tms@cfc.com - CFC, Center Line, MI | | | |__ | The flame at the heart of a pawnbroker's diamond |_| |_|\/|_|____| is a cold fire. - Rush
Newsgroups: comp.sys.next.bugs From: alex@cs.umd.edu (Alex Blakemore) Subject: Re: /bin/strip is an ugly beast.... Message-ID: <CpB5zC.8pp@genoa.com> Sender: alex@genoa.com (Alex Blakemore) Organization: Genoa Software Systems References: <TMS.94May4095501@litespeed.cfc.com> Date: Thu, 5 May 1994 02:38:00 GMT Todd M. Swan writes > I think the key phrase up there is "on a file it creates." In your > example chmod didn't create testfile. quite right. sorry to muddy the waters. -- Alex Blakemore alex@cs.umd.edu NeXT mail accepted
Newsgroups: comp.sys.next.bugs From: karl@trapac.com (Karl Kraft) Subject: Re: SECURITY: NFS Mount Vulnerability (with patch) [1/2] Message-ID: <CpAKr6.160@trapac.com> Sender: karl@trapac.com (Karl Kraft) Organization: Trans Pacific Container Service Corporation References: <2q6apn$pue@nic-nac.CSU.net> Date: Wed, 4 May 1994 18:59:29 GMT In article <2q6apn$pue@nic-nac.CSU.net> eps@futon.SFSU.EDU (Eric P. Scott) writes: > A serious security vulnerability exists in NEXTSTEP 3.2 > and previous releases (going at least as far back as 2.1) > that affects networked multiuser systems. > > This package contains a description of the problem along with > source code for a patch to neutralize it. > Eric, I hope you realize that some freeware, some NeXT software, and some future commercial software depend on this feature. It's introduction was deliberate, it is not accidental... Also, one does not need to be networked for this to be an issue. The same machine that mounts the NFS volume can export one to itself, or for that matter, intercept the RPC requests and feed a logical volume instead of a physical one...... And yes, it does date back to 2.x, because that is when the first NeXT program to use this feature became available. -- Karl Kraft Karl_Kraft@trapac.com Karl_Kraft@ensuing.com [My opinions are my own]
Newsgroups: comp.sys.next.bugs From: timo@swabsib.stgt.sub.org (Timo Hennerich) Subject: multiple forward:: problem Message-ID: <1994May5.161413.12091@swabsib.stgt.sub.org> Keywords: foward performv Organization: Swabian Siberian Software Solutions Date: Thu, 5 May 1994 16:14:13 GMT The documentation says: forward:: - forward:(SEL)aSelector :(marg_list)argFrame Implemented by subclasses to forward messages to other objects. When an object is sent an aSelector message, and the run-time system can't find an implementation of the method for the receiving object, it sends the object a forward:: message to give it an opportunity to delegate the message to another receiver. That's true. But the following sentence is a bug: (If the delegated receiver can't respond to the message either, it too will be given a chance to forward it.) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ If in a forward:: method a message is resent via a [delegate performv:aSelector :argFrame]; and the delegate doesn't respond to aSelector, it has no possibility to forward it a second time. It just gets a doesNotRecognize: message and quits 8-( A possible workaround is (if the delegate uses a forward:: method, too): if([delegate respondsTo:aSelector]) [delegate performv:aSelector :argFrame]; else [delegate forward:aSelector :argFrame]; Now, is this a bug in performv (== objc_msgSendv?) or only a misleading documentation? Greetings T.Hennerich -- || \\__/ '> \_ _/ Timo Hennerich, +49 711 7353788, timo@swabsib.stgt.sub.org __][__
From: eps@futon.SFSU.EDU (Eric P. Scott) Newsgroups: comp.sys.next.bugs Subject: Re: SECURITY: NFS Mount Vulnerability (with patch) [1/2] Date: 6 May 1994 05:41:02 GMT Organization: San Francisco State University Message-ID: <2qcl9e$mde@nic-nac.CSU.net> References: <2q6apn$pue@nic-nac.CSU.net> <CpAKr6.160@trapac.com> In article <CpAKr6.160@trapac.com> karl@trapac.com (Karl Kraft) writes: > I hope you realize that some freeware, some NeXT software, >and some future commercial software depend on this feature. It's >introduction was deliberate, it is not accidental... Why isn't there a check for the mount point (or parent, in the case of a virtual mount) being owned by/writable to the caller? Why doesn't mount implement the "nosub" option? Actually, I'm not convinced there's *any* way to make direct non-root mounts safe, nor do I believe there's any real need. [Hmm... I wonder if it would be helpful to extend NXWorkspaceRequestProtocol?] As for "future commercial software," there's no reason it can't include a component that runs with superuser privileges and provides reasonable sanity checking. In any case, the current situation is completely unacceptable. > Also, one does not need to be networked for this to be an >issue. The same machine that mounts the NFS volume can export one to >itself, or for that matter, intercept the RPC requests and feed a >logical volume instead of a physical one...... There's a serious problem with this, and it results in kernel deadlocks. We had some very, very bad experiences with this when some students got their hands on a copy of "Touch." As for logical volume stuff, that's best accomplished with a /usr/filesystems/ driver. NFS is designed to do one thing well; it's not intended to be a generic file system interface. [No, I'm not happy with autoNetWaremount.] -=EPS=-
Newsgroups: comp.sys.next.bugs,comp.sys.next.misc From: Eduardo Ostertag Subject: NeXTstation freezing and/or panicking with NFS backup Message-ID: <1994May6.221140.18847@tolten.puc.cl> Sender: news@tolten.puc.cl (News Manager) Organization: Pontificia Universidad Catolica de Chile Date: Fri, 6 May 1994 22:11:40 GMT I have a Non-Turbo NeXTstation connected via Ethernet to a network of heterogenous machines (Unixes and PCs). Each time we attempt to backup my hard disk via NFS from another Unix machine, the NeXT either freezes (no respond to <command>-<command>-~) or panics with an "unexpected kernel page fault failure" message. I have tried doing the same from various Unix machines (HPs, Suns, NCRs), but the result is the same. In fact, it seems this NeXTstation hates NFS. I once tried to use it as a File/Mail server and each time the load got high (lots of clients using the NeXTstation) the NeXT died. I have checked the hardware using a software tool sent to me by someone at NeXT (dvt040), but all tests finish successfully. I have checked the wiring and its OK. I have read all the docums I have access to (FAQ, NeXTanswer, Addison-Wesley docum on NeXT) but nothing helps. If I can't do backup then my NeXT will become the most expensive VT100 of my organization. Please, help!! and thanx in advance. NeXTstation configuration is: + NeXT ROM Monitor 2.2 v63 + NeXT Mach 3.1 Fri Apr 30 08:31:15 PDT 1993 + FPU version 0x40 + physical memory = 20.00 megabytes + using 126 buffers containing 0.98 megabytes of memory + Disk: SEAGATE ST11200N Rev 8770 (1GB capacity; 48% free) + Standard NeXT Printer + No other SCCI devices -Eduardo Ostertag (ostertag@macha.orden.cl)
Newsgroups: comp.sys.next.bugs Subject: CDPlayer.app and NEC From: damon.cooper@canrem.com (Damon Cooper) Distribution: world Message-ID: <60.1421.4901.0N19DCC6@canrem.com> Date: Mon, 9 May 94 23:31:00 -0400 Organization: CRS Online (Toronto, Ontario) After reading someone's mention of getting the CDPlayer.app working with a NEC CD-ROM drive, I pulled the cover off my box and flicked the suggested dip switche in an effort to get the CDPlayer.app to work... Previous to this attempt, I had the Public Sound Server "ON", tried logging in as root, etc, etc, but all could ever get was: Probing for CDROM Probing for DOS Probing for mac Probing for cdaudio Then I got the famous "EJECT DISK" msg. Okay. Now to the current day. I read that somebody got it to work by flipping the CD-ROM drive switch that puts the drive into SCSI-2 mode instead of SCSI-1. I flipped the switch, but then when I re-boot the machine, the adaptor scans the SCSI bus (as usual) but can't find the CD-ROM. The CD-ROM "busy" light also flickers constantly when this switch is on. Help! If anyone has a similar config to below with CDPlayer.app working, I'd really like to hear about it! My config: -> Fujitsu 1G SCSI-2 -> NEC 84-I Internal CD-ROM drive -> BusLogic 747S (the earlier ver that works with the Adaptec 1542 driver) Thanks in advance!!! Damon F. Cooper damon.cooper@canrem.com --- * CmpQwk #UNREG* UNREGISTERED EVALUATION COPY
From: sanguish@digifix.com Newsgroups: comp.sys.next.announce,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.bugs,comp.sys.next.programmer,comp.sys.next.marketplace Subject: Quick Guide to NEXTSTEP information on the Internet Date: 10 May 1994 02:36:35 -0400 Organization: Next Announcements Message-ID: <2qna1j$i84@digifix.digifix.com> This post is made weekly, to help 'point' users to more NEXTSTEP information Topics include: Stepwise NEXTSTEP/OpenStep Information Server comp.sys.next newsgroups related newsgroups comp.sys.next newsgroups mailing list ftp sites NeXTanswers New Information --------------- Stepwise NEXTSTEP/OpenStep Information Server A product directory built around the World Wide Web system, this will allow full multimedia announcements by NEXTSTEP developers, as well as the ability to browse the available products for NEXTSTEP. This service is online now, and can be reached at http://digifix.digifix.com/ it can be reached using OmniWeb (available from ftp.omnigroup.com) or Mosaic. The entries are coming in quite quickly, and I'll be moving the entire NEXTSTEP Third Party Catalog contents in as soon as they become available. Additionally the NEXTSTEP Product Information Mail Server will be stocked full file files in the next week... you can get information on using the mail server at ns-products@digifix.com Suggestions or comments can be directed to me at sanguish@digifix.com comp.sys.next.* newsgroups -------------------------- Comp.Sys.Next.Advocacy This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. Comp.Sys.Next.Announce Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) The NEXTSTEP FAQs are posted here monthly as well. This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Comp.Sys.Next.Bugs A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT- specific groups as well. Comp.Sys.Next.Hardware Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. Comp.Sys.Next.Marketplace NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. Comp.Sys.Next.Misc For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! Comp.Sys.Next.Programmer Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. The NEXTSTEP programmer FAQs are posted here. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. Comp.Sys.Next.Software This is a place to talk about [third party] software products that run on NEXTSTEP systems. Comp.Sys.Next.Sysadmin Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. related Newsgroups ------------------ Comp.Soft-Sys.Nextstep Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. Comp.Lang.Objective-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. Comp.Object Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original Comp.Sys.Next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News ------------------------------------------- Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep-d next-advocacy-d next-announce-d next-bugs-d next-hardware-d next-marketplace-d next-misc-d next-programmer-d next-software-d next-sysadmin-d (For a full description, send mail saying LISTS to <digestif@antigone.com>). The subscription syntax is essentially the same as LISTSERV's. To subscribe, send a message to <digestif@antigone.com> saying: SUB Listname YourName Example: SUB next-hardware-d John Doe The ftp sites ------------- cs.orst.edu: The main site for North American submissions nova.cc.purdue.edu: Lots of older stuff, but very short on disk space ftp.informatik.uni-muenchen.de: In Germany. terra.stack.urc.tue.nl (Dutch NEXTSTEP User Group) and cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp.next.com: See the below ftp.next.com and NextAnswers@next.com ------------------------------------- From the document 1000_Help from ftp.next.com Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, or you can transfer them by anonymous ftp. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to NeXTanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO <your-address> If you have any problem with the system or suggestions for improvement, please send mail to NeXTanswers-request@NeXT.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to NeXTanswers-request@NeXT.com. Written by: Eric P. Scott eps@toaster.SFSU.EDU and Scott Anguish sanguish@digifix.com Additions from: Greg Anderson (Greg_Anderson@afs.com) and Michael Pizolato (Michael_Pizolato@afs.com)
Newsgroups: comp.sys.next.bugs From: yoda@cis.uni-muenchen.de (Marc Guenther) Subject: Re: SECURITY: NFS Mount Vulnerability (with patch) [1/2] Sender: news@informatik.uni-muenchen.de (News System) Message-ID: <CpMxIy.2K3@informatik.uni-muenchen.de> Date: Wed, 11 May 1994 11:06:33 GMT References: <CpAKr6.160@trapac.com> Organization: Institut fuer Informatik der Universitaet Muenchen In article <CpAKr6.160@trapac.com> karl@trapac.com (Karl Kraft) writes: > In article <2q6apn$pue@nic-nac.CSU.net> eps@futon.SFSU.EDU (Eric P. > Scott) writes: > > A serious security vulnerability exists in NEXTSTEP 3.2 > > and previous releases (going at least as far back as 2.1) > > that affects networked multiuser systems. > > > Eric, > > I hope you realize that some freeware, some NeXT software, > and some future commercial software depend on this feature. It's > introduction was deliberate, it is not accidental... Something which allows all users to gain root access is definitely NOT a feature !!!! Anyway, which software depends on this ??? -- Marc Guenther ------------------------------------------------------------------------- Centrum fuer Informations | Wagmuellerstr. 23 | Phone: +49 89 211 0670 und Sprachverarbeitung | 80538 M"unchen | Fax: +49 89 211 0674 University of Munich | Germany | yoda@cis.uni-muenchen.de -------------------------------------------------------------------------
From: quarrie@darth.lbl.gov (David R. Quarrie) Newsgroups: comp.sys.next.bugs,comp.sys.next.software Subject: C Compiler Bug? Date: 11 May 1994 17:20:55 GMT Organization: Lawrence Berkeley Laboratory, Berkeley CA Distribution: world Message-ID: <38984@dog.ee.lbl.gov> The following fragments of code and i386 assembler illustrate a suspected bug that I've not been able to isolate in a simple test program. There are two C code fragments (which are not meant to be human-readable, they're automatically generated), the first of which fails, the second, which uses a static char* variable to hold an intermediate result, succeeds. I enclose the C code fragments and the (hopefully) relevant assembler output. If anyone understands what the problem is, and even better, a solution I would be eternally grateful. I've tried several switches to the C compiler without success (e.g. no -O, -fno-defer-pop -fno-function-cse -fvolatile etc.). I should add that this is running NEXTSTEP 3.2 using the standard NeXT C compiler. David ----------------------------------------------------------------------------- Here's the Intel (i386) assembler output for the BAD case: C code: Eakt1Ho(l[1], l[0]); ti1 = *(long *) (l[1] + (EbNpqj8-497)[dtype]); EamwkjW(l[4], ti1); EazcALc(EaeIDNx(l[1]), l[4]); tc1 = Eax7a6T(EadPGtu(l[1])); Assembler: call _Eakt1Ho addl $8,%esp movl -92(%ebp),%eax addl $4,%eax movl -88(%ebp),%edx movl (%eax),%eax addl _EbNpqj8-1988(,%edx,4),%eax movl (%eax),%ebx movl %ebx,-84(%ebp) pushl %ebx movl -92(%ebp),%eax addl $16,%eax movl (%eax),%ebx pushl %ebx call _EamwkjW addl $8,%esp movl -92(%ebp),%eax addl $16,%eax movl (%eax),%ebx pushl %ebx movl -92(%ebp),%eax addl $4,%eax movl (%eax),%ebx pushl %ebx call _EaeIDNx addl $4,%esp movl %eax,%eax pushl %eax call _EazcALc addl $8,%esp movl -92(%ebp),%eax addl $4,%eax movl (%eax),%ebx pushl %ebx call _EadPGtu addl $4,%esp movl %eax,%eax pushl %eax call _Eax7a6T ----------------------------------------------------------------------------- Here's the GOOD case: C Code: char* temp; [....] Eakt1Ho(l[1], l[0]); ti1 = *(long *) (l[1] + (EbNpqj8-497)[dtype]); EamwkjW(l[4], ti1); temp = EaeIDNx(l[1]); EazcALc(temp, l[4]); tc1 = Eax7a6T(EadPGtu(l[1])); Assembler: call _Eakt1Ho addl $8,%esp movl -96(%ebp),%eax addl $4,%eax movl -92(%ebp),%edx movl (%eax),%eax addl _EbNpqj8-1988(,%edx,4),%eax movl (%eax),%ebx movl %ebx,-88(%ebp) pushl %ebx movl -96(%ebp),%eax addl $16,%eax movl (%eax),%ebx pushl %ebx call _EamwkjW addl $8,%esp movl -96(%ebp),%eax addl $4,%eax movl (%eax),%ebx pushl %ebx call _EaeIDNx addl $4,%esp movl %eax,-80(%ebp) movl -96(%ebp),%eax addl $16,%eax movl (%eax),%ebx pushl %ebx movl -80(%ebp),%ebx pushl %ebx call _EazcALc addl $8,%esp movl -96(%ebp),%eax addl $4,%eax movl (%eax),%ebx pushl %ebx call _EadPGtu addl $4,%esp movl %eax,%eax pushl %eax call _Eax7a6T --------------------------------------------------------------------- David R. Quarrie Lawrence Berkeley Lab MS 50B-3238 (NeXTmail ok) 1 Cyclotron Road Internet: DRQuarrie@LBL.Gov Berkeley, CA 94720 DECnet : lbl::drquarrie Tel: (510) 486-4868
From: sean@zapotec.math.byu.edu (Sean O. Luke) Newsgroups: comp.sys.next.bugs,comp.sys.next.sysadmin Subject: Novell REALLY REALLY SLOW Date: 11 May 1994 17:45:29 GMT Organization: Brigham Young University Message-ID: <2qr5jp$7t4@hamblin.math.byu.edu> I've noticed to major bugs in NeXTSTEP 3.2, and I'm sure they've been noticed before on _some_ users' system, so I'm posting to see if I could get some help. 1) I'm using a Microsoft serial mouse (I know, I know), and an Intel system (a System E straight _from_ Intel). The Workspace Manager seems to be having a really bad time with clicking on icons and drag-and-drop. Namely, double-clicking on an icon freaks it out, and drag-and-drop no longer work. It might have something to do with the evil serial mouse driver, but I don't think so. 2) Novell services are SO SLOW AS TO BE UNUSABLE!!!! Contrast this to my coworkers who have no problem at all accessing the exact same network at blinding speed in Windows. This is unacceptable. Can anyone tell me why it would take a full minute to open a directory containing three files on an underutilized Novell server? +---------------------------------------------------------+ | Sean Luke This signature no verb | | sean@digaudio.byu.edu ,,, | | sean@zapotec.math.byu.edu (o o) ??? in September | +------------------------oOO--(_)--OOo--------------------+
From: quarrie@darth.lbl.gov (David R. Quarrie) Newsgroups: comp.sys.next.software,comp.sys.next.bugs Subject: Re: C Compiler Bug? Date: 11 May 1994 21:25:31 GMT Organization: Lawrence Berkeley Laboratory, Berkeley CA Distribution: world Message-ID: <38989@dog.ee.lbl.gov> Please ignore my recent post regarding a possible C compiler bug. It turns out that the problem has been understood independently. I (naturally) only heard of that after I'd hit the "send" key. Sorry for the waste of bandwidth. David --------------------------------------------------------------------- David R. Quarrie Lawrence Berkeley Lab MS 50B-3238 (NeXTmail ok) 1 Cyclotron Road Internet: DRQuarrie@LBL.Gov Berkeley, CA 94720 DECnet : lbl::drquarrie Tel: (510) 486-4868
Newsgroups: comp.sys.next.bugs From: irving@Happy-Man.com (Irving_Wolfe) Subject: Re: FAX problems Date: Thu, 12 May 1994 02:42:20 GMT Message-ID: <Cpo4uL.AKD@Happy-Man.com> References: <2q8de5$e3@keen.ccit.duq.edu> Organization: Happy Man Corp, 4410 Pt Robinson, Vashon, WA 98070 206/463-9399 In <2q8de5$e3@keen.ccit.duq.edu> mcquill@next.duq.edu (Tod McQuillin) writes: >There are some destinations to which I >attempt to send a fax which yield the following message in my fax log: >Unsupported (local) >and the fax does not go through. BTW my fax modem is a DoveFax. We found this too when I started to send faxes to a particular model of Brother fax machine. It means there are complementary bugs in the fax machine and the DoveFax, that seems to make the DoveFax think the fax machine is requiring a communications protocol that it doesn't support, even though this is not really the case. My solution was to sell the DoveFax cheap to a friend and buy the NXFax/Zyxel combination from Black & White Software. This is a costly solution, but easy, and it works. (contact info@bandw.com) Regards, - Irving -- Irving_Wolfe@Happy-Man.com 206/463-9399 x101 fax extension 108 Happy Man Corp. 4410 SW Pt. Robinson Rd., Vashon, WA 98070-7399 In SOLID VALUE, we show intelligent investors under-priced stocks Printed Info Free: Send POSTAL address: Solid-Value@Happy-Man.com
From: wmorse@erasure-sl.cc.emory.edu (William Morse) Newsgroups: comp.sys.next.bugs Subject: "lookupd" dies under heavy use... Date: 12 May 1994 04:07:42 GMT Organization: Emory University, Dept of Math and CS Distribution: world Message-ID: <2qsa2e$d34@emory.mathcs.emory.edu> Problem: The "/usr/etc/lookupd" daemon dies under heavy use. Results: Lookupd handles system to DNS and system to Netinfo communication. Thus, when the daemon dies, DNS dies and the system "forgets" everything in netinfo. So, you get the following (you can simulate the daemon crash by under root, executing "kill -9 <lookupd pid>" NOTE: remain logged in as root so you can restart it with "/usr/etc/lookupd"): ----- Example #1: houdini:4# finger Login Name TTY Idle When Office smyerbur ??? smyerbur ??? wmorse ??? ----- Example #2: houdini:5# telnet houdini houdini: Unknown host ----- Example #3: houdini> whoami Intruder alert. ----- Example #4: houdini:10# ps -aux USER PID %CPU %MEM VSIZE RSIZE TT STAT TIME COMMAND UNKNOWN 178 85.2 36.4 14.0M 5.82M ? R 125:18 - console (WindowServer) UNKNOWN 187 11.0 7.9 5.16M 1.27M ? R N 16:02 root -1 1.9 29.2 34.9M 4.66M ? U 1:47 <mach-task> root 265 0.9 0.5 1.69M 80K p3 S 0:00 telnetd root -1 0.0 0.0 0K 0K ? ?W< 0:00 <mach-task> root 1 0.0 0.2 744K 40K ? S 0:00 /usr/etc/init -xx root 0 0.0 29.2 34.9M 4.66M ? R N 1:45 (kernel idle) .. The UNKNOWN are users who the system has forgotten who they are. As a result, wheel cannot "su" to root when lookupd dies. Thus, (s)he cannot restart it with "/usr/etc/lookupd". Further, no new users can log into the machine when this occurs. Mail also begins to bounce. In other words, when this daemon dies, bad things happen. ----- Note: this the system did not "forget" the users in NS 3.0. So, this bug would have been less nasty. Unfortunately, there is no way to reproduce what causes the daemon to die. The log on it just stops. Increasing the daemon's cache flush does no good. It just seems to go when there are heavy DNS requests and heavy Netinfo system calls. Ah well... This bug has been reported to NeXT. WORKAROUND: An rough and ready gawk script... saved as a file and run in cron. (You will need to compile gawk for this to work) - ps -aux | gawk ' BEGIN { x="" } index($10,"lookupd") { x="a" } END { print x "b" } ' | gawk ' !(index($0,"ab")) { print "lookupd restarted" system("/usr/etc/lookupd") } ' I am interested in feedback on this problem... - William Morse Sysadmin. / Emory Public NeXT Network Consult. / Emory ITD
Newsgroups: comp.sys.next.bugs,comp.sys.next.sysadmin From: csaldanh@mae.carleton.ca (Chris Saldanha) Subject: Re: Novell REALLY REALLY SLOW Message-ID: <CpoCp7.Dww@cunews.carleton.ca> Followup-To: comp.sys.next.bugs,comp.sys.next.sysadmin Sender: news@cunews.carleton.ca (News Administrator) Organization: Carleton University References: <2qr5jp$7t4@hamblin.math.byu.edu> Date: Thu, 12 May 1994 05:31:54 GMT Sean O. Luke (sean@zapotec.math.byu.edu) wrote: : I've noticed to major bugs in NeXTSTEP 3.2, and I'm sure they've : been noticed before on _some_ users' system, so I'm posting to : see if I could get some help. : 1) I'm using a Microsoft serial mouse (I know, I know), and an : Intel system (a System E straight _from_ Intel). The Workspace : Manager seems to be having a really bad time with clicking on icons : and drag-and-drop. Namely, double-clicking on an icon freaks it out, : and drag-and-drop no longer work. It might have something to do with : the evil serial mouse driver, but I don't think so. : 2) Novell services are SO SLOW AS TO BE UNUSABLE!!!! Contrast this : to my coworkers who have no problem at all accessing the exact same : network at blinding speed in Windows. This is unacceptable. Can anyone : tell me why it would take a full minute to open a directory containing : three files on an underutilized Novell server? Well, I'm running on a non-certified, non-listed clone 486, also with a Micro$oft serial mouse, and I seem to be able to double-click just fine. Same goes for dragging. What does 'freaks it out' mean anyway? Would you be referring to what gets put in the console as 'Drag service protocol error'? Please be specific about bugs/problems. I have just recently been doing things with one of our clients' NetWare network through NeXTSTEP (Intel & Motorola). Access through WorkSpace Manager is excellent. I have also been using Netware through SoftPC, and have had access to the server for login & disk access at a _Very_ acceptible speed. --Chris Chris Saldanha -------------------------------------- Carleton University (Comp. Sci) |"The eternal silence of these infinite| chris@computerActive.on.ca (NeXTMail) | spaces terrifies me." -Blaise Pascal| csaldanh@mae.carleton.ca (NeXT/MIME) ------------------:-o-----------------
From: lemson@ux1.cso.uiuc.edu (David Lemson) Newsgroups: comp.sys.next.bugs Subject: Re: "lookupd" dies under heavy use... Date: 12 May 1994 06:53:32 GMT Organization: University of Illinois at Urbana Message-ID: <2qsjpc$9pj@vixen.cso.uiuc.edu> References: <2qsa2e$d34@emory.mathcs.emory.edu> wmorse@erasure-sl.cc.emory.edu (William Morse) writes: >Problem: The "/usr/etc/lookupd" daemon dies under heavy use. Just as a data point, I have one machine that is the top level server for about 2700 users (a very large database)... the lookupd there gets hit VERY hard, and has never died. You might try running lookupd with the -m 0 flag... I've never tried this, but since this turns off caching, it might make it safer in your situation. -- David Lemson University of Illinois Computing & Comm Services Office System Administrator Internet : lemson@uiuc.edu UUCP :...!uiucuxc!uiucux1!lemson NeXTMail & MIME accepted BITNET : LEMSON@UIUCVMD
Newsgroups: comp.sys.next.bugs From: francisr@stupid.ucs.indiana.edu (Rob Francis) Subject: Re: "lookupd" dies under heavy use... Message-ID: <CpoxyI.9u3@usenet.ucs.indiana.edu> Sender: news@usenet.ucs.indiana.edu (USENET News System) Organization: Indiana University References: <2qsa2e$d34@emory.mathcs.emory.edu> <2qsjpc$9pj@vixen.cso.uiuc.edu> Date: Thu, 12 May 1994 13:11:05 GMT In article <2qsjpc$9pj@vixen.cso.uiuc.edu>, David Lemson <lemson@uiuc.edu> wrote: >wmorse@erasure-sl.cc.emory.edu (William Morse) writes: > >>Problem: The "/usr/etc/lookupd" daemon dies under heavy use. > >Just as a data point, I have one machine that is the top level >server for about 2700 users (a very large database)... the lookupd >there gets hit VERY hard, and has never died. I can say the same thing here. I have one machine as the top level server for over 3000 users and haven't had any trouble either. -rob --- Rob Francis Unix Workstation Support Group | francisr@indiana.edu (812)855-0327 University Computing Services, Indiana University
From: vasu@beancounter.commerce.ubc.ca (Vasu Krishnamurthy) Newsgroups: comp.sys.next.bugs Subject: NewsGrazer Crashes Date: 12 May 1994 18:15:54 GMT Organization: The University of British Columbia Message-ID: <2qtroq$5cf@nntp.ucs.ubc.ca> During the last few months there has been a steady stream of posts on frequent crashes by NewsGrazer. Since NG wasn't exhibiting this behavior on my machine, I ignored the discussion. However, NG inexplicably caught the disease yesterday, so would some kind soul e-mail me any known fixes? I am running NG version 75 on NeXT hardware with NS 3.2. Thanks in advance. -- Vasu Krishnamurthy Assistant Professor Faculty of Commerce and Business Administration University of British Columbia Vancouver, BC V6T 1Z2 CANADA
From: wmorse@erasure-sl.cc.emory.edu (William Morse) Newsgroups: comp.sys.next.bugs Subject: Re: "lookupd" dies under heavy use... Date: 12 May 1994 22:12:23 GMT Organization: Emory University, Dept of Math and CS Distribution: world Message-ID: <2qu9k7$5e9@emory.mathcs.emory.edu> References: <CpoxyI.9u3@usenet.ucs.indiana.edu> In article <CpoxyI.9u3@usenet.ucs.indiana.edu> francisr@stupid.ucs.indiana.edu (Rob Francis) writes: > In article <2qsjpc$9pj@vixen.cso.uiuc.edu>, > David Lemson <lemson@uiuc.edu> wrote: > >wmorse@erasure-sl.cc.emory.edu (William Morse) writes: > > > >>Problem: The "/usr/etc/lookupd" daemon dies under heavy use. > > > >Just as a data point, I have one machine that is the top level > >server for about 2700 users (a very large database)... the lookupd > >there gets hit VERY hard, and has never died. > > I can say the same thing here. I have one machine as the top level > server for over 3000 users and haven't had any trouble either. > > -rob > --- > Rob Francis > Unix Workstation Support Group | francisr@indiana.edu (812)855-0327 > University Computing Services, Indiana University ---- I spent a great deal of time going over the logs and checking into the connections made when the NeXT server lookupd would die. This is what I found: (Note: the following has been fixed and no, I did not set up): -- dooley> host euccmicro euccmicro.cc.emory.edu has address 128.140.9.7 euccmicro.cc.emory.edu has address 128.140.9.8 euccmicro.cc.emory.edu has address 128.140.9.9 euccmicro.cc.emory.edu has address 128.140.9.10 euccmicro.cc.emory.edu has address 128.140.9.11 euccmicro.cc.emory.edu has address 128.140.9.12 euccmicro.cc.emory.edu has address 128.140.9.13 euccmicro.cc.emory.edu has address 128.140.9.14 euccmicro.cc.emory.edu has address 128.140.9.15 euccmicro.cc.emory.edu has address 128.140.9.16 euccmicro.cc.emory.edu has address 128.140.9.17 euccmicro.cc.emory.edu has address 128.140.9.18 euccmicro.cc.emory.edu has address 128.140.9.19 euccmicro.cc.emory.edu has address 128.140.9.20 euccmicro.cc.emory.edu has address 128.140.9.21 euccmicro.cc.emory.edu has address 128.140.9.22 euccmicro.cc.emory.edu has address 128.140.9.23 euccmicro.cc.emory.edu has address 128.140.9.24 euccmicro.cc.emory.edu has address 128.140.9.25 euccmicro.cc.emory.edu has address 128.140.9.26 euccmicro.cc.emory.edu has address 128.140.9.27 euccmicro.cc.emory.edu has address 128.140.9.28 euccmicro.cc.emory.edu has address 128.140.9.29 euccmicro.cc.emory.edu has address 128.140.9.30 euccmicro.cc.emory.edu has address 128.140.9.31 euccmicro.cc.emory.edu has address 128.140.9.32 euccmicro.cc.emory.edu has address 128.140.9.33 euccmicro.cc.emory.edu has address 128.140.9.34 euccmicro.cc.emory.edu has address 128.140.9.35 euccmicro.cc.emory.edu has addresspacket size error (0xdffff644 != 0xdffff744) -- There were 120 addresses like this which were dynamically assigned to machines as needed on logon. Obviously, doing the above was a mistake. But, to have lookupd die when it tried to resolve the ip numbers of these machines is also not good. Such vulnerabilities should be noted and corrected. Also note: the death of lookupd because of a DNS problem also took out system queries to netinfo. Maybe the functions should be separated? In any case, this daemon could at least be "multithreaded" so as to prevent a request to a broken on bad DNS server to freeze NS Netinfo queries as is currently the case. Anyway, I welcome further comments. - William Sysadmin. / Emory Public NeXT Network Consult. / ITD
From: Steve_Wessels@pol.com (Stephan B. Wessels) Newsgroups: comp.sys.next.bugs Subject: Gator FTP + / intel Followup-To: comp.sys.next.bugs Date: 12 May 1994 22:03:00 GMT Organization: Plum Street Software Message-ID: <Steve_Wessels-120594180315@portable.plum.pol.com> I'm running the GatorFTP+ app for Intel and just recently started to get a problem showing up. Everytime I actually go and try and do a file transfer, via get or uncompress/get, the program just exits. It used to work. Perhaps I have some kind of bad configuration or setup? Anyone know what this might be? Thanks in advance, - Steve
Newsgroups: comp.sys.next.bugs From: paul@seer.demon.co.uk (Paul Lynch) Subject: Re: Gator FTP + / intel Message-ID: <1994May13.063248.12234@seer.demon.co.uk> Sender: news@seer.demon.co.uk Organization: P & L Systems References: <Steve_Wessels-120594180315@portable.plum.pol.com> Date: Fri, 13 May 1994 06:32:48 GMT In article <Steve_Wessels-120594180315@portable.plum.pol.com> Steve_Wessels@pol.com (Stephan B. Wessels) writes: > I'm running the GatorFTP+ app for Intel and just recently started to get a > problem showing up. Everytime I actually go and try and do a file > transfer, via get or uncompress/get, the program just exits. > > It used to work. > > Perhaps I have some kind of bad configuration or setup? Anyone know what > this might be? If you delete the configuration file in your home directory (.gatorftprc) it should just start working again. This happens a lot. Paul -- Paul Lynch P & L Systems (NeXTmail) paul@seer.demon.co.uk Tel: (0494)671501 9 Stable Lane, Seer Green, Fax: (0494)680228 Bucks, HP9 2YT, UK
From: peter@rhea.teorfys.uu.se (Peter Trygg) Newsgroups: comp.sys.next.bugs Subject: rg property in printcap does not work Date: 13 May 1994 14:06:15 GMT Organization: Uppsala University Message-ID: <2r01gn$t7k@columba.udac.uu.se> Is the rg (resticted group property) supposed to work on the NeXT:s? I added the property rg with the value printer to all printers. If a user is not a member of the group printer, he can not print on a network printer BUT he can print on Local_Printer (Yes I added rg to that also!). Anyone but me facing that problem? Peter Trygg peter@rhea.teorfys.uu.se Dept. of Theoretical Physics Uppsala University
From: schmo1@info.isbiel.ch (Olivier Schmid) Newsgroups: comp.sys.next.bugs Subject: Re: Novell REALLY REALLY SLOW Date: 13 May 1994 14:57:35 GMT Organization: Biel School of Engineering, CH-2501 Biel, Switzerland Distribution: world Message-ID: <2r04gv$ker@vega.info.isbiel.ch> I've exactly the same problem. My machine works fine (even SoftPC) runs well but the NFS-access is very slow (sometimes a few minutes to read a directory !). A slow 386 on the same net works nice. Oli
From: lestat@cs.utexas.edu (Faizel Dakri) Newsgroups: comp.sys.next.bugs Subject: Re: Gator FTP + / intel Date: 13 May 1994 19:11:00 GMT Organization: The University of Texas at Austin, Austin, Texas Message-ID: <2r0jc4$rl4@geraldo.cc.utexas.edu> References: <Steve_Wessels-120594180315@portable.plum.pol.com> In article <Steve_Wessels-120594180315@portable.plum.pol.com> Steve_Wessels@pol.com (Stephan B. Wessels) writes: > I'm running the GatorFTP+ app for Intel and just recently started to get a > problem showing up. Everytime I actually go and try and do a file > transfer, via get or uncompress/get, the program just exits. > > It used to work. > > Perhaps I have some kind of bad configuration or setup? Anyone know what > this might be? > > Thanks in advance, > > - Steve If you're running version 1.4, then this happens quite a bit. You need to remove your .gatorftprc file everytime you run GatorFTP. BUT, there is a new version of GatorFTP (ver. 1.5) that seems to work okay now. The only difference is that the pie chart no longer updates. But, at least now you can save your sites between sessions. You can get it from: uhoop.egr.uh.edu and look in /pub/misc/GatorFTP Hope that helps, -Faizel
Newsgroups: comp.sys.next.bugs Subject: Re: "lookupd" dies under heavy use... Message-ID: <MATTHEW.94May15155658@dyson.dp.ox.ac.uk> From: matthew@dyson.dp.ox.ac.uk (Matthew Seaman) Date: 15 May 1994 14:56:57 GMT References: <2qsa2e$d34@emory.mathcs.emory.edu> Distribution: world Organization: Dyson Perrins Laboratory (Organic Chemistry), Oxford University In-reply-to: wmorse@erasure-sl.cc.emory.edu's message of 12 May 1994 04:07:42 GMT >>>>> "WM" == William Morse <wmorse@erasure-sl.cc.emory.edu> writes: WM> Problem: The "/usr/etc/lookupd" daemon dies under heavy use. [Munch] WM> As a result, wheel cannot "su" to root when lookupd dies. Thus, WM> (s)he cannot restart it with "/usr/etc/lookupd". Further, no new WM> users can log into the machine when this occurs. Mail also begins WM> to bounce. I too have been bitten by lookupd failures --- caused by a fault with a hard disk rather than any problem intrinsic to lookupd. However my experience differs in some important respects. The machine came up with the 'me' account logged in. Very odd, as I had religiously removed that account from netinfo. Even worse: su would allow 'me' to become root *without* a password. I believe that what was happening was that in the absence of lookupd, the system reverted to the standard flat file /etc/passwd, /etc/groups etc. system of most Unices. I had left the default /etc/passwd file as shipped with NS 3.2 in place, and that contains a password-less root account. The moral of the story is: always keep valid flat file versions of /etc/passwd, /etc/group and other important system databases around. Perhaps limit them to just the root account (with password...) and the system managers' accounts, so that only a responsible person could log on in the event of a lookupd failure. WM> In other words, when this daemon dies, bad things happen. Indeed. Any cracker that can successfully subvert your lookupd has your system at their mercy. Matthew -- Matthew Seaman <Matthew.Seaman@dyson.ox.ac.uk> Non-NeXT mail <matthew@viriconium.ocms.ox.ac.uk> NeXT mail Inorganic Chemistry Laboratory, S. Parks Rd., Oxford, OX1 3QR, England Tel +44 (0)865 272640 Fax +44 (0)865 272690
From: davidwr@dip.eecs.umich.edu (David Richardson) Newsgroups: comp.sys.next.bugs Subject: Intel: Adaptec 1542CF + ATI vid. drivers = HD errors Date: 16 May 1994 18:42:44 GMT Organization: University of Michigan EECS Dept. Message-ID: <2r8er4$d4c@zip.eecs.umich.edu> I am running an Intel 486 box, NeXTSTEP 3.2. Video: ATI VMEBus card SCSI: Adaptec 1542CF, running 154x drivers dated 1994. Problem: When I use either of the ATI video drivers ("Ultra" or "On-board"), I get disk errors on the hard disk. With the default VGA adapter, I have no problems (but no color, & no high-resolution!). This *sounds like* the problem in NeXT's NeXTAnswers #1108 "Adaptec 1542C and Adaptec 1540C SCSI Adapters", in which ATI Ultra Pro EISA cards are known to have SCSI data corruption when used with Adaptec 15xx series SCSI cards. Anybody have any suggested workarounds? I've contacted Adaptec, ATI, and the people who put our computer together. All have been helpful, but nobody seems to have an answer (yet). Reply-to: David Richardson, richardson@laa.com +1 313 995 5590
From: mcquill@next.duq.edu (Tod McQuillin) Newsgroups: comp.sys.next.bugs Subject: NS3.[12] emacs Date: 16 May 1994 19:43:47 -0400 Organization: Duquesne University, Pittsburgh, PA USA Message-ID: <2r90fj$61l@keen.ccit.duq.edu> Thankfully, NeXTStep 3.1/3.2 developer includes emacs 18.59, instead of the older 18.55 included with earlier releases, but M-x manual-entry is broken. Since they changed the system type from berkeley-unix to next-mach, paths.el defines manual-program as /usr/bin/man instead of /usr/ucb/man. What lossage. My fix was just to symlink /usr/ucb/man to /usr/bin/man, since paths.el gets loaded before emacs is dumped.
From: mcquill@next.duq.edu (Tod McQuillin) Newsgroups: comp.sys.next.bugs Subject: GNU iscreen 3.5.2/NeXTStep Date: 16 May 1994 19:54:39 -0400 Organization: Duquesne University, Pittsburgh, PA USA Message-ID: <2r913v$645@keen.ccit.duq.edu> Has anyone got GNU iscreen 3.5.2 working under NeXTStep 3.[012]? The attacher process dies when trying to reconnect to a detached screen. The detached screen process writes this failure to the debug log: Read error: 5 - SigHup()ing! which indicates EIO reading from the d_userfd for the window. screen 3.2b works fine.
From: erikkay@next.com (Erik Kay) Newsgroups: comp.sys.next.bugs Subject: Re: NS3.[12] emacs Date: 17 May 1994 00:36:51 GMT Organization: NeXT, Inc. Message-ID: <2r93j3$8km@rosie.next.com> References: <2r90fj$61l@keen.ccit.duq.edu> In article <2r90fj$61l@keen.ccit.duq.edu> mcquill@next.duq.edu (Tod McQuillin) writes: > Thankfully, NeXTStep 3.1/3.2 developer includes emacs 18.59, instead of > the older 18.55 included with earlier releases, but M-x manual-entry > is broken. Since they changed the system type from berkeley-unix to > next-mach, paths.el defines manual-program as /usr/bin/man instead of > /usr/ucb/man. What lossage. This bug has already been fixed for our next release. Erik
From: Lennart_Lovstrand@NeXT.COM Newsgroups: comp.sys.next.bugs Subject: Re: NS3.[12] emacs Date: 17 May 1994 00:41:49 GMT Organization: NeXT, Inc. Message-ID: <2r93se$8kq@rosie.next.com> References: <2r90fj$61l@keen.ccit.duq.edu> In article <2r90fj$61l@keen.ccit.duq.edu> mcquill@next.duq.edu (Tod McQuillin) writes: > Thankfully, NeXTStep 3.1/3.2 developer includes emacs 18.59, instead of > the older 18.55 included with earlier releases, Thanks! ;-) > but M-x manual-entry > is broken. Since they changed the system type from berkeley-unix to > next-mach, paths.el defines manual-program as /usr/bin/man instead of > /usr/ucb/man. What lossage. Not our working, that came from FSF. Fixed in 3.3, though. > My fix was just to symlink /usr/ucb/man to /usr/bin/man, since > paths.el gets loaded before emacs is dumped. Or just put (setq manual-program "/usr/ucb/man") in your ~/.emacs file. Cheers, --Lennart -- (let ((foo '(nothing can go wrong))) (rplacd (cdddr foo) (cddr foo)) foo)
From: hdwoernd@faui06d.informatik.uni-erlangen.de (Harald Woerndl-Aichriedler) Newsgroups: comp.sys.next.sysadmin,de.comp.sys.next,comp.sys.next.hardware,comp.sys.next.bugs Subject: Bugfix: ATI rev.3 @1120x832 Date: Tue, 17 May 1994 16:26:47 +0200 Organization: Student Pool, CSD, University of Erlangen, Germany Message-ID: <2rak77$fns@faui06d.informatik.uni-erlangen.de> Summary: ATI r3 now works great! Keywords: ATI bug viedeo Good news for all folks using an older ATI, revision 3 Card: Now it works nearly CLEAN at 1120x832. You only have to recompile the 'ATI-driver in /NextDeveloper/Examples/Driverkit/ATI...' with the following changes in the file "ATIconfig.c": static const ATI_CRTSetup ATI_crt_1120_68 = { 1120, 832, 173, 139, 138, 15, 1725, 1663, 1666, 6, 0x23, 0x062c, 0, }; the values mean: 173 (* 8) pixels total (before 171) 139 (* 8) pixels visable (unchanged) 138 (* 8) pixels when hSync comes (before 145) 15 (* 8) pixels hSync width (before 12) 1725 (/ 2) lines total (before 1736) 1663 (/ 2) lines visable (unchanged) 1666 (/ 2) lines when vSync comes (before 1669) 6 (/ 2) lines vSync width (unchanged) 0x23 .. control flags (unchanged) 0x062c .. clock select for 80 MHz (unchanged) 0 .. mux-flag (??? - unchanged) The Settings also better adjust the width for low-cost 17"-displays. WARNING: The display must work with HSync-signals up to 60 KHz, a pixel-clock up to 80 MHz and vSync-signals up to 67.5 Hz. The new vertical refresh rate is 67Hz, the old one is 67.7 Hz (not 68). My System now runs nearly with no black or white lines at the left border of the screen. (You don't recognice them) If you want to experience with new resolutions, you will like this setting for 17" displays (1280x960 @ 67Hz, BW:8): static const ATI_CRTSetup ATI_crt_1280_67 = { 1280, 960, 206, 159, 162, 23, 1992, 1920, 1926, 12, 0x23, 0x0628, 1, }; To install a new resolution, fill in the CRTSetup and update the table 'ATIMode[]'. You also have to add one line into the file 'localiceable.string'. If someone is able to run the ATI interlaced, please send me the description, how you did it, because I got an old 19" fix-frequence greyscale monitor (80Hz i). Waiting for other (re)solutions... Bye, Harry
From: chris@nice.usergroup.ethz.ch (Christian Limpach) Newsgroups: comp.sys.next.bugs Subject: Re: GNU iscreen 3.5.2/NeXTStep Date: 17 May 1994 18:15:38 GMT Organization: welcome to nowhere... Message-ID: <CHRIS.94May17201538arkin@nice.usergroup.ethz.ch> References: <2r913v$645@keen.ccit.duq.edu> In-reply-to: mcquill@next.duq.edu's message of 16 May 1994 19:54:39 -0400 Originator: ARKIN@nice In article <2r913v$645@keen.ccit.duq.edu> mcquill@next.duq.edu (Tod McQuillin) writes: > Has anyone got GNU iscreen 3.5.2 working under NeXTStep 3.[012]? The > attacher process dies when trying to reconnect to a detached screen. it's although possible to reattach from a different tty... > screen 3.2b works fine. I think that even 3.5.0 worked fine... christian
From: mcquill@next.duq.edu (Tod McQuillin) Newsgroups: comp.sys.next.bugs Subject: Re: NS3.[12] emacs Date: 17 May 1994 19:11:49 -0400 Organization: Duquesne University, Pittsburgh, PA USA Message-ID: <2rbivl$2li@keen.ccit.duq.edu> References: <2r90fj$61l@keen.ccit.duq.edu> <2r93j3$8km@rosie.next.com> In article <2r93j3$8km@rosie.next.com>, Erik Kay <erikkay@next.com> wrote: > In article <2r90fj$61l@keen.ccit.duq.edu> mcquill@next.duq.edu (Tod > McQuillin) writes: > > [...] Since they changed the system type from berkeley-unix to > > next-mach, paths.el defines manual-program as /usr/bin/man instead of > > /usr/ucb/man. What lossage. > > This bug has already been fixed for our next release. As says also Lennart Lovstrand. Thanks for the prompt response, guys. Sorry if I sounded a bit harsh. Any chance of having Carl Edman's "Emacs for NeXTstep 4.0" available in a future NeXTStep release? (see sonata.cc.purdue.edu:/pub/next/submissions/Emacs_for_NeXTstep_4.0.README for details). -- Tod McQuillin Duquesne University
From: mcquill@next.duq.edu (Tod McQuillin) Newsgroups: comp.sys.next.bugs Subject: Re: GNU iscreen 3.5.2/NeXTStep Date: 17 May 1994 19:25:36 -0400 Organization: Duquesne University, Pittsburgh, PA USA Message-ID: <2rbjpg$2ne@keen.ccit.duq.edu> References: <2r913v$645@keen.ccit.duq.edu> <CHRIS.94May17201538arkin@nice.usergroup.ethz.ch> In article <CHRIS.94May17201538arkin@nice.usergroup.ethz.ch>, Christian Limpach <chris@nice.usergroup.ethz.ch> wrote: > I think that even 3.5.0 worked fine... And Alexander Swietlicki has informed me that there are yet newer beta versions of screen available, and the latest of these, 3.5.11, works fine as far as I can tell under NS 3.2! At last! Available from ftp.uni-erlangen.de:/pub/utilities/screen For U.S. folks, I have made screen 3.5.11 available from ftp.duq.edu:/pub/packages/screen-3.5.11beta.tar.gz -- Tod McQuillin Duquesne University
From: rgc@twix.cs.umd.edu (Ross Garrett Cutler) Newsgroups: comp.sys.next.sysadmin,de.comp.sys.next,comp.sys.next.hardware,comp.sys.next.bugs Subject: Re: Bugfix: ATI rev.3 @1120x832 Followup-To: comp.sys.next.sysadmin,de.comp.sys.next,comp.sys.next.hardware,comp.sys.next.bugs Date: 18 May 1994 01:20:34 GMT Organization: U of Maryland, Dept. of Computer Science, Coll. Pk., MD 20742 Message-ID: <2rbqh2$nvn@mimsy.cs.umd.edu> References: <2rak77$fns@faui06d.informatik.uni-erlangen.de> Harald Woerndl-Aichriedler (hdwoernd@faui06d.informatik.uni-erlangen.de) wrote: : Good news for all folks using an older ATI, revision 3 Card: : Now it works nearly CLEAN at 1120x832. You only have to recompile : the 'ATI-driver in /NextDeveloper/Examples/Driverkit/ATI...' : with the following changes in the file "ATIconfig.c": ATIModes.c : static const ATI_CRTSetup ATI_crt_1120_68 = { static const ATI_CRTCSetup ATI_crt_1120_68 = { ^ : 1120, 832, : 173, 139, 138, 15, : 1725, 1663, 1666, 6, : 0x23, 0x062c, 0, }; : the values mean: : 173 (* 8) pixels total (before 171) : 139 (* 8) pixels visable (unchanged) : 138 (* 8) pixels when hSync comes (before 145) : 15 (* 8) pixels hSync width (before 12) : 1725 (/ 2) lines total (before 1736) : 1663 (/ 2) lines visable (unchanged) : 1666 (/ 2) lines when vSync comes (before 1669) : 6 (/ 2) lines vSync width (unchanged) : 0x23 .. control flags (unchanged) : 0x062c .. clock select for 80 MHz (unchanged) : 0 .. mux-flag (??? - unchanged) : The Settings also better adjust the width for low-cost 17"-displays. : WARNING: The display must work with HSync-signals up to 60 KHz, : a pixel-clock up to 80 MHz and vSync-signals up to 67.5 Hz. : The new vertical refresh rate is 67Hz, the old one is 67.7 Hz (not 68). : My System now runs nearly with no black or white lines at the left : border of the screen. (You don't recognice them) Unfortunately, I still very noticeable lines on my rev 3 ATI card (with a NEC Multisync 5fg). Have to wait for the GUT drivers...
Newsgroups: comp.sys.next.bugs From: kramer@fragile.termfrost.org (Mike Andrews) Subject: Re: GNU iscreen 3.5.2/NeXTStep Message-ID: <CpzB58.Cx@fragile.termfrost.org> Organization: Terminal Frost, Springfield OH References: <2r913v$645@keen.ccit.duq.edu> Date: Wed, 18 May 1994 03:31:55 GMT mcquill@next.duq.edu (Tod McQuillin) writes: >Has anyone got GNU iscreen 3.5.2 working under NeXTStep 3.[012]? The >attacher process dies when trying to reconnect to a detached screen. > >The detached screen process writes this failure to the debug log: >Read error: 5 - SigHup()ing! This patch was forwarded along to me a while back (though NOT directly from this person...) Seemed to do the trick for me... Mike -------------------------------------------------- Finally we have a patch for NeXT systems. There is a bug in the NeXT kernel. If one opens a named pipe for reading and does a select() on it and then another process opens the pipe for writing and immediatly closes then other writes to the pipe don't release the process from the select(). Real strange... We have modified the configure script to detect this behaviour and now sockets get selected. The appended patch fixes also some other problems (stupid cpp, signal() prototype, ...) Good luck, Michael --------------------------------------------------------------- Michael Schroeder mlschroe@faui43.informatik.uni-erlangen.de main(a){while(a=~getchar())putchar(~a-1/(~(a|32)/13*2-11)*13);} *** configure.in.orig Thu Aug 19 17:59:09 1993 --- configure.in Sat Jan 15 17:22:12 1994 *************** *** 1,6 **** dnl Process this file with autoconf to produce a configure script. dnl ! dnl $Id: configure.in,v 1.11 1993/08/16 14:51:20 mlschroe Exp $ FAU dnl dnl Many thanks to David MacKenzie for writing autoconf and dnl providing a sample configure.in file for screen. --- 1,6 ---- dnl Process this file with autoconf to produce a configure script. dnl ! dnl $Id: configure.in,v 1.15 1994/01/13 23:45:54 mlschroe Exp $ FAU dnl dnl Many thanks to David MacKenzie for writing autoconf and dnl providing a sample configure.in file for screen. *************** *** 36,41 **** --- 36,50 ---- AC_TEST_PROGRAM(main(){exit(0);},,echo "Can't run the compiler - sorry";exit) + + AC_TEST_PROGRAM([ + main() + { + int __something_strange_(); + __something_strange_(0); + } + ],echo "Your compiler does not set the exit status - sorry";exit) + AC_PROG_AWK AC_PROG_INSTALL *************** *** 47,62 **** AC_DEFINE(ISC) LIBS="$LIBS -linet" fi ! echo checking for OSF1 ! if test -f /bin/uname ; then ! if test `/bin/uname` = OSF1 || test -f /osf_boot; then ! AC_DEFINE(OSF1) # this disables MIPS again.... ! fi ! fi echo checking for MIPS if test -f /lib/libmld.a || test -f /usr/lib/libmld.a || test -f /usr/lib/cmplrs/cc/libmld.a; then ! LIBS="$LIBS -lmld" # for nlist. if test -r /dev/ptc; then AC_DEFINE(MIPS) AC_COMPILE_CHECK(wait3, , [wait3();], , --- 56,73 ---- AC_DEFINE(ISC) LIBS="$LIBS -linet" fi ! dnl echo checking for OSF1 ! dnl if test -f /bin/uname ; then ! dnl if test `/bin/uname` = OSF1 || test -f /osf_boot; then ! dnl AC_DEFINE(OSF1) # this disables MIPS again.... ! dnl fi ! dnl fi echo checking for MIPS if test -f /lib/libmld.a || test -f /usr/lib/libmld.a || test -f /usr/lib/cmplrs/cc/libmld.a; then ! test -f /bin/mx || LIBS="$LIBS -lmld" # for nlist. But not on alpha. ! dnl djm@eng.umd.edu: "... for one thing, it doubles the size of the executable" ! dnl if test -r /dev/ptc; then AC_DEFINE(MIPS) AC_COMPILE_CHECK(wait3, , [wait3();], , *************** *** 77,82 **** --- 88,99 ---- #endif ], ULTRIX=1) + if test -f /usr/lib/libpyr.a ; then + oldlibs="$LIBS" + LIBS="$LIBS -lpyr" + AC_COMPILE_CHECK(Pyramid OSX,,[open_controlling_pty("");],AC_DEFINE(OSX),LIBS="oldlibs") + fi + dnl ghazi@caip.rutgers.edu (Kaveh R. Ghazi): dnl BBN butterfly is not POSIX, but a MACH BSD system. dnl Do not define POSIX and TERMIO. *************** *** 99,105 **** #ifdef _POSIX_VERSION yes #endif ! ], echo "- you have a POSIX system";AC_DEFINE(POSIX)) fi AC_COMPILE_CHECK([System V], --- 116,122 ---- #ifdef _POSIX_VERSION yes #endif ! ], echo "- you have a POSIX system";posix=1;AC_DEFINE(POSIX)) fi AC_COMPILE_CHECK([System V], *************** *** 204,218 **** int f; (void)alarm(5); if (mknod(fin, S_IFIFO|0777, 0)) exit(1); if (stat(fin, &stb) || (stb.st_mode & S_IFIFO) != S_IFIFO) exit(1); close(0); ! if (open(fin, O_RDWR | O_NDELAY)) ! exit(1); ! if (write(0, "TEST", 4) == -1) exit(1); f = 1; if (select(1, &f, 0, 0, 0) == -1) exit(1); --- 221,257 ---- int f; (void)alarm(5); + #ifdef POSIX + if (mkfifo(fin, 0777)) + #else if (mknod(fin, S_IFIFO|0777, 0)) + #endif exit(1); if (stat(fin, &stb) || (stb.st_mode & S_IFIFO) != S_IFIFO) exit(1); close(0); ! #ifdef __386BSD__ ! /* ! * The next test fails under 386BSD, but screen works using fifos. ! * Fifos in O_RDWR mode are only used for the BROKEN_PIPE case and for ! * the select() configuration test. ! */ ! exit(0); ! #endif ! if (open(fin, O_RDONLY | O_NDELAY)) exit(1); + if (fork() == 0) + { + close(0); + if (open(fin, O_WRONLY | O_NDELAY)) + exit(1); + close(0); + if (open(fin, O_WRONLY | O_NDELAY)) + exit(1); + if (write(0, "TEST", 4) == -1) + exit(1); + exit(0); + } f = 1; if (select(1, &f, 0, 0, 0) == -1) exit(1); *************** *** 243,249 **** --- 282,292 ---- struct timeval tv; int r, x; + #ifdef POSIX + if (mkfifo(fin, 0600)) + #else if (mknod(fin, S_IFIFO|0600, 0)) + #endif exit(1); close(0); if (open(fin, O_RDONLY|O_NDELAY)) *************** *** 404,410 **** --- 447,457 ---- int l; (void)alarm(5); + #ifdef POSIX + if (mkfifo(nam, 0777)) + #else if (mknod(nam, S_IFIFO|0777, 0)) + #endif exit(1); close(0); if (open(nam, O_RDWR | O_NDELAY)) *************** *** 464,475 **** dnl echo searching for tgetent olibs="$LIBS" - LIBS="-ltermcap $LIBS" - AC_COMPILE_CHECK(libtermcap,,tgetent((char *)0, (char *)0);,, LIBS="-lcurses $olibs" AC_COMPILE_CHECK(libcurses,,tgetent((char *)0, (char *)0);,, ! echo "!!! no tgetent - no screen";exit) ! ) TERMCAP="xx|scrdumm:xx:" TERM=scrdumm export TERMCAP --- 511,524 ---- dnl echo searching for tgetent olibs="$LIBS" LIBS="-lcurses $olibs" AC_COMPILE_CHECK(libcurses,,tgetent((char *)0, (char *)0);,, ! LIBS="-ltermcap $olibs" ! AC_COMPILE_CHECK(libtermcap,,tgetent((char *)0, (char *)0);,, ! LIBS="-ltermlib $olibs" ! AC_COMPILE_CHECK(libtermcap,,tgetent((char *)0, (char *)0);,, ! echo "!!! no tgetent - no screen";exit))) ! TERMCAP="xx|scrdumm:xx:" TERM=scrdumm export TERMCAP *************** *** 502,507 **** --- 551,603 ---- AC_DEFINE_UNQUOTED(PTYRANGE1,\"$p1\") fi + dnl **** pty mode/group handling **** + dnl + dnl support provided by Luke Mewburn <lm@rmit.edu.au>, 931222 + echo "checking default tty permissions/group" + rm -f conftest_grp + AC_TEST_PROGRAM([ + #include <sys/types.h> + #include <sys/stat.h> + #include <stdio.h> + main() + { + struct stat sb; + char *x; + int om, m; + FILE *fp; + + if (!(x = ttyname(0))) exit(1); + if (stat(x, &sb)) exit(1); + om = sb.st_mode; + if (chmod(x, om & ~077)) exit(1); + m = system("mesg y"); + if (m == -1 || m == 127) exit(1); + if (stat(x, &sb)) exit(1); + m = sb.st_mode; + if (chmod(x, om)) exit(1); + if (!(m & 002)) + { + if (!(fp=fopen("conftest_grp", "w"))) + exit(1); + fprintf(fp, "%d\n", sb.st_gid); + fclose(fp); + } + exit(0); + } + ],[ + if test -f conftest_grp; then + ptygrp=`cat conftest_grp` + echo "- pty mode: 0620, group: $ptygrp" + AC_DEFINE(PTYMODE, 0620) + AC_DEFINE_UNQUOTED(PTYGROUP,$ptygrp) + else + echo "- ptys are world accessable" + fi + ], + echo "- can't determine - assume ptys are world accessable" + ) + rm -f conftest_grp dnl dnl **** utmp handling **** *************** *** 509,516 **** dnl linux has a void pututline, grrr, gcc will error when evaluating it. AC_COMPILE_CHECK(getutent, [#include <time.h> /* to get time_t on SCO */ #include <sys/types.h> ! #ifdef SVR4 #include <utmpx.h> #else #include <utmp.h> #endif --- 605,613 ---- dnl linux has a void pututline, grrr, gcc will error when evaluating it. AC_COMPILE_CHECK(getutent, [#include <time.h> /* to get time_t on SCO */ #include <sys/types.h> ! #if defined(SVR4) && !defined(DGUX) #include <utmpx.h> + #define utmp utmpx #else #include <utmp.h> #endif *************** *** 521,527 **** [int x = DEAD_PROCESS; struct utmp *y = pututline((struct utmp *)0); getutent();], AC_DEFINE(GETUTENT)) AC_COMPILE_CHECK(ut_host, [#include <time.h> #include <sys/types.h> ! #ifdef SVR4 #include <utmpx.h> #define utmp utmpx #else --- 618,624 ---- [int x = DEAD_PROCESS; struct utmp *y = pututline((struct utmp *)0); getutent();], AC_DEFINE(GETUTENT)) AC_COMPILE_CHECK(ut_host, [#include <time.h> #include <sys/types.h> ! #if defined(SVR4) && !defined(DGUX) #include <utmpx.h> #define utmp utmpx #else *************** *** 565,571 **** AC_DEFINE(NLIST_NAME_UNION))]) echo checking for nlist declaration ! AC_PROGRAM_EGREP([nlist( | )( | )*.*\(],[ #ifdef NLIST_STRUCT # include <nlist.h> #else --- 662,668 ---- AC_DEFINE(NLIST_NAME_UNION))]) echo checking for nlist declaration ! AC_PROGRAM_EGREP([nlist(( | )( | )*.*\(|\()],[ #ifdef NLIST_STRUCT # include <nlist.h> #else *************** *** 653,660 **** dnl dnl **** signal handling **** dnl ! AC_HEADER_EGREP([(void|sighandler_t).*signal], signal.h, ! AC_DEFINE(SIGVOID)) AC_COMPILE_CHECK(sigset, [ #include <sys/types.h> #include <signal.h> --- 750,762 ---- dnl dnl **** signal handling **** dnl ! AC_COMPILE_CHECK([return type of signal handlers], ! [#include <sys/types.h> ! #include <signal.h> ! #ifdef signal ! #undef signal ! #endif ! extern void (*signal ()) ();], [int i;], AC_DEFINE(SIGVOID)) AC_COMPILE_CHECK(sigset, [ #include <sys/types.h> #include <signal.h> *************** *** 670,677 **** #include <sys/types.h> #include <signal.h> ! #ifndef SIGCHLD ! #define SIGCHLD SIGCLD #endif #ifdef USESIGSET #define signal sigset --- 772,779 ---- #include <sys/types.h> #include <signal.h> ! #ifndef SIGCLD ! #define SIGCLD SIGCHLD #endif #ifdef USESIGSET #define signal sigset *************** *** 689,697 **** main() { ! (void)signal(SIGCHLD, hand); ! kill(getpid(), SIGCHLD); ! kill(getpid(), SIGCHLD); if (got < 2) exit(1); exit(0); --- 791,799 ---- main() { ! (void)signal(SIGCLD, hand); ! kill(getpid(), SIGCLD); ! kill(getpid(), SIGCLD); if (got < 2) exit(1); exit(0); *************** *** 727,735 **** if test -z "$butterfly"; then echo checking for termio or termios ! AC_TEST_CPP([#include <termios.h>], AC_DEFINE(TERMIO), ! AC_TEST_CPP([#include <termio.h>], AC_DEFINE(TERMIO))) fi dnl AC_HEADER_CHECK(shadow.h, AC_DEFINE(SHADOWPW)) AC_COMPILE_CHECK(getspnam, [#include <shadow.h>], [getspnam("x");], --- 829,840 ---- if test -z "$butterfly"; then echo checking for termio or termios ! AC_TEST_CPP([#include <termio.h>], AC_DEFINE(TERMIO), ! if test -n "$posix"; then ! AC_TEST_CPP([#include <termios.h>], AC_DEFINE(TERMIO)) fi + ) + fi dnl AC_HEADER_CHECK(shadow.h, AC_DEFINE(SHADOWPW)) AC_COMPILE_CHECK(getspnam, [#include <shadow.h>], [getspnam("x");], *************** *** 806,811 **** --- 911,918 ---- AC_DEFINE(HAVE_DEV_PTC) fi + AC_COMPILE_CHECK(strerror, ,[strerror(0);], AC_DEFINE(HAVE_STRERROR)) + dnl dnl **** the end **** dnl *** configure.orig Thu Aug 19 17:59:17 1993 --- configure Sat Jan 15 17:22:12 1994 *************** *** 17,28 **** # along with this program; if not, write to the Free Software # Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. ! # Usage: configure [--srcdir=DIR] [--host=HOST] [--gas] [--nfp] [--no-create] ! # [--prefix=PREFIX] [--exec-prefix=PREFIX] [--with-PACKAGE] [TARGET] ! # Ignores all args except --srcdir, --prefix, --exec-prefix, --no-create, and ! # --with-PACKAGE unless this script has special code to handle it. - for arg do # Handle --exec-prefix with a space before the argument. --- 17,27 ---- # along with this program; if not, write to the Free Software # Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. ! # Usage: configure [--srcdir=DIR] [--host=HOST] [--gas] [--nfp] ! # [--prefix=PREFIX] [--exec-prefix=PREFIX] [--with-PACKAGE[=VALUE]] ! # Ignores all args except --srcdir, --prefix, --exec-prefix, and ! # --with-PACKAGE[=VALUE] unless this script has special code to handle it. for arg do # Handle --exec-prefix with a space before the argument. *************** *** 35,41 **** elif test x$next_srcdir = xyes; then srcdir=$arg; next_srcdir= else case $arg in ! # For backward compatibility, also recognize exact --exec_prefix. -exec-prefix=* | --exec_prefix=* | --exec-prefix=* | --exec-prefi=* | --exec-pref=* | --exec-pre=* | --exec-pr=* | --exec-p=* | --exec-=* | --exec=* | --exe=* | --ex=* | --e=*) exec_prefix=`echo $arg | sed 's/[-a-z_]*=//'` ;; -exec-prefix | --exec_prefix | --exec-prefix | --exec-prefi | --exec-pref | --exec-pre | --exec-pr | --exec-p | --exec- | --exec | --exe | --ex | --e) --- 34,40 ---- elif test x$next_srcdir = xyes; then srcdir=$arg; next_srcdir= else case $arg in ! # For backward compatibility, recognize -exec-prefix and --exec_prefix. -exec-prefix=* | --exec_prefix=* | --exec-prefix=* | --exec-prefi=* | --exec-pref=* | --exec-pre=* | --exec-pr=* | --exec-p=* | --exec-=* | --exec=* | --exe=* | --ex=* | --e=*) exec_prefix=`echo $arg | sed 's/[-a-z_]*=//'` ;; -exec-prefix | --exec_prefix | --exec-prefix | --exec-prefi | --exec-pref | --exec-pre | --exec-pr | --exec-p | --exec- | --exec | --exe | --ex | --e) *************** *** 49,57 **** -nfp | --nfp | --nf) ;; - -no-create | --no-create | --no-creat | --no-crea | --no-cre | --no-cr | --no-c | --no- | --no) - no_create=1 ;; - -prefix=* | --prefix=* | --prefi=* | --pref=* | --pre=* | --pr=* | --p=*) prefix=`echo $arg | sed 's/[-a-z_]*=//'` ;; -prefix | --prefix | --prefi | --pref | --pre | --pr | --p) --- 48,53 ---- *************** *** 63,74 **** next_srcdir=yes ;; -with-* | --with-*) ! package=`echo $arg|sed 's/-*with-//'` ! # Delete all the valid chars; see if any are left. ! if test -n "`echo $package|sed 's/[-a-zA-Z0-9_]*//g'`"; then echo "configure: $package: invalid package name" >&2; exit 1 fi ! eval "with_`echo $package|sed s/-/_/g`=1" ;; -v | -verbose | --verbose | --verbos | --verbo | --verb | --ver | --ve | --v) verbose=yes ;; --- 59,75 ---- next_srcdir=yes ;; -with-* | --with-*) ! package=`echo $arg|sed -e 's/-*with-//' -e 's/=.*//'` ! # Reject names that aren't valid shell variable names. ! if test -n "`echo $package| sed 's/[-a-zA-Z0-9_]//g'`"; then echo "configure: $package: invalid package name" >&2; exit 1 fi ! package=`echo $package| sed 's/-/_/g'` ! case "$arg" in ! *=*) val="`echo $arg|sed 's/[^=]*=//'`" ;; ! *) val=1 ;; ! esac ! eval "with_$package='$val'" ;; -v | -verbose | --verbose | --verbos | --verbo | --verb | --ver | --ve | --v) verbose=yes ;; *************** *** 78,84 **** fi done ! trap 'rm -fr conftest* core; exit 1' 1 3 15 # NLS nuisances. # These must not be set unconditionally because not all systems understand --- 79,86 ---- fi done ! trap 'rm -fr conftest* confdefs* core; exit 1' 1 3 15 ! trap 'rm -f confdefs*' 0 # NLS nuisances. # These must not be set unconditionally because not all systems understand *************** *** 86,93 **** if test "${LC_ALL+set}" = 'set' ; then LC_ALL=C; export LC_ALL; fi if test "${LANG+set}" = 'set' ; then LANG=C; export LANG; fi ! rm -f conftest* ! compile='${CC-cc} $CFLAGS $DEFS conftest.c -o conftest $LIBS >/dev/null 2>&1' # A filename unique to this package, relative to the directory that # configure is in, which we can look for to find out if srcdir is correct. --- 88,97 ---- if test "${LC_ALL+set}" = 'set' ; then LC_ALL=C; export LC_ALL; fi if test "${LANG+set}" = 'set' ; then LANG=C; export LANG; fi ! rm -f conftest* confdefs.h ! # AIX cpp loses on an empty file, so make sure it contains at least a newline. ! echo > confdefs.h ! compile='${CC-cc} $CFLAGS conftest.c -o conftest $LIBS >/dev/null 2>&1' # A filename unique to this package, relative to the directory that # configure is in, which we can look for to find out if srcdir is correct. *************** *** 120,125 **** --- 124,130 ---- *) srcdir=`cd $srcdir; pwd` ;; # Make relative path absolute. esac + # Save the original args to write them into config.status later. configure_args="$*" *************** *** 148,154 **** IFS="$saveifs" fi test -z "$CC" && CC="cc" ! test -n "$CC" -a -n "$verbose" && echo " setting CC to $CC" # Find out if we are using GNU C, under whatever name. cat > conftest.c <<EOF --- 153,159 ---- IFS="$saveifs" fi test -z "$CC" && CC="cc" ! test -n "$CC" && test -n "$verbose" && echo " setting CC to $CC" # Find out if we are using GNU C, under whatever name. cat > conftest.c <<EOF *************** *** 169,181 **** # make. It must be expanded now. CPP="${CC-cc} -E" cat > conftest.c <<EOF #include <stdio.h> Syntax Error EOF ! err=`eval "($CPP \$DEFS conftest.c >/dev/null) 2>&1"` if test -z "$err"; then : else CPP=/lib/cpp fi rm -f conftest* --- 174,188 ---- # make. It must be expanded now. CPP="${CC-cc} -E" cat > conftest.c <<EOF + #include "confdefs.h" #include <stdio.h> Syntax Error EOF ! err=`eval "($CPP conftest.c >/dev/null) 2>&1"` if test -z "$err"; then : else + rm -rf conftest* CPP=/lib/cpp fi rm -f conftest* *************** *** 188,198 **** prog='#include <sgtty.h> Autoconf TIOCGETP' cat > conftest.c <<EOF $prog EOF ! eval "$CPP \$DEFS conftest.c > conftest.out 2>&1" if egrep "$pattern" conftest.out >/dev/null 2>&1; then need_trad=1 fi rm -f conftest* --- 195,208 ---- prog='#include <sgtty.h> Autoconf TIOCGETP' cat > conftest.c <<EOF + #include "confdefs.h" $prog EOF ! eval "$CPP conftest.c > conftest.out 2>&1" if egrep "$pattern" conftest.out >/dev/null 2>&1; then + rm -rf conftest* need_trad=1 + fi rm -f conftest* *************** *** 201,211 **** prog='#include <termio.h> Autoconf TCGETA' cat > conftest.c <<EOF $prog EOF ! eval "$CPP \$DEFS conftest.c > conftest.out 2>&1" if egrep "$pattern" conftest.out >/dev/null 2>&1; then need_trad=1 fi rm -f conftest* --- 211,224 ---- prog='#include <termio.h> Autoconf TCGETA' cat > conftest.c <<EOF + #include "confdefs.h" $prog EOF ! eval "$CPP conftest.c > conftest.out 2>&1" if egrep "$pattern" conftest.out >/dev/null 2>&1; then + rm -rf conftest* need_trad=1 + fi rm -f conftest* *************** *** 222,227 **** --- 235,241 ---- { test -n "$verbose" && \ echo " defining _POSIX_SOURCE" + echo "#define" _POSIX_SOURCE 1 >> confdefs.h DEFS="$DEFS -D_POSIX_SOURCE=1" SEDDEFS="${SEDDEFS}\${SEDdA}_POSIX_SOURCE\${SEDdB}_POSIX_SOURCE\${SEDdC}1\${SEDdD} \${SEDuA}_POSIX_SOURCE\${SEDuB}_POSIX_SOURCE\${SEDuC}1\${SEDuD} *************** *** 239,244 **** --- 253,259 ---- cat > conftest.c <<EOF + #include "confdefs.h" main(){exit(0);} EOF eval $compile *************** *** 247,253 **** else echo "Can't run the compiler - sorry";exit fi ! rm -f conftest* for p in mawk gawk nawk awk do if test -z "$AWK"; then --- 262,286 ---- else echo "Can't run the compiler - sorry";exit fi ! rm -fr conftest* ! ! cat > conftest.c <<EOF ! #include "confdefs.h" ! ! main() ! { ! int __something_strange_(); ! __something_strange_(0); ! } ! ! EOF ! eval $compile ! if test -s conftest && (./conftest; exit) 2>/dev/null; then ! echo "Your compiler does not set the exit status - sorry";exit ! ! fi ! rm -fr conftest* ! for p in mawk gawk nawk awk do if test -z "$AWK"; then *************** *** 265,271 **** IFS="$saveifs" fi ! test -n "$AWK" -a -n "$verbose" && echo " setting AWK to $AWK" test -n "$AWK" && break done --- 298,304 ---- IFS="$saveifs" fi ! test -n "$AWK" && test -n "$verbose" && echo " setting AWK to $AWK" test -n "$AWK" && break done *************** *** 308,315 **** --- 341,351 ---- IFS="$saveifs" fi INSTALL=${INSTALL-cp} + test -n "$verbose" && echo " setting INSTALL to $INSTALL" INSTALL_PROGRAM=${INSTALL_PROGRAM-'$(INSTALL)'} + test -n "$verbose" && echo " setting INSTALL_PROGRAM to $INSTALL_PROGRAM" INSTALL_DATA=${INSTALL_DATA-'$(INSTALL)'} + test -n "$verbose" && echo " setting INSTALL_DATA to $INSTALL_DATA" if test -n "$ISC"; then *************** *** 317,322 **** --- 353,359 ---- { test -n "$verbose" && \ echo " defining ISC" + echo "#define" ISC 1 >> confdefs.h DEFS="$DEFS -DISC=1" SEDDEFS="${SEDDEFS}\${SEDdA}ISC\${SEDdB}ISC\${SEDdC}1\${SEDdD} \${SEDuA}ISC\${SEDuB}ISC\${SEDuC}1\${SEDuD} *************** *** 326,356 **** LIBS="$LIBS -linet" fi - echo checking for OSF1 - if test -f /bin/uname ; then - if test `/bin/uname` = OSF1 || test -f /osf_boot; then - - { - test -n "$verbose" && \ - echo " defining OSF1" - DEFS="$DEFS -DOSF1=1" - SEDDEFS="${SEDDEFS}\${SEDdA}OSF1\${SEDdB}OSF1\${SEDdC}1\${SEDdD} - \${SEDuA}OSF1\${SEDuB}OSF1\${SEDuC}1\${SEDuD} - \${SEDeA}OSF1\${SEDeB}OSF1\${SEDeC}1\${SEDeD} - " - } - # this disables MIPS again.... - fi - fi echo checking for MIPS if test -f /lib/libmld.a || test -f /usr/lib/libmld.a || test -f /usr/lib/cmplrs/cc/libmld.a; then ! LIBS="$LIBS -lmld" # for nlist. if test -r /dev/ptc; then { test -n "$verbose" && \ echo " defining MIPS" DEFS="$DEFS -DMIPS=1" SEDDEFS="${SEDDEFS}\${SEDdA}MIPS\${SEDdB}MIPS\${SEDdC}1\${SEDdD} \${SEDuA}MIPS\${SEDuB}MIPS\${SEDuC}1\${SEDuD} --- 363,378 ---- LIBS="$LIBS -linet" fi echo checking for MIPS if test -f /lib/libmld.a || test -f /usr/lib/libmld.a || test -f /usr/lib/cmplrs/cc/libmld.a; then ! test -f /bin/mx || LIBS="$LIBS -lmld" # for nlist. But not on alpha. if test -r /dev/ptc; then { test -n "$verbose" && \ echo " defining MIPS" + echo "#define" MIPS 1 >> confdefs.h DEFS="$DEFS -DMIPS=1" SEDDEFS="${SEDDEFS}\${SEDdA}MIPS\${SEDdB}MIPS\${SEDdC}1\${SEDdD} \${SEDuA}MIPS\${SEDuB}MIPS\${SEDuC}1\${SEDuD} *************** *** 360,365 **** --- 382,388 ---- echo checking for wait3 cat > conftest.c <<EOF + #include "confdefs.h" int main() { exit(0); } int t() { wait3(); } *************** *** 367,383 **** --- 390,410 ---- if eval $compile; then : else + rm -rf conftest* echo checking for wait2 cat > conftest.c <<EOF + #include "confdefs.h" int main() { exit(0); } int t() { wait2(); } EOF if eval $compile; then + rm -rf conftest* { test -n "$verbose" && \ echo " defining USE_WAIT2" + echo "#define" USE_WAIT2 1 >> confdefs.h DEFS="$DEFS -DUSE_WAIT2=1" SEDDEFS="${SEDDEFS}\${SEDdA}USE_WAIT2\${SEDdB}USE_WAIT2\${SEDdC}1\${SEDdD} \${SEDuA}USE_WAIT2\${SEDuB}USE_WAIT2\${SEDuC}1\${SEDuD} *************** *** 386,391 **** --- 413,419 ---- } LIBS="$LIBS -lbsd" ; CC="$CC -I/usr/include/bsd" + fi rm -f conftest* *************** *** 397,424 **** echo checking for Ultrix cat > conftest.c <<EOF #if defined(ultrix) || defined(__ultrix) yes #endif EOF ! eval "$CPP \$DEFS conftest.c > conftest.out 2>&1" if egrep "yes" conftest.out >/dev/null 2>&1; then ULTRIX=1 fi rm -f conftest* echo checking for butterfly cat > conftest.c <<EOF #if defined(butterfly) yes #endif EOF ! eval "$CPP \$DEFS conftest.c > conftest.out 2>&1" if egrep "yes" conftest.out >/dev/null 2>&1; then butterfly=1 fi rm -f conftest* --- 425,491 ---- echo checking for Ultrix cat > conftest.c <<EOF + #include "confdefs.h" #if defined(ultrix) || defined(__ultrix) yes #endif EOF ! eval "$CPP conftest.c > conftest.out 2>&1" if egrep "yes" conftest.out >/dev/null 2>&1; then + rm -rf conftest* ULTRIX=1 + + fi + rm -f conftest* + + + if test -f /usr/lib/libpyr.a ; then + oldlibs="$LIBS" + LIBS="$LIBS -lpyr" + echo checking for Pyramid OSX + cat > conftest.c <<EOF + #include "confdefs.h" + + int main() { exit(0); } + int t() { open_controlling_pty(""); } + EOF + if eval $compile; then + rm -rf conftest* + + { + test -n "$verbose" && \ + echo " defining OSX" + echo "#define" OSX 1 >> confdefs.h + DEFS="$DEFS -DOSX=1" + SEDDEFS="${SEDDEFS}\${SEDdA}OSX\${SEDdB}OSX\${SEDdC}1\${SEDdD} + \${SEDuA}OSX\${SEDuB}OSX\${SEDuC}1\${SEDuD} + \${SEDeA}OSX\${SEDeB}OSX\${SEDeC}1\${SEDeD} + " + } + + + else + rm -rf conftest* + LIBS="oldlibs" fi rm -f conftest* + fi echo checking for butterfly cat > conftest.c <<EOF + #include "confdefs.h" #if defined(butterfly) yes #endif EOF ! eval "$CPP conftest.c > conftest.out 2>&1" if egrep "yes" conftest.out >/dev/null 2>&1; then + rm -rf conftest* butterfly=1 + fi rm -f conftest* *************** *** 429,434 **** --- 496,502 ---- fi echo checking for POSIX.1 cat > conftest.c <<EOF + #include "confdefs.h" #include <sys/types.h> #include <unistd.h> main () { *************** *** 437,448 **** #endif EOF ! eval "$CPP \$DEFS conftest.c > conftest.out 2>&1" if egrep "yes" conftest.out >/dev/null 2>&1; then ! echo "- you have a POSIX system"; { test -n "$verbose" && \ echo " defining POSIX" DEFS="$DEFS -DPOSIX=1" SEDDEFS="${SEDDEFS}\${SEDdA}POSIX\${SEDdB}POSIX\${SEDdC}1\${SEDdD} \${SEDuA}POSIX\${SEDuB}POSIX\${SEDuC}1\${SEDuD} --- 505,518 ---- #endif EOF ! eval "$CPP conftest.c > conftest.out 2>&1" if egrep "yes" conftest.out >/dev/null 2>&1; then ! rm -rf conftest* ! echo "- you have a POSIX system";posix=1; { test -n "$verbose" && \ echo " defining POSIX" + echo "#define" POSIX 1 >> confdefs.h DEFS="$DEFS -DPOSIX=1" SEDDEFS="${SEDDEFS}\${SEDdA}POSIX\${SEDdB}POSIX\${SEDdC}1\${SEDdD} \${SEDuA}POSIX\${SEDuB}POSIX\${SEDuC}1\${SEDuD} *************** *** 450,455 **** --- 520,526 ---- " } + fi rm -f conftest* *************** *** 457,462 **** --- 528,534 ---- echo checking for System V cat > conftest.c <<EOF + #include "confdefs.h" #include <sys/types.h> #include <signal.h> #include <fcntl.h> *************** *** 466,475 **** --- 538,549 ---- if eval $compile; then : else + rm -rf conftest* { test -n "$verbose" && \ echo " defining SYSV" + echo "#define" SYSV 1 >> confdefs.h DEFS="$DEFS -DSYSV=1" SEDDEFS="${SEDDEFS}\${SEDdA}SYSV\${SEDdB}SYSV\${SEDdC}1\${SEDdD} \${SEDuA}SYSV\${SEDuB}SYSV\${SEDuC}1\${SEDuD} *************** *** 483,496 **** echo checking for sequent/ptx cat > conftest.c <<EOF #ifdef _SEQUENT_ yes #endif EOF ! eval "$CPP \$DEFS conftest.c > conftest.out 2>&1" if egrep "yes" conftest.out >/dev/null 2>&1; then LIBS="$LIBS -lsocket -linet";seqptx=1 fi rm -f conftest* --- 557,573 ---- echo checking for sequent/ptx cat > conftest.c <<EOF + #include "confdefs.h" #ifdef _SEQUENT_ yes #endif EOF ! eval "$CPP conftest.c > conftest.out 2>&1" if egrep "yes" conftest.out >/dev/null 2>&1; then + rm -rf conftest* LIBS="$LIBS -lsocket -linet";seqptx=1 + fi rm -f conftest* *************** *** 499,519 **** LIBS="$LIBS -lelf" echo checking for SVR4 cat > conftest.c <<EOF int main() { exit(0); } int t() { } EOF if eval $compile; then echo checking for dwarf.h cat > conftest.c <<EOF #include <dwarf.h> EOF ! err=`eval "($CPP \$DEFS conftest.c >/dev/null) 2>&1"` if test -z "$err"; then { test -n "$verbose" && \ echo " defining SVR4" DEFS="$DEFS -DSVR4=1" SEDDEFS="${SEDDEFS}\${SEDdA}SVR4\${SEDdB}SVR4\${SEDdC}1\${SEDdD} \${SEDuA}SVR4\${SEDuB}SVR4\${SEDuC}1\${SEDuD} --- 576,601 ---- LIBS="$LIBS -lelf" echo checking for SVR4 cat > conftest.c <<EOF + #include "confdefs.h" int main() { exit(0); } int t() { } EOF if eval $compile; then + rm -rf conftest* echo checking for dwarf.h cat > conftest.c <<EOF + #include "confdefs.h" #include <dwarf.h> EOF ! err=`eval "($CPP conftest.c >/dev/null) 2>&1"` if test -z "$err"; then + rm -rf conftest* { test -n "$verbose" && \ echo " defining SVR4" + echo "#define" SVR4 1 >> confdefs.h DEFS="$DEFS -DSVR4=1" SEDDEFS="${SEDDEFS}\${SEDdA}SVR4\${SEDdB}SVR4\${SEDdC}1\${SEDdD} \${SEDuA}SVR4\${SEDuB}SVR4\${SEDuC}1\${SEDuD} *************** *** 524,529 **** --- 606,612 ---- { test -n "$verbose" && \ echo " defining BUGGYGETLOGIN" + echo "#define" BUGGYGETLOGIN 1 >> confdefs.h DEFS="$DEFS -DBUGGYGETLOGIN=1" SEDDEFS="${SEDDEFS}\${SEDdA}BUGGYGETLOGIN\${SEDdB}BUGGYGETLOGIN\${SEDdC}1\${SEDdD} \${SEDuA}BUGGYGETLOGIN\${SEDuB}BUGGYGETLOGIN\${SEDuC}1\${SEDuD} *************** *** 531,547 **** " } else echo checking for elf.h cat > conftest.c <<EOF #include <elf.h> EOF ! err=`eval "($CPP \$DEFS conftest.c >/dev/null) 2>&1"` if test -z "$err"; then { test -n "$verbose" && \ echo " defining SVR4" DEFS="$DEFS -DSVR4=1" SEDDEFS="${SEDDEFS}\${SEDdA}SVR4\${SEDdB}SVR4\${SEDdC}1\${SEDdD} \${SEDuA}SVR4\${SEDuB}SVR4\${SEDuC}1\${SEDuD} --- 614,635 ---- " } + else + rm -rf conftest* echo checking for elf.h cat > conftest.c <<EOF + #include "confdefs.h" #include <elf.h> EOF ! err=`eval "($CPP conftest.c >/dev/null) 2>&1"` if test -z "$err"; then + rm -rf conftest* { test -n "$verbose" && \ echo " defining SVR4" + echo "#define" SVR4 1 >> confdefs.h DEFS="$DEFS -DSVR4=1" SEDDEFS="${SEDDEFS}\${SEDdA}SVR4\${SEDdB}SVR4\${SEDdC}1\${SEDdD} \${SEDuA}SVR4\${SEDuB}SVR4\${SEDuC}1\${SEDuD} *************** *** 552,557 **** --- 640,646 ---- { test -n "$verbose" && \ echo " defining BUGGYGETLOGIN" + echo "#define" BUGGYGETLOGIN 1 >> confdefs.h DEFS="$DEFS -DBUGGYGETLOGIN=1" SEDDEFS="${SEDDEFS}\${SEDdA}BUGGYGETLOGIN\${SEDdB}BUGGYGETLOGIN\${SEDdC}1\${SEDdD} \${SEDuA}BUGGYGETLOGIN\${SEDuB}BUGGYGETLOGIN\${SEDuC}1\${SEDuD} *************** *** 559,564 **** --- 648,654 ---- " } + fi rm -f conftest* *************** *** 566,572 **** --- 656,664 ---- rm -f conftest* + else + rm -rf conftest* LIBS="$oldlibs" fi rm -f conftest* *************** *** 575,589 **** echo checking for pid_t cat > conftest.c <<EOF #include <sys/types.h> EOF ! eval "$CPP \$DEFS conftest.c > conftest.out 2>&1" if egrep "pid_t" conftest.out >/dev/null 2>&1; then { test -n "$verbose" && \ echo " defining PID_T_DEFINED" DEFS="$DEFS -DPID_T_DEFINED=1" SEDDEFS="${SEDDEFS}\${SEDdA}PID_T_DEFINED\${SEDdB}PID_T_DEFINED\${SEDdC}1\${SEDdD} \${SEDuA}PID_T_DEFINED\${SEDuB}PID_T_DEFINED\${SEDuC}1\${SEDuD} --- 667,684 ---- echo checking for pid_t cat > conftest.c <<EOF + #include "confdefs.h" #include <sys/types.h> EOF ! eval "$CPP conftest.c > conftest.out 2>&1" if egrep "pid_t" conftest.out >/dev/null 2>&1; then + rm -rf conftest* { test -n "$verbose" && \ echo " defining PID_T_DEFINED" + echo "#define" PID_T_DEFINED 1 >> confdefs.h DEFS="$DEFS -DPID_T_DEFINED=1" SEDDEFS="${SEDDEFS}\${SEDdA}PID_T_DEFINED\${SEDdB}PID_T_DEFINED\${SEDdC}1\${SEDdD} \${SEDuA}PID_T_DEFINED\${SEDuB}PID_T_DEFINED\${SEDuC}1\${SEDuD} *************** *** 591,612 **** " } fi rm -f conftest* echo checking for sig_t cat > conftest.c <<EOF #include <sys/types.h> #include <signal.h> EOF ! eval "$CPP \$DEFS conftest.c > conftest.out 2>&1" if egrep "sig_t" conftest.out >/dev/null 2>&1; then { test -n "$verbose" && \ echo " defining SIG_T_DEFINED" DEFS="$DEFS -DSIG_T_DEFINED=1" SEDDEFS="${SEDDEFS}\${SEDdA}SIG_T_DEFINED\${SEDdB}SIG_T_DEFINED\${SEDdC}1\${SEDdD} \${SEDuA}SIG_T_DEFINED\${SEDuB}SIG_T_DEFINED\${SEDuC}1\${SEDuD} --- 686,711 ---- " } + fi rm -f conftest* echo checking for sig_t cat > conftest.c <<EOF + #include "confdefs.h" #include <sys/types.h> #include <signal.h> EOF ! eval "$CPP conftest.c > conftest.out 2>&1" if egrep "sig_t" conftest.out >/dev/null 2>&1; then + rm -rf conftest* { test -n "$verbose" && \ echo " defining SIG_T_DEFINED" + echo "#define" SIG_T_DEFINED 1 >> confdefs.h DEFS="$DEFS -DSIG_T_DEFINED=1" SEDDEFS="${SEDDEFS}\${SEDdA}SIG_T_DEFINED\${SEDdB}SIG_T_DEFINED\${SEDdC}1\${SEDdD} \${SEDuA}SIG_T_DEFINED\${SEDuB}SIG_T_DEFINED\${SEDuC}1\${SEDuD} *************** *** 614,634 **** " } fi rm -f conftest* echo checking for uid_t cat > conftest.c <<EOF #include <sys/types.h> EOF ! eval "$CPP \$DEFS conftest.c > conftest.out 2>&1" if egrep "uid_t" conftest.out >/dev/null 2>&1; then { test -n "$verbose" && \ echo " defining UID_T_DEFINED" DEFS="$DEFS -DUID_T_DEFINED=1" SEDDEFS="${SEDDEFS}\${SEDdA}UID_T_DEFINED\${SEDdB}UID_T_DEFINED\${SEDdC}1\${SEDdD} \${SEDuA}UID_T_DEFINED\${SEDuB}UID_T_DEFINED\${SEDuC}1\${SEDuD} --- 713,737 ---- " } + fi rm -f conftest* echo checking for uid_t cat > conftest.c <<EOF + #include "confdefs.h" #include <sys/types.h> EOF ! eval "$CPP conftest.c > conftest.out 2>&1" if egrep "uid_t" conftest.out >/dev/null 2>&1; then + rm -rf conftest* { test -n "$verbose" && \ echo " defining UID_T_DEFINED" + echo "#define" UID_T_DEFINED 1 >> confdefs.h DEFS="$DEFS -DUID_T_DEFINED=1" SEDDEFS="${SEDDEFS}\${SEDdA}UID_T_DEFINED\${SEDdB}UID_T_DEFINED\${SEDdC}1\${SEDdD} \${SEDuA}UID_T_DEFINED\${SEDuB}UID_T_DEFINED\${SEDuC}1\${SEDuD} *************** *** 636,641 **** --- 739,745 ---- " } + fi rm -f conftest* *************** *** 643,648 **** --- 747,753 ---- echo checking for BSD job control cat > conftest.c <<EOF + #include "confdefs.h" #include <sys/types.h> #include <sys/ioctl.h> *************** *** 661,670 **** --- 766,777 ---- } EOF if eval $compile; then + rm -rf conftest* echo "- you have jobcontrol" { test -n "$verbose" && \ echo " defining BSDJOBS" + echo "#define" BSDJOBS 1 >> confdefs.h DEFS="$DEFS -DBSDJOBS=1" SEDDEFS="${SEDDEFS}\${SEDdA}BSDJOBS\${SEDdB}BSDJOBS\${SEDdC}1\${SEDdD} \${SEDuA}BSDJOBS\${SEDuB}BSDJOBS\${SEDuC}1\${SEDuD} *************** *** 672,678 **** --- 779,787 ---- " } + else + rm -rf conftest* echo "- you don't have jobcontrol" fi rm -f conftest* *************** *** 680,685 **** --- 789,795 ---- echo checking for setreuid cat > conftest.c <<EOF + #include "confdefs.h" int main() { exit(0); } int t() { *************** *** 693,702 **** --- 803,814 ---- if eval $compile; then : else + rm -rf conftest* { test -n "$verbose" && \ echo " defining NOREUID" + echo "#define" NOREUID 1 >> confdefs.h DEFS="$DEFS -DNOREUID=1" SEDDEFS="${SEDDEFS}\${SEDdA}NOREUID\${SEDdB}NOREUID\${SEDdC}1\${SEDdD} \${SEDuA}NOREUID\${SEDuB}NOREUID\${SEDuC}1\${SEDuD} *************** *** 712,717 **** --- 824,830 ---- echo checking for select cat > conftest.c <<EOF + #include "confdefs.h" int main() { exit(0); } int t() { select(0, 0, 0, 0, 0); } *************** *** 719,727 **** --- 832,842 ---- if eval $compile; then : else + rm -rf conftest* LIBS="$LIBS -lnet -lnsl" echo checking for select with $LIBS cat > conftest.c <<EOF + #include "confdefs.h" int main() { exit(0); } int t() { select(0, 0, 0, 0, 0); } *************** *** 729,734 **** --- 844,850 ---- if eval $compile; then : else + rm -rf conftest* echo '!!! no select - no screen';exit fi rm -f conftest* *************** *** 740,745 **** --- 856,862 ---- echo checking fifos cat > conftest.c <<EOF + #include "confdefs.h" #include <sys/types.h> #include <sys/stat.h> *************** *** 760,774 **** int f; (void)alarm(5); if (mknod(fin, S_IFIFO|0777, 0)) exit(1); if (stat(fin, &stb) || (stb.st_mode & S_IFIFO) != S_IFIFO) exit(1); close(0); ! if (open(fin, O_RDWR | O_NDELAY)) ! exit(1); ! if (write(0, "TEST", 4) == -1) exit(1); f = 1; if (select(1, &f, 0, 0, 0) == -1) exit(1); --- 877,913 ---- int f; (void)alarm(5); + #ifdef POSIX + if (mkfifo(fin, 0777)) + #else if (mknod(fin, S_IFIFO|0777, 0)) + #endif exit(1); if (stat(fin, &stb) || (stb.st_mode & S_IFIFO) != S_IFIFO) exit(1); close(0); ! #ifdef __386BSD__ ! /* ! * The next test fails under 386BSD, but screen works using fifos. ! * Fifos in O_RDWR mode are only used for the BROKEN_PIPE case and for ! * the select() configuration test. ! */ ! exit(0); ! #endif ! if (open(fin, O_RDONLY | O_NDELAY)) exit(1); + if (fork() == 0) + { + close(0); + if (open(fin, O_WRONLY | O_NDELAY)) + exit(1); + close(0); + if (open(fin, O_WRONLY | O_NDELAY)) + exit(1); + if (write(0, "TEST", 4) == -1) + exit(1); + exit(0); + } f = 1; if (select(1, &f, 0, 0, 0) == -1) exit(1); *************** *** 779,793 **** eval $compile if test -s conftest && (./conftest; exit) 2>/dev/null; then echo "- your fifos are usable"; fifo=1 else echo "- your fifos are not usable" fi ! rm -f conftest* rm -f /tmp/conftest* if test -n "$fifo"; then echo "checking for broken fifo implementation" cat > conftest.c <<EOF #include <sys/types.h> #include <fcntl.h> --- 918,934 ---- eval $compile if test -s conftest && (./conftest; exit) 2>/dev/null; then echo "- your fifos are usable"; fifo=1 + else echo "- your fifos are not usable" fi ! rm -fr conftest* rm -f /tmp/conftest* if test -n "$fifo"; then echo "checking for broken fifo implementation" cat > conftest.c <<EOF + #include "confdefs.h" #include <sys/types.h> #include <fcntl.h> *************** *** 808,814 **** --- 949,959 ---- struct timeval tv; int r, x; + #ifdef POSIX + if (mkfifo(fin, 0600)) + #else if (mknod(fin, S_IFIFO|0600, 0)) + #endif exit(1); close(0); if (open(fin, O_RDONLY|O_NDELAY)) *************** *** 825,835 **** --- 970,982 ---- eval $compile if test -s conftest && (./conftest; exit) 2>/dev/null; then echo "- your implementation is ok" + else echo "- you have a broken implementation" { test -n "$verbose" && \ echo " defining BROKEN_PIPE" + echo "#define" BROKEN_PIPE 1 >> confdefs.h DEFS="$DEFS -DBROKEN_PIPE=1" SEDDEFS="${SEDDEFS}\${SEDdA}BROKEN_PIPE\${SEDdB}BROKEN_PIPE\${SEDdC}1\${SEDdD} \${SEDuA}BROKEN_PIPE\${SEDuB}BROKEN_PIPE\${SEDuC}1\${SEDuD} *************** *** 838,850 **** } fifobr=1 fi ! rm -f conftest* rm -f /tmp/conftest* fi echo checking sockets cat > conftest.c <<EOF #include <sys/types.h> #include <sys/socket.h> --- 985,998 ---- } fifobr=1 fi ! rm -fr conftest* rm -f /tmp/conftest* fi echo checking sockets cat > conftest.c <<EOF + #include "confdefs.h" #include <sys/types.h> #include <sys/socket.h> *************** *** 898,912 **** eval $compile if test -s conftest && (./conftest; exit) 2>/dev/null; then echo "- your sockets are usable"; sock=1 else echo "- your sockets are not usable" fi ! rm -f conftest* rm -f /tmp/conftest* if test -n "$sock"; then echo "checking socket implementation" cat > conftest.c <<EOF #include <sys/types.h> #include <sys/stat.h> --- 1046,1062 ---- eval $compile if test -s conftest && (./conftest; exit) 2>/dev/null; then echo "- your sockets are usable"; sock=1 + else echo "- your sockets are not usable" fi ! rm -fr conftest* rm -f /tmp/conftest* if test -n "$sock"; then echo "checking socket implementation" cat > conftest.c <<EOF + #include "confdefs.h" #include <sys/types.h> #include <sys/stat.h> *************** *** 937,948 **** --- 1087,1100 ---- eval $compile if test -s conftest && (./conftest; exit) 2>/dev/null; then echo "- you are normal" + else echo "- unix domain sockets are not kept in the filesystem" { test -n "$verbose" && \ echo " defining SOCK_NOT_IN_FS" + echo "#define" SOCK_NOT_IN_FS 1 >> confdefs.h DEFS="$DEFS -DSOCK_NOT_IN_FS=1" SEDDEFS="${SEDDEFS}\${SEDdA}SOCK_NOT_IN_FS\${SEDdB}SOCK_NOT_IN_FS\${SEDdC}1\${SEDdD} \${SEDuA}SOCK_NOT_IN_FS\${SEDuB}SOCK_NOT_IN_FS\${SEDuC}1\${SEDuD} *************** *** 951,957 **** } socknofs=1 fi ! rm -f conftest* rm -f /tmp/conftest* fi --- 1103,1109 ---- } socknofs=1 fi ! rm -fr conftest* rm -f /tmp/conftest* fi *************** *** 964,969 **** --- 1116,1122 ---- { test -n "$verbose" && \ echo " defining NAMEDPIPE" + echo "#define" NAMEDPIPE 1 >> confdefs.h DEFS="$DEFS -DNAMEDPIPE=1" SEDDEFS="${SEDDEFS}\${SEDdA}NAMEDPIPE\${SEDdB}NAMEDPIPE\${SEDdC}1\${SEDdD} \${SEDuA}NAMEDPIPE\${SEDuB}NAMEDPIPE\${SEDuC}1\${SEDuD} *************** *** 979,984 **** --- 1132,1138 ---- { test -n "$verbose" && \ echo " defining NAMEDPIPE" + echo "#define" NAMEDPIPE 1 >> confdefs.h DEFS="$DEFS -DNAMEDPIPE=1" SEDDEFS="${SEDDEFS}\${SEDdA}NAMEDPIPE\${SEDdB}NAMEDPIPE\${SEDdC}1\${SEDdD} \${SEDuA}NAMEDPIPE\${SEDuB}NAMEDPIPE\${SEDuC}1\${SEDuD} *************** *** 993,998 **** --- 1147,1153 ---- { test -n "$verbose" && \ echo " defining NAMEDPIPE" + echo "#define" NAMEDPIPE 1 >> confdefs.h DEFS="$DEFS -DNAMEDPIPE=1" SEDDEFS="${SEDDEFS}\${SEDdA}NAMEDPIPE\${SEDdB}NAMEDPIPE\${SEDdC}1\${SEDdD} \${SEDuA}NAMEDPIPE\${SEDuB}NAMEDPIPE\${SEDuC}1\${SEDuD} *************** *** 1011,1016 **** --- 1166,1172 ---- echo "checking select return value" cat > conftest.c <<EOF + #include "confdefs.h" #include <sys/types.h> #include <sys/stat.h> *************** *** 1033,1039 **** --- 1189,1199 ---- int l; (void)alarm(5); + #ifdef POSIX + if (mkfifo(nam, 0777)) + #else if (mknod(nam, S_IFIFO|0777, 0)) + #endif exit(1); close(0); if (open(nam, O_RDWR | O_NDELAY)) *************** *** 1090,1101 **** --- 1250,1263 ---- eval $compile if test -s conftest && (./conftest; exit) 2>/dev/null; then echo "- select is ok" + else echo "- it is not usable" { test -n "$verbose" && \ echo " defining SELECT_BROKEN" + echo "#define" SELECT_BROKEN 1 >> confdefs.h DEFS="$DEFS -DSELECT_BROKEN=1" SEDDEFS="${SEDDEFS}\${SEDdA}SELECT_BROKEN\${SEDdB}SELECT_BROKEN\${SEDdC}1\${SEDdD} \${SEDuA}SELECT_BROKEN\${SEDuB}SELECT_BROKEN\${SEDuC}1\${SEDuD} *************** *** 1104,1116 **** } fi ! rm -f conftest* echo searching for tgetent olibs="$LIBS" ! LIBS="-ltermcap $LIBS" echo checking for libtermcap cat > conftest.c <<EOF int main() { exit(0); } int t() { tgetent((char *)0, (char *)0); } --- 1266,1291 ---- } fi ! rm -fr conftest* echo searching for tgetent olibs="$LIBS" ! LIBS="-lcurses $olibs" ! echo checking for libcurses ! cat > conftest.c <<EOF ! #include "confdefs.h" ! ! int main() { exit(0); } ! int t() { tgetent((char *)0, (char *)0); } ! EOF ! if eval $compile; then ! : ! else ! rm -rf conftest* ! LIBS="-ltermcap $olibs" echo checking for libtermcap cat > conftest.c <<EOF + #include "confdefs.h" int main() { exit(0); } int t() { tgetent((char *)0, (char *)0); } *************** *** 1118,1126 **** if eval $compile; then : else ! LIBS="-lcurses $olibs" ! echo checking for libcurses cat > conftest.c <<EOF int main() { exit(0); } int t() { tgetent((char *)0, (char *)0); } --- 1293,1303 ---- if eval $compile; then : else ! rm -rf conftest* ! LIBS="-ltermlib $olibs" ! echo checking for libtermcap cat > conftest.c <<EOF + #include "confdefs.h" int main() { exit(0); } int t() { tgetent((char *)0, (char *)0); } *************** *** 1128,1146 **** --- 1305,1328 ---- if eval $compile; then : else + rm -rf conftest* echo "!!! no tgetent - no screen";exit fi rm -f conftest* + fi + rm -f conftest* fi rm -f conftest* + TERMCAP="xx|scrdumm:xx:" TERM=scrdumm export TERMCAP export TERM cat > conftest.c <<EOF + #include "confdefs.h" main() { *************** *** 1153,1164 **** --- 1335,1348 ---- eval $compile if test -s conftest && (./conftest; exit) 2>/dev/null; then echo "- you use the termcap database" + else echo "- you use the terminfo database" { test -n "$verbose" && \ echo " defining TERMINFO" + echo "#define" TERMINFO 1 >> confdefs.h DEFS="$DEFS -DTERMINFO=1" SEDDEFS="${SEDDEFS}\${SEDdA}TERMINFO\${SEDdB}TERMINFO\${SEDdC}1\${SEDdD} \${SEDuA}TERMINFO\${SEDuB}TERMINFO\${SEDuC}1\${SEDuD} *************** *** 1167,1175 **** } fi ! rm -f conftest* echo checking for ospeed cat > conftest.c <<EOF extern short ospeed; int main() { exit(0); } int t() { ospeed=5; } --- 1351,1360 ---- } fi ! rm -fr conftest* echo checking for ospeed cat > conftest.c <<EOF + #include "confdefs.h" extern short ospeed; int main() { exit(0); } int t() { ospeed=5; } *************** *** 1177,1186 **** --- 1362,1373 ---- if eval $compile; then : else + rm -rf conftest* { test -n "$verbose" && \ echo " defining NEED_OSPEED" + echo "#define" NEED_OSPEED 1 >> confdefs.h DEFS="$DEFS -DNEED_OSPEED=1" SEDDEFS="${SEDDEFS}\${SEDdA}NEED_OSPEED\${SEDdB}NEED_OSPEED\${SEDdC}1\${SEDdD} \${SEDuA}NEED_OSPEED\${SEDuB}NEED_OSPEED\${SEDuC}1\${SEDuD} *************** *** 1205,1211 **** { test -n "$verbose" && \ ! echo " defining PTYRANGE0 to be \"$p0\"" DEFS="$DEFS -DPTYRANGE0=\"$p0\"" SEDDEFS="${SEDDEFS}\${SEDdA}PTYRANGE0\${SEDdB}PTYRANGE0\${SEDdC}\"$p0\"\${SEDdD} \${SEDuA}PTYRANGE0\${SEDuB}PTYRANGE0\${SEDuC}\"$p0\"\${SEDuD} --- 1392,1399 ---- { test -n "$verbose" && \ ! echo " defining" PTYRANGE0 to be \"$p0\" ! echo "#define" PTYRANGE0 \"$p0\" >> confdefs.h DEFS="$DEFS -DPTYRANGE0=\"$p0\"" SEDDEFS="${SEDDEFS}\${SEDdA}PTYRANGE0\${SEDdB}PTYRANGE0\${SEDdC}\"$p0\"\${SEDdD} \${SEDuA}PTYRANGE0\${SEDuB}PTYRANGE0\${SEDuC}\"$p0\"\${SEDuD} *************** *** 1216,1222 **** { test -n "$verbose" && \ ! echo " defining PTYRANGE1 to be \"$p1\"" DEFS="$DEFS -DPTYRANGE1=\"$p1\"" SEDDEFS="${SEDDEFS}\${SEDdA}PTYRANGE1\${SEDdB}PTYRANGE1\${SEDdC}\"$p1\"\${SEDdD} \${SEDuA}PTYRANGE1\${SEDuB}PTYRANGE1\${SEDuC}\"$p1\"\${SEDuD} --- 1404,1411 ---- { test -n "$verbose" && \ ! echo " defining" PTYRANGE1 to be \"$p1\" ! echo "#define" PTYRANGE1 \"$p1\" >> confdefs.h DEFS="$DEFS -DPTYRANGE1=\"$p1\"" SEDDEFS="${SEDDEFS}\${SEDdA}PTYRANGE1\${SEDdB}PTYRANGE1\${SEDdC}\"$p1\"\${SEDdD} \${SEDuA}PTYRANGE1\${SEDuB}PTYRANGE1\${SEDuC}\"$p1\"\${SEDuD} *************** *** 1226,1238 **** fi echo checking for getutent cat > conftest.c <<EOF #include <time.h> /* to get time_t on SCO */ #include <sys/types.h> ! #ifdef SVR4 #include <utmpx.h> #else #include <utmp.h> #endif --- 1415,1505 ---- fi + echo "checking default tty permissions/group" + rm -f conftest_grp + cat > conftest.c <<EOF + #include "confdefs.h" + + #include <sys/types.h> + #include <sys/stat.h> + #include <stdio.h> + main() + { + struct stat sb; + char *x; + int om, m; + FILE *fp; + + if (!(x = ttyname(0))) exit(1); + if (stat(x, &sb)) exit(1); + om = sb.st_mode; + if (chmod(x, om & ~077)) exit(1); + m = system("mesg y"); + if (m == -1 || m == 127) exit(1); + if (stat(x, &sb)) exit(1); + m = sb.st_mode; + if (chmod(x, om)) exit(1); + if (!(m & 002)) + { + if (!(fp=fopen("conftest_grp", "w"))) + exit(1); + fprintf(fp, "%d\n", sb.st_gid); + fclose(fp); + } + exit(0); + } + + EOF + eval $compile + if test -s conftest && (./conftest; exit) 2>/dev/null; then + + if test -f conftest_grp; then + ptygrp=`cat conftest_grp` + echo "- pty mode: 0620, group: $ptygrp" + + { + test -n "$verbose" && \ + echo " defining" PTYMODE to be 0620 + echo "#define" PTYMODE 0620 >> confdefs.h + DEFS="$DEFS -DPTYMODE=0620" + SEDDEFS="${SEDDEFS}\${SEDdA}PTYMODE\${SEDdB}PTYMODE\${SEDdC}0620\${SEDdD} + \${SEDuA}PTYMODE\${SEDuB}PTYMODE\${SEDuC}0620\${SEDuD} + \${SEDeA}PTYMODE\${SEDeB}PTYMODE\${SEDeC}0620\${SEDeD} + " + } + + + { + test -n "$verbose" && \ + echo " defining" PTYGROUP to be $ptygrp + echo "#define" PTYGROUP $ptygrp >> confdefs.h + DEFS="$DEFS -DPTYGROUP=$ptygrp" + SEDDEFS="${SEDDEFS}\${SEDdA}PTYGROUP\${SEDdB}PTYGROUP\${SEDdC}$ptygrp\${SEDdD} + \${SEDuA}PTYGROUP\${SEDuB}PTYGROUP\${SEDuC}$ptygrp\${SEDuD} + \${SEDeA}PTYGROUP\${SEDeB}PTYGROUP\${SEDeC}$ptygrp\${SEDeD} + " + } + + else + echo "- ptys are world accessable" + fi + + + else + echo "- can't determine - assume ptys are world accessable" + + fi + rm -fr conftest* + rm -f conftest_grp echo checking for getutent cat > conftest.c <<EOF + #include "confdefs.h" #include <time.h> /* to get time_t on SCO */ #include <sys/types.h> ! #if defined(SVR4) && !defined(DGUX) #include <utmpx.h> + #define utmp utmpx #else #include <utmp.h> #endif *************** *** 1244,1253 **** --- 1511,1522 ---- int t() { int x = DEAD_PROCESS; struct utmp *y = pututline((struct utmp *)0); getutent(); } EOF if eval $compile; then + rm -rf conftest* { test -n "$verbose" && \ echo " defining GETUTENT" + echo "#define" GETUTENT 1 >> confdefs.h DEFS="$DEFS -DGETUTENT=1" SEDDEFS="${SEDDEFS}\${SEDdA}GETUTENT\${SEDdB}GETUTENT\${SEDdC}1\${SEDdD} \${SEDuA}GETUTENT\${SEDuB}GETUTENT\${SEDuC}1\${SEDuD} *************** *** 1255,1268 **** " } fi rm -f conftest* echo checking for ut_host cat > conftest.c <<EOF #include <time.h> #include <sys/types.h> ! #ifdef SVR4 #include <utmpx.h> #define utmp utmpx #else --- 1524,1539 ---- " } + fi rm -f conftest* echo checking for ut_host cat > conftest.c <<EOF + #include "confdefs.h" #include <time.h> #include <sys/types.h> ! #if defined(SVR4) && !defined(DGUX) #include <utmpx.h> #define utmp utmpx #else *************** *** 1273,1282 **** --- 1544,1555 ---- int t() { struct utmp u; u.ut_host[0] = 0; } EOF if eval $compile; then + rm -rf conftest* { test -n "$verbose" && \ echo " defining UTHOST" + echo "#define" UTHOST 1 >> confdefs.h DEFS="$DEFS -DUTHOST=1" SEDDEFS="${SEDDEFS}\${SEDdA}UTHOST\${SEDdB}UTHOST\${SEDdC}1\${SEDdD} \${SEDuA}UTHOST\${SEDuB}UTHOST\${SEDuC}1\${SEDuD} *************** *** 1284,1289 **** --- 1557,1563 ---- " } + fi rm -f conftest* *************** *** 1295,1309 **** --- 1569,1586 ---- echo checking for getloadavg cat > conftest.c <<EOF + #include "confdefs.h" int main() { exit(0); } int t() { getloadavg((double *)0, 0); } EOF if eval $compile; then + rm -rf conftest* { test -n "$verbose" && \ echo " defining LOADAV_GETLOADAVG" + echo "#define" LOADAV_GETLOADAVG 1 >> confdefs.h DEFS="$DEFS -DLOADAV_GETLOADAVG=1" SEDDEFS="${SEDDEFS}\${SEDdA}LOADAV_GETLOADAVG\${SEDdB}LOADAV_GETLOADAVG\${SEDdC}1\${SEDdD} \${SEDuA}LOADAV_GETLOADAVG\${SEDuB}LOADAV_GETLOADAVG\${SEDuC}1\${SEDuD} *************** *** 1311,1329 **** " } load=1 fi rm -f conftest* if test -z "$load" ; then cat > conftest.c <<EOF #if defined(NeXT) || defined(apollo) || defined(linux) yes #endif EOF ! eval "$CPP \$DEFS conftest.c > conftest.out 2>&1" if egrep "yes" conftest.out >/dev/null 2>&1; then load=1 fi rm -f conftest* --- 1588,1610 ---- " } load=1 + fi rm -f conftest* if test -z "$load" ; then cat > conftest.c <<EOF + #include "confdefs.h" #if defined(NeXT) || defined(apollo) || defined(linux) yes #endif EOF ! eval "$CPP conftest.c > conftest.out 2>&1" if egrep "yes" conftest.out >/dev/null 2>&1; then + rm -rf conftest* load=1 + fi rm -f conftest* *************** *** 1342,1348 **** { test -n "$verbose" && \ ! echo " defining LOADAV_UNIX to be \"$core\"" DEFS="$DEFS -DLOADAV_UNIX=\"$core\"" SEDDEFS="${SEDDEFS}\${SEDdA}LOADAV_UNIX\${SEDdB}LOADAV_UNIX\${SEDdC}\"$core\"\${SEDdD} \${SEDuA}LOADAV_UNIX\${SEDuB}LOADAV_UNIX\${SEDuC}\"$core\"\${SEDuD} --- 1623,1630 ---- { test -n "$verbose" && \ ! echo " defining" LOADAV_UNIX to be \"$core\" ! echo "#define" LOADAV_UNIX \"$core\" >> confdefs.h DEFS="$DEFS -DLOADAV_UNIX=\"$core\"" SEDDEFS="${SEDDEFS}\${SEDdA}LOADAV_UNIX\${SEDdB}LOADAV_UNIX\${SEDdC}\"$core\"\${SEDdD} \${SEDuA}LOADAV_UNIX\${SEDuB}LOADAV_UNIX\${SEDuC}\"$core\"\${SEDuD} *************** *** 1352,1365 **** echo checking for nlist.h cat > conftest.c <<EOF #include <nlist.h> EOF ! err=`eval "($CPP \$DEFS conftest.c >/dev/null) 2>&1"` if test -z "$err"; then { test -n "$verbose" && \ echo " defining NLIST_STRUCT" DEFS="$DEFS -DNLIST_STRUCT=1" SEDDEFS="${SEDDEFS}\${SEDdA}NLIST_STRUCT\${SEDdB}NLIST_STRUCT\${SEDdC}1\${SEDdD} \${SEDuA}NLIST_STRUCT\${SEDuB}NLIST_STRUCT\${SEDuC}1\${SEDuD} --- 1634,1650 ---- echo checking for nlist.h cat > conftest.c <<EOF + #include "confdefs.h" #include <nlist.h> EOF ! err=`eval "($CPP conftest.c >/dev/null) 2>&1"` if test -z "$err"; then + rm -rf conftest* { test -n "$verbose" && \ echo " defining NLIST_STRUCT" + echo "#define" NLIST_STRUCT 1 >> confdefs.h DEFS="$DEFS -DNLIST_STRUCT=1" SEDDEFS="${SEDDEFS}\${SEDdA}NLIST_STRUCT\${SEDdB}NLIST_STRUCT\${SEDdC}1\${SEDdD} \${SEDuA}NLIST_STRUCT\${SEDuB}NLIST_STRUCT\${SEDuC}1\${SEDuD} *************** *** 1369,1383 **** --- 1654,1671 ---- echo checking for n_un in struct nlist cat > conftest.c <<EOF + #include "confdefs.h" #include <nlist.h> int main() { exit(0); } int t() { struct nlist n; n.n_un.n_name = 0; } EOF if eval $compile; then + rm -rf conftest* { test -n "$verbose" && \ echo " defining NLIST_NAME_UNION" + echo "#define" NLIST_NAME_UNION 1 >> confdefs.h DEFS="$DEFS -DNLIST_NAME_UNION=1" SEDDEFS="${SEDDEFS}\${SEDdA}NLIST_NAME_UNION\${SEDdB}NLIST_NAME_UNION\${SEDdC}1\${SEDdD} \${SEDuA}NLIST_NAME_UNION\${SEDuB}NLIST_NAME_UNION\${SEDuC}1\${SEDuD} *************** *** 1385,1399 **** --- 1673,1690 ---- " } + fi rm -f conftest* + fi rm -f conftest* echo checking for nlist declaration cat > conftest.c <<EOF + #include "confdefs.h" #ifdef NLIST_STRUCT # include <nlist.h> *************** *** 1402,1413 **** #endif EOF ! eval "$CPP \$DEFS conftest.c > conftest.out 2>&1" ! if egrep "nlist( | )( | )*.*\(" conftest.out >/dev/null 2>&1; then { test -n "$verbose" && \ echo " defining NLIST_DECLARED" DEFS="$DEFS -DNLIST_DECLARED=1" SEDDEFS="${SEDDEFS}\${SEDdA}NLIST_DECLARED\${SEDdB}NLIST_DECLARED\${SEDdC}1\${SEDdD} \${SEDuA}NLIST_DECLARED\${SEDuB}NLIST_DECLARED\${SEDuC}1\${SEDuD} --- 1693,1706 ---- #endif EOF ! eval "$CPP conftest.c > conftest.out 2>&1" ! if egrep "nlist(( | )( | )*.*\(|\()" conftest.out >/dev/null 2>&1; then ! rm -rf conftest* { test -n "$verbose" && \ echo " defining NLIST_DECLARED" + echo "#define" NLIST_DECLARED 1 >> confdefs.h DEFS="$DEFS -DNLIST_DECLARED=1" SEDDEFS="${SEDDEFS}\${SEDdA}NLIST_DECLARED\${SEDdB}NLIST_DECLARED\${SEDdC}1\${SEDdD} \${SEDuA}NLIST_DECLARED\${SEDuB}NLIST_DECLARED\${SEDuC}1\${SEDuD} *************** *** 1415,1420 **** --- 1708,1714 ---- " } + fi rm -f conftest* *************** *** 1422,1427 **** --- 1716,1722 ---- echo searching for avenrun symbol for av in avenrun _avenrun _Loadavg ; do cat > conftest.c <<EOF + #include "confdefs.h" #include <sys/types.h> #ifdef NLIST_STRUCT *************** *** 1449,1456 **** eval $compile if test -s conftest && (./conftest; exit) 2>/dev/null; then avensym=$av;break fi ! rm -f conftest* done if test -z "$avensym" ; then echo "- no avenrun symbol found" --- 1744,1752 ---- eval $compile if test -s conftest && (./conftest; exit) 2>/dev/null; then avensym=$av;break + fi ! rm -fr conftest* done if test -z "$avensym" ; then echo "- no avenrun symbol found" *************** *** 1459,1465 **** { test -n "$verbose" && \ ! echo " defining LOADAV_AVENRUN to be \"$avensym\"" DEFS="$DEFS -DLOADAV_AVENRUN=\"$avensym\"" SEDDEFS="${SEDDEFS}\${SEDdA}LOADAV_AVENRUN\${SEDdB}LOADAV_AVENRUN\${SEDdC}\"$avensym\"\${SEDdD} \${SEDuA}LOADAV_AVENRUN\${SEDuB}LOADAV_AVENRUN\${SEDuC}\"$avensym\"\${SEDuD} --- 1755,1762 ---- { test -n "$verbose" && \ ! echo " defining" LOADAV_AVENRUN to be \"$avensym\" ! echo "#define" LOADAV_AVENRUN \"$avensym\" >> confdefs.h DEFS="$DEFS -DLOADAV_AVENRUN=\"$avensym\"" SEDDEFS="${SEDDEFS}\${SEDdA}LOADAV_AVENRUN\${SEDdB}LOADAV_AVENRUN\${SEDdC}\"$avensym\"\${SEDdD} \${SEDuA}LOADAV_AVENRUN\${SEDuB}LOADAV_AVENRUN\${SEDuC}\"$avensym\"\${SEDuD} *************** *** 1519,1524 **** --- 1816,1822 ---- { test -n "$verbose" && \ echo " defining LOADAV" + echo "#define" LOADAV 1 >> confdefs.h DEFS="$DEFS -DLOADAV=1" SEDDEFS="${SEDDEFS}\${SEDdA}LOADAV\${SEDdB}LOADAV\${SEDdC}1\${SEDdD} \${SEDuA}LOADAV\${SEDuB}LOADAV\${SEDuC}1\${SEDuD} *************** *** 1529,1535 **** if test -n "$loadtype" ; then { test -n "$verbose" && \ ! echo " defining LOADAV_TYPE to be $loadtype" DEFS="$DEFS -DLOADAV_TYPE=$loadtype" SEDDEFS="${SEDDEFS}\${SEDdA}LOADAV_TYPE\${SEDdB}LOADAV_TYPE\${SEDdC}$loadtype\${SEDdD} \${SEDuA}LOADAV_TYPE\${SEDuB}LOADAV_TYPE\${SEDuC}$loadtype\${SEDuD} --- 1827,1834 ---- if test -n "$loadtype" ; then { test -n "$verbose" && \ ! echo " defining" LOADAV_TYPE to be $loadtype ! echo "#define" LOADAV_TYPE $loadtype >> confdefs.h DEFS="$DEFS -DLOADAV_TYPE=$loadtype" SEDDEFS="${SEDDEFS}\${SEDdA}LOADAV_TYPE\${SEDdB}LOADAV_TYPE\${SEDdC}$loadtype\${SEDdD} \${SEDuA}LOADAV_TYPE\${SEDuB}LOADAV_TYPE\${SEDuC}$loadtype\${SEDuD} *************** *** 1540,1546 **** if test -n "$loadnum" ; then { test -n "$verbose" && \ ! echo " defining LOADAV_NUM to be $loadnum" DEFS="$DEFS -DLOADAV_NUM=$loadnum" SEDDEFS="${SEDDEFS}\${SEDdA}LOADAV_NUM\${SEDdB}LOADAV_NUM\${SEDdC}$loadnum\${SEDdD} \${SEDuA}LOADAV_NUM\${SEDuB}LOADAV_NUM\${SEDuC}$loadnum\${SEDuD} --- 1839,1846 ---- if test -n "$loadnum" ; then { test -n "$verbose" && \ ! echo " defining" LOADAV_NUM to be $loadnum ! echo "#define" LOADAV_NUM $loadnum >> confdefs.h DEFS="$DEFS -DLOADAV_NUM=$loadnum" SEDDEFS="${SEDDEFS}\${SEDdA}LOADAV_NUM\${SEDdB}LOADAV_NUM\${SEDdC}$loadnum\${SEDdD} \${SEDuA}LOADAV_NUM\${SEDuB}LOADAV_NUM\${SEDuC}$loadnum\${SEDuD} *************** *** 1551,1557 **** if test -n "$loadscale" ; then { test -n "$verbose" && \ ! echo " defining LOADAV_SCALE to be $loadscale" DEFS="$DEFS -DLOADAV_SCALE=$loadscale" SEDDEFS="${SEDDEFS}\${SEDdA}LOADAV_SCALE\${SEDdB}LOADAV_SCALE\${SEDdC}$loadscale\${SEDdD} \${SEDuA}LOADAV_SCALE\${SEDuB}LOADAV_SCALE\${SEDuC}$loadscale\${SEDuD} --- 1851,1858 ---- if test -n "$loadscale" ; then { test -n "$verbose" && \ ! echo " defining" LOADAV_SCALE to be $loadscale ! echo "#define" LOADAV_SCALE $loadscale >> confdefs.h DEFS="$DEFS -DLOADAV_SCALE=$loadscale" SEDDEFS="${SEDDEFS}\${SEDdA}LOADAV_SCALE\${SEDdB}LOADAV_SCALE\${SEDdC}$loadscale\${SEDdD} \${SEDuA}LOADAV_SCALE\${SEDuB}LOADAV_SCALE\${SEDuC}$loadscale\${SEDuD} *************** *** 1561,1573 **** fi ! echo '#include <signal.h>' > conftest.c ! eval "$CPP \$DEFS conftest.c > conftest.out 2>&1" ! if egrep "(void|sighandler_t).*signal" conftest.out >/dev/null 2>&1; then { test -n "$verbose" && \ echo " defining SIGVOID" DEFS="$DEFS -DSIGVOID=1" SEDDEFS="${SEDDEFS}\${SEDdA}SIGVOID\${SEDdB}SIGVOID\${SEDdC}1\${SEDdD} \${SEDuA}SIGVOID\${SEDuB}SIGVOID\${SEDuC}1\${SEDuD} --- 1862,1886 ---- fi ! echo checking for return type of signal handlers ! cat > conftest.c <<EOF ! #include "confdefs.h" ! #include <sys/types.h> ! #include <signal.h> ! #ifdef signal ! #undef signal ! #endif ! extern void (*signal ()) (); ! int main() { exit(0); } ! int t() { int i; } ! EOF ! if eval $compile; then ! rm -rf conftest* { test -n "$verbose" && \ echo " defining SIGVOID" + echo "#define" SIGVOID 1 >> confdefs.h DEFS="$DEFS -DSIGVOID=1" SEDDEFS="${SEDDEFS}\${SEDdA}SIGVOID\${SEDdB}SIGVOID\${SEDdC}1\${SEDdD} \${SEDuA}SIGVOID\${SEDuB}SIGVOID\${SEDuC}1\${SEDuD} *************** *** 1575,1585 **** --- 1888,1900 ---- " } + fi rm -f conftest* echo checking for sigset cat > conftest.c <<EOF + #include "confdefs.h" #include <sys/types.h> #include <signal.h> *************** *** 1594,1603 **** --- 1909,1920 ---- } EOF if eval $compile; then + rm -rf conftest* { test -n "$verbose" && \ echo " defining USESIGSET" + echo "#define" USESIGSET 1 >> confdefs.h DEFS="$DEFS -DUSESIGSET=1" SEDDEFS="${SEDDEFS}\${SEDdA}USESIGSET\${SEDdB}USESIGSET\${SEDdC}1\${SEDdD} \${SEDuA}USESIGSET\${SEDuB}USESIGSET\${SEDuC}1\${SEDuD} *************** *** 1605,1621 **** " } fi rm -f conftest* echo checking signal implementation cat > conftest.c <<EOF #include <sys/types.h> #include <signal.h> ! #ifndef SIGCHLD ! #define SIGCHLD SIGCLD #endif #ifdef USESIGSET #define signal sigset --- 1922,1940 ---- " } + fi rm -f conftest* echo checking signal implementation cat > conftest.c <<EOF + #include "confdefs.h" #include <sys/types.h> #include <signal.h> ! #ifndef SIGCLD ! #define SIGCLD SIGCHLD #endif #ifdef USESIGSET #define signal sigset *************** *** 1633,1641 **** main() { ! (void)signal(SIGCHLD, hand); ! kill(getpid(), SIGCHLD); ! kill(getpid(), SIGCHLD); if (got < 2) exit(1); exit(0); --- 1952,1960 ---- main() { ! (void)signal(SIGCLD, hand); ! kill(getpid(), SIGCLD); ! kill(getpid(), SIGCLD); if (got < 2) exit(1); exit(0); *************** *** 1650,1655 **** --- 1969,1975 ---- { test -n "$verbose" && \ echo " defining SYSVSIGS" + echo "#define" SYSVSIGS 1 >> confdefs.h DEFS="$DEFS -DSYSVSIGS=1" SEDDEFS="${SEDDEFS}\${SEDdA}SYSVSIGS\${SEDdB}SYSVSIGS\${SEDdC}1\${SEDdD} \${SEDuA}SYSVSIGS\${SEDuB}SYSVSIGS\${SEDuC}1\${SEDuD} *************** *** 1658,1664 **** } fi ! rm -f conftest* echo checking for crypt and sec libraries --- 1978,1984 ---- } fi ! rm -fr conftest* echo checking for crypt and sec libraries *************** *** 1670,1675 **** --- 1990,1996 ---- LIBS="$LIBS -lsun" echo checking for IRIX sun library cat > conftest.c <<EOF + #include "confdefs.h" int main() { exit(0); } int t() { } *************** *** 1677,1682 **** --- 1998,2004 ---- if eval $compile; then : else + rm -rf conftest* LIBS="$oldlibs" fi rm -f conftest* *************** *** 1685,1690 **** --- 2007,2013 ---- echo checking for wait union cat > conftest.c <<EOF + #include "confdefs.h" #include <sys/types.h> #include <sys/wait.h> *************** *** 1698,1707 **** --- 2021,2032 ---- } EOF if eval $compile; then + rm -rf conftest* { test -n "$verbose" && \ echo " defining BSDWAIT" + echo "#define" BSDWAIT 1 >> confdefs.h DEFS="$DEFS -DBSDWAIT=1" SEDDEFS="${SEDDEFS}\${SEDdA}BSDWAIT\${SEDdB}BSDWAIT\${SEDdC}1\${SEDdD} \${SEDuA}BSDWAIT\${SEDuB}BSDWAIT\${SEDuC}1\${SEDuD} *************** *** 1709,1714 **** --- 2034,2040 ---- " } + fi rm -f conftest* *************** *** 1716,1729 **** if test -z "$butterfly"; then echo checking for termio or termios cat > conftest.c <<EOF ! #include <termios.h> EOF ! err=`eval "($CPP \$DEFS conftest.c >/dev/null) 2>&1"` if test -z "$err"; then { test -n "$verbose" && \ echo " defining TERMIO" DEFS="$DEFS -DTERMIO=1" SEDDEFS="${SEDDEFS}\${SEDdA}TERMIO\${SEDdB}TERMIO\${SEDdC}1\${SEDdD} \${SEDuA}TERMIO\${SEDuB}TERMIO\${SEDuC}1\${SEDuD} --- 2042,2058 ---- if test -z "$butterfly"; then echo checking for termio or termios cat > conftest.c <<EOF ! #include "confdefs.h" ! #include <termio.h> EOF ! err=`eval "($CPP conftest.c >/dev/null) 2>&1"` if test -z "$err"; then + rm -rf conftest* { test -n "$verbose" && \ echo " defining TERMIO" + echo "#define" TERMIO 1 >> confdefs.h DEFS="$DEFS -DTERMIO=1" SEDDEFS="${SEDDEFS}\${SEDdA}TERMIO\${SEDdB}TERMIO\${SEDdC}1\${SEDdD} \${SEDuA}TERMIO\${SEDuB}TERMIO\${SEDuC}1\${SEDuD} *************** *** 1731,1746 **** " } else ! cat > conftest.c <<EOF ! #include <termio.h> EOF ! err=`eval "($CPP \$DEFS conftest.c >/dev/null) 2>&1"` if test -z "$err"; then { test -n "$verbose" && \ echo " defining TERMIO" DEFS="$DEFS -DTERMIO=1" SEDDEFS="${SEDDEFS}\${SEDdA}TERMIO\${SEDdB}TERMIO\${SEDdC}1\${SEDdD} \${SEDuA}TERMIO\${SEDuB}TERMIO\${SEDuC}1\${SEDuD} --- 2060,2081 ---- " } + else ! rm -rf conftest* ! if test -n "$posix"; then ! cat > conftest.c <<EOF ! #include "confdefs.h" ! #include <termios.h> EOF ! err=`eval "($CPP conftest.c >/dev/null) 2>&1"` if test -z "$err"; then + rm -rf conftest* { test -n "$verbose" && \ echo " defining TERMIO" + echo "#define" TERMIO 1 >> confdefs.h DEFS="$DEFS -DTERMIO=1" SEDDEFS="${SEDDEFS}\${SEDdA}TERMIO\${SEDdB}TERMIO\${SEDdC}1\${SEDdD} \${SEDuA}TERMIO\${SEDuB}TERMIO\${SEDuC}1\${SEDuD} *************** *** 1748,1770 **** --- 2083,2111 ---- " } + fi rm -f conftest* fi + + fi rm -f conftest* fi echo checking for getspnam cat > conftest.c <<EOF + #include "confdefs.h" #include <shadow.h> int main() { exit(0); } int t() { getspnam("x"); } EOF if eval $compile; then + rm -rf conftest* { test -n "$verbose" && \ echo " defining SHADOWPW" + echo "#define" SHADOWPW 1 >> confdefs.h DEFS="$DEFS -DSHADOWPW=1" SEDDEFS="${SEDDEFS}\${SEDdA}SHADOWPW\${SEDdB}SHADOWPW\${SEDdC}1\${SEDdD} \${SEDuA}SHADOWPW\${SEDuB}SHADOWPW\${SEDuC}1\${SEDuD} *************** *** 1772,1792 **** --- 2113,2137 ---- " } + fi rm -f conftest* echo checking for getttyent cat > conftest.c <<EOF + #include "confdefs.h" int main() { exit(0); } int t() { getttyent(); } EOF if eval $compile; then + rm -rf conftest* { test -n "$verbose" && \ echo " defining GETTTYENT" + echo "#define" GETTTYENT 1 >> confdefs.h DEFS="$DEFS -DGETTTYENT=1" SEDDEFS="${SEDDEFS}\${SEDdA}GETTTYENT\${SEDdB}GETTTYENT\${SEDdC}1\${SEDdD} \${SEDuA}GETTTYENT\${SEDuB}GETTTYENT\${SEDuC}1\${SEDuD} *************** *** 1794,1805 **** --- 2139,2152 ---- " } + fi rm -f conftest* echo checking whether memcpy/memmove/bcopy handles overlapping arguments cat > conftest.c <<EOF + #include "confdefs.h" main() { char buf[10]; *************** *** 1820,1825 **** --- 2167,2173 ---- { test -n "$verbose" && \ echo " defining USEBCOPY" + echo "#define" USEBCOPY 1 >> confdefs.h DEFS="$DEFS -DUSEBCOPY=1" SEDDEFS="${SEDDEFS}\${SEDdA}USEBCOPY\${SEDdB}USEBCOPY\${SEDdC}1\${SEDdD} \${SEDuA}USEBCOPY\${SEDuB}USEBCOPY\${SEDuC}1\${SEDuD} *************** *** 1827,1836 **** " } fi ! rm -f conftest* cat > conftest.c <<EOF #define bcopy(s,d,l) memmove(d,s,l) main() { --- 2175,2186 ---- " } + fi ! rm -fr conftest* cat > conftest.c <<EOF + #include "confdefs.h" #define bcopy(s,d,l) memmove(d,s,l) main() { *************** *** 1852,1857 **** --- 2202,2208 ---- { test -n "$verbose" && \ echo " defining USEMEMMOVE" + echo "#define" USEMEMMOVE 1 >> confdefs.h DEFS="$DEFS -DUSEMEMMOVE=1" SEDDEFS="${SEDDEFS}\${SEDdA}USEMEMMOVE\${SEDdB}USEMEMMOVE\${SEDdC}1\${SEDdD} \${SEDuA}USEMEMMOVE\${SEDuB}USEMEMMOVE\${SEDuC}1\${SEDuD} *************** *** 1859,1869 **** " } fi ! rm -f conftest* cat > conftest.c <<EOF #define bcopy(s,d,l) memcpy(d,s,l) main() { --- 2210,2222 ---- " } + fi ! rm -fr conftest* cat > conftest.c <<EOF + #include "confdefs.h" #define bcopy(s,d,l) memcpy(d,s,l) main() { *************** *** 1885,1890 **** --- 2238,2244 ---- { test -n "$verbose" && \ echo " defining USEMEMCPY" + echo "#define" USEMEMCPY 1 >> confdefs.h DEFS="$DEFS -DUSEMEMCPY=1" SEDDEFS="${SEDDEFS}\${SEDdA}USEMEMCPY\${SEDdB}USEMEMCPY\${SEDdC}1\${SEDdD} \${SEDuA}USEMEMCPY\${SEDuB}USEMEMCPY\${SEDuC}1\${SEDuD} *************** *** 1892,1899 **** " } fi ! rm -f conftest* echo checking for long file names (echo 1 > /tmp/conftest9012345) 2>/dev/null --- 2246,2254 ---- " } + fi ! rm -fr conftest* echo checking for long file names (echo 1 > /tmp/conftest9012345) 2>/dev/null *************** *** 1903,1909 **** else { test -n "$verbose" && \ ! echo " defining NAME_MAX to be 14" DEFS="$DEFS -DNAME_MAX=14" SEDDEFS="${SEDDEFS}\${SEDdA}NAME_MAX\${SEDdB}NAME_MAX\${SEDdC}14\${SEDdD} \${SEDuA}NAME_MAX\${SEDuB}NAME_MAX\${SEDuC}14\${SEDuD} --- 2258,2265 ---- else { test -n "$verbose" && \ ! echo " defining" NAME_MAX to be 14 ! echo "#define" NAME_MAX 14 >> confdefs.h DEFS="$DEFS -DNAME_MAX=14" SEDDEFS="${SEDDEFS}\${SEDdA}NAME_MAX\${SEDdB}NAME_MAX\${SEDdC}14\${SEDdD} \${SEDuA}NAME_MAX\${SEDuB}NAME_MAX\${SEDuC}14\${SEDuD} *************** *** 1916,1931 **** --- 2272,2290 ---- echo checking for vsprintf cat > conftest.c <<EOF + #include "confdefs.h" #include <varargs.h> #include <stdio.h> int main() { exit(0); } int t() { vsprintf(); } EOF if eval $compile; then + rm -rf conftest* { test -n "$verbose" && \ echo " defining USEVARARGS" + echo "#define" USEVARARGS 1 >> confdefs.h DEFS="$DEFS -DUSEVARARGS=1" SEDDEFS="${SEDDEFS}\${SEDdA}USEVARARGS\${SEDdB}USEVARARGS\${SEDdC}1\${SEDdD} \${SEDuA}USEVARARGS\${SEDuB}USEVARARGS\${SEDuC}1\${SEDuD} *************** *** 1933,1938 **** --- 2292,2298 ---- " } + fi rm -f conftest* *************** *** 1942,1957 **** if test -z "$dirheader"; then echo checking for dirent.h cat > conftest.c <<EOF #include <sys/types.h> #include <dirent.h> int main() { exit(0); } ! int t() { DIR *dirp = opendir ("/"); } EOF if eval $compile; then { test -n "$verbose" && \ echo " defining DIRENT" DEFS="$DEFS -DDIRENT=1" SEDDEFS="${SEDDEFS}\${SEDdA}DIRENT\${SEDdB}DIRENT\${SEDdC}1\${SEDdD} \${SEDuA}DIRENT\${SEDuB}DIRENT\${SEDuC}1\${SEDuD} --- 2302,2320 ---- if test -z "$dirheader"; then echo checking for dirent.h cat > conftest.c <<EOF + #include "confdefs.h" #include <sys/types.h> #include <dirent.h> int main() { exit(0); } ! int t() { DIR *dirp = 0; } EOF if eval $compile; then + rm -rf conftest* { test -n "$verbose" && \ echo " defining DIRENT" + echo "#define" DIRENT 1 >> confdefs.h DEFS="$DEFS -DDIRENT=1" SEDDEFS="${SEDDEFS}\${SEDdA}DIRENT\${SEDdB}DIRENT\${SEDdC}1\${SEDdD} \${SEDuA}DIRENT\${SEDuB}DIRENT\${SEDuC}1\${SEDuD} *************** *** 1959,1980 **** " } dirheader=dirent.h fi rm -f conftest* fi if test -z "$dirheader"; then echo checking for sys/ndir.h cat > conftest.c <<EOF #include <sys/types.h> #include <sys/ndir.h> int main() { exit(0); } ! int t() { DIR *dirp = opendir ("/"); } EOF if eval $compile; then { test -n "$verbose" && \ echo " defining SYSNDIR" DEFS="$DEFS -DSYSNDIR=1" SEDDEFS="${SEDDEFS}\${SEDdA}SYSNDIR\${SEDdB}SYSNDIR\${SEDdC}1\${SEDdD} \${SEDuA}SYSNDIR\${SEDuB}SYSNDIR\${SEDuC}1\${SEDuD} --- 2322,2347 ---- " } dirheader=dirent.h + fi rm -f conftest* fi if test -z "$dirheader"; then echo checking for sys/ndir.h cat > conftest.c <<EOF + #include "confdefs.h" #include <sys/types.h> #include <sys/ndir.h> int main() { exit(0); } ! int t() { DIR *dirp = 0; } EOF if eval $compile; then + rm -rf conftest* { test -n "$verbose" && \ echo " defining SYSNDIR" + echo "#define" SYSNDIR 1 >> confdefs.h DEFS="$DEFS -DSYSNDIR=1" SEDDEFS="${SEDDEFS}\${SEDdA}SYSNDIR\${SEDdB}SYSNDIR\${SEDdC}1\${SEDdD} \${SEDuA}SYSNDIR\${SEDuB}SYSNDIR\${SEDuC}1\${SEDuD} *************** *** 1982,2003 **** " } dirheader=sys/ndir.h fi rm -f conftest* fi if test -z "$dirheader"; then echo checking for sys/dir.h cat > conftest.c <<EOF #include <sys/types.h> #include <sys/dir.h> int main() { exit(0); } ! int t() { DIR *dirp = opendir ("/"); } EOF if eval $compile; then { test -n "$verbose" && \ echo " defining SYSDIR" DEFS="$DEFS -DSYSDIR=1" SEDDEFS="${SEDDEFS}\${SEDdA}SYSDIR\${SEDdB}SYSDIR\${SEDdC}1\${SEDdD} \${SEDuA}SYSDIR\${SEDuB}SYSDIR\${SEDuC}1\${SEDuD} --- 2349,2374 ---- " } dirheader=sys/ndir.h + fi rm -f conftest* fi if test -z "$dirheader"; then echo checking for sys/dir.h cat > conftest.c <<EOF + #include "confdefs.h" #include <sys/types.h> #include <sys/dir.h> int main() { exit(0); } ! int t() { DIR *dirp = 0; } EOF if eval $compile; then + rm -rf conftest* { test -n "$verbose" && \ echo " defining SYSDIR" + echo "#define" SYSDIR 1 >> confdefs.h DEFS="$DEFS -DSYSDIR=1" SEDDEFS="${SEDDEFS}\${SEDdA}SYSDIR\${SEDdB}SYSDIR\${SEDdC}1\${SEDdD} \${SEDuA}SYSDIR\${SEDuB}SYSDIR\${SEDuC}1\${SEDuD} *************** *** 2005,2026 **** " } dirheader=sys/dir.h fi rm -f conftest* fi if test -z "$dirheader"; then echo checking for ndir.h cat > conftest.c <<EOF #include <sys/types.h> #include <ndir.h> int main() { exit(0); } ! int t() { DIR *dirp = opendir ("/"); } EOF if eval $compile; then { test -n "$verbose" && \ echo " defining NDIR" DEFS="$DEFS -DNDIR=1" SEDDEFS="${SEDDEFS}\${SEDdA}NDIR\${SEDdB}NDIR\${SEDdC}1\${SEDdD} \${SEDuA}NDIR\${SEDuB}NDIR\${SEDuC}1\${SEDuD} --- 2376,2401 ---- " } dirheader=sys/dir.h + fi rm -f conftest* fi if test -z "$dirheader"; then echo checking for ndir.h cat > conftest.c <<EOF + #include "confdefs.h" #include <sys/types.h> #include <ndir.h> int main() { exit(0); } ! int t() { DIR *dirp = 0; } EOF if eval $compile; then + rm -rf conftest* { test -n "$verbose" && \ echo " defining NDIR" + echo "#define" NDIR 1 >> confdefs.h DEFS="$DEFS -DNDIR=1" SEDDEFS="${SEDDEFS}\${SEDdA}NDIR\${SEDdB}NDIR\${SEDdC}1\${SEDdD} \${SEDuA}NDIR\${SEDuB}NDIR\${SEDuC}1\${SEDuD} *************** *** 2028,2039 **** --- 2403,2416 ---- " } dirheader=ndir.h + fi rm -f conftest* fi echo checking for closedir return value cat > conftest.c <<EOF + #include "confdefs.h" #include <sys/types.h> #include <$dirheader> int closedir(); main() { exit(closedir(opendir(".")) != 0); } *************** *** 2046,2051 **** --- 2423,2429 ---- { test -n "$verbose" && \ echo " defining VOID_CLOSEDIR" + echo "#define" VOID_CLOSEDIR 1 >> confdefs.h DEFS="$DEFS -DVOID_CLOSEDIR=1" SEDDEFS="${SEDDEFS}\${SEDdA}VOID_CLOSEDIR\${SEDdB}VOID_CLOSEDIR\${SEDdC}1\${SEDdD} \${SEDuA}VOID_CLOSEDIR\${SEDuB}VOID_CLOSEDIR\${SEDuC}1\${SEDuD} *************** *** 2054,2071 **** } fi ! rm -f conftest* echo checking for Xenix cat > conftest.c <<EOF #if defined(M_XENIX) && !defined(M_UNIX) yes #endif EOF ! eval "$CPP \$DEFS conftest.c > conftest.out 2>&1" if egrep "yes" conftest.out >/dev/null 2>&1; then XENIX=1 fi rm -f conftest* --- 2432,2452 ---- } fi ! rm -fr conftest* echo checking for Xenix cat > conftest.c <<EOF + #include "confdefs.h" #if defined(M_XENIX) && !defined(M_UNIX) yes #endif EOF ! eval "$CPP conftest.c > conftest.out 2>&1" if egrep "yes" conftest.out >/dev/null 2>&1; then + rm -rf conftest* XENIX=1 + fi rm -f conftest* *************** *** 2080,2094 **** --- 2461,2478 ---- echo checking for setenv cat > conftest.c <<EOF + #include "confdefs.h" int main() { exit(0); } int t() { setenv((char *)0,(char *)0);unsetenv((char *)0); } EOF if eval $compile; then + rm -rf conftest* { test -n "$verbose" && \ echo " defining USESETENV" + echo "#define" USESETENV 1 >> confdefs.h DEFS="$DEFS -DUSESETENV=1" SEDDEFS="${SEDDEFS}\${SEDdA}USESETENV\${SEDdB}USESETENV\${SEDdC}1\${SEDdD} \${SEDuA}USESETENV\${SEDuB}USESETENV\${SEDuC}1\${SEDuD} *************** *** 2096,2104 **** --- 2480,2491 ---- " } + else + rm -rf conftest* echo checking for putenv cat > conftest.c <<EOF + #include "confdefs.h" int main() { exit(0); } int t() { putenv((char *)0);unsetenv((char *)0); } *************** *** 2106,2115 **** --- 2493,2504 ---- if eval $compile; then : else + rm -rf conftest* { test -n "$verbose" && \ echo " defining NEEDPUTENV" + echo "#define" NEEDPUTENV 1 >> confdefs.h DEFS="$DEFS -DNEEDPUTENV=1" SEDDEFS="${SEDDEFS}\${SEDdA}NEEDPUTENV\${SEDdB}NEEDPUTENV\${SEDdC}1\${SEDdD} \${SEDuA}NEEDPUTENV\${SEDuB}NEEDPUTENV\${SEDuC}1\${SEDuD} *************** *** 2129,2134 **** --- 2518,2524 ---- { test -n "$verbose" && \ echo " defining HAVE_DEV_PTC" + echo "#define" HAVE_DEV_PTC 1 >> confdefs.h DEFS="$DEFS -DHAVE_DEV_PTC=1" SEDDEFS="${SEDDEFS}\${SEDdA}HAVE_DEV_PTC\${SEDdB}HAVE_DEV_PTC\${SEDdC}1\${SEDdD} \${SEDuA}HAVE_DEV_PTC\${SEDuB}HAVE_DEV_PTC\${SEDuC}1\${SEDuD} *************** *** 2138,2146 **** --- 2528,2563 ---- fi + echo checking for strerror + cat > conftest.c <<EOF + #include "confdefs.h" + + int main() { exit(0); } + int t() { strerror(0); } + EOF + if eval $compile; then + rm -rf conftest* + + { + test -n "$verbose" && \ + echo " defining HAVE_STRERROR" + echo "#define" HAVE_STRERROR 1 >> confdefs.h + DEFS="$DEFS -DHAVE_STRERROR=1" + SEDDEFS="${SEDDEFS}\${SEDdA}HAVE_STRERROR\${SEDdB}HAVE_STRERROR\${SEDdC}1\${SEDdD} + \${SEDuA}HAVE_STRERROR\${SEDuB}HAVE_STRERROR\${SEDuC}1\${SEDuD} + \${SEDeA}HAVE_STRERROR\${SEDeB}HAVE_STRERROR\${SEDeC}1\${SEDeD} + " + } + + + fi + rm -f conftest* + + test -n "$seqptx" && LIBS="-ltermcap -lc -lsocket -linet -lsec -lseq" cat > conftest.c <<EOF + #include "confdefs.h" main(){exit(0);} EOF eval $compile *************** *** 2149,2160 **** else echo "Can't run the compiler - internal error. Sorry.";exit fi ! rm -f conftest* if test -n "$prefix"; then { test -n "$verbose" && \ ! echo " defining ETCSCREENRC to be \"$prefix/etc/screenrc\"" DEFS="$DEFS -DETCSCREENRC=\"$prefix/etc/screenrc\"" SEDDEFS="${SEDDEFS}\${SEDdA}ETCSCREENRC\${SEDdB}ETCSCREENRC\${SEDdC}\"$prefix/etc/screenrc\"\${SEDdD} \${SEDuA}ETCSCREENRC\${SEDuB}ETCSCREENRC\${SEDuC}\"$prefix/etc/screenrc\"\${SEDuD} --- 2566,2578 ---- else echo "Can't run the compiler - internal error. Sorry.";exit fi ! rm -fr conftest* if test -n "$prefix"; then { test -n "$verbose" && \ ! echo " defining" ETCSCREENRC to be \"$prefix/etc/screenrc\" ! echo "#define" ETCSCREENRC \"$prefix/etc/screenrc\" >> confdefs.h DEFS="$DEFS -DETCSCREENRC=\"$prefix/etc/screenrc\"" SEDDEFS="${SEDDEFS}\${SEDdA}ETCSCREENRC\${SEDdB}ETCSCREENRC\${SEDdC}\"$prefix/etc/screenrc\"\${SEDdD} \${SEDuA}ETCSCREENRC\${SEDuB}ETCSCREENRC\${SEDuC}\"$prefix/etc/screenrc\"\${SEDuD} *************** *** 2164,2169 **** --- 2582,2588 ---- fi + # Set default prefixes. if test -n "$prefix"; then test -z "$exec_prefix" && exec_prefix='${prefix}' prsub="s%^prefix\\([ ]*\\)=\\([ ]*\\).*$%prefix\\1=\\2$prefix%" *************** *** 2172,2183 **** --- 2591,2604 ---- prsub="$prsub s%^exec_prefix\\([ ]*\\)=\\([ ]*\\).*$%exec_prefix\\1=\\2$exec_prefix%" fi + # Quote sed substitution magic chars in DEFS. cat >conftest.def <<EOF $DEFS EOF escape_ampersand_and_backslash='s%[&\\]%\\&%g' DEFS=`sed "$escape_ampersand_and_backslash" <conftest.def` rm -f conftest.def + # Substitute for predefined variables. trap 'rm -f config.status; exit 1' 1 3 15 echo creating config.status *************** *** 2195,2201 **** do case "\$arg" in -recheck | --recheck | --rechec | --reche | --rech | --rec | --re | --r) ! exec /bin/sh $0 $configure_args ;; *) echo "Usage: config.status --recheck" 2>&1; exit 1 ;; esac done --- 2616,2623 ---- do case "\$arg" in -recheck | --recheck | --rechec | --reche | --rech | --rec | --re | --r) ! echo running \${CONFIG_SHELL-/bin/sh} $0 $configure_args ! exec \${CONFIG_SHELL-/bin/sh} $0 $configure_args ;; *) echo "Usage: config.status --recheck" 2>&1; exit 1 ;; esac done *************** *** 2213,2227 **** prefix='$prefix' exec_prefix='$exec_prefix' prsub='$prsub' EOF cat >> config.status <<\EOF top_srcdir=$srcdir ! # Allow make-time overrides of the generated file list. ! test -n "$gen_files" || gen_files="Makefile doc/Makefile" ! ! for file in .. $gen_files; do if [ "x$file" != "x.." ]; then srcdir=$top_srcdir # Remove last slash and all that follows it. Not all systems have dirname. dir=`echo $file|sed 's%/[^/][^/]*$%%'` --- 2635,2648 ---- prefix='$prefix' exec_prefix='$exec_prefix' prsub='$prsub' + extrasub='$extrasub' EOF cat >> config.status <<\EOF top_srcdir=$srcdir ! CONFIG_FILES=${CONFIG_FILES-"Makefile doc/Makefile"} ! for file in .. ${CONFIG_FILES}; do if test "x$file" != x..; then srcdir=$top_srcdir # Remove last slash and all that follows it. Not all systems have dirname. dir=`echo $file|sed 's%/[^/][^/]*$%%'` *************** *** 2234,2239 **** --- 2655,2661 ---- echo "# Generated automatically from `echo $file|sed 's|.*/||'`.in by configure." > $file sed -e " $prsub + $extrasub s%@VERSION@%$VERSION%g s%@CC@%$CC%g s%@CPP@%$CPP%g *************** *** 2245,2252 **** s%@srcdir@%$srcdir%g s%@DEFS@%-DHAVE_CONFIG_H%" $top_srcdir/${file}.in >> $file fi; done ! test -n "$gen_config" || gen_config=config.h ! echo creating $gen_config # These sed commands are put into SEDDEFS when defining a macro. # They are broken into pieces to make the sed script easier to manage. # They are passed to sed as "A NAME B NAME C VALUE D", where NAME --- 2667,2677 ---- s%@srcdir@%$srcdir%g s%@DEFS@%-DHAVE_CONFIG_H%" $top_srcdir/${file}.in >> $file fi; done ! ! CONFIG_HEADERS=${CONFIG_HEADERS-"config.h"} ! for file in .. ${CONFIG_HEADERS}; do if test "x$file" != x..; then ! echo creating $file ! # These sed commands are put into SEDDEFS when defining a macro. # They are broken into pieces to make the sed script easier to manage. # They are passed to sed as "A NAME B NAME C VALUE D", where NAME *************** *** 2276,2289 **** $SEDDEFS EOF - # Maximum number of lines to put in a single here document. - maxshlines=9 - # Break up $SEDDEFS (now in conftest.sh) because some shells have a limit # on the size of here documents. while : do lines=`grep -c . conftest.sh` if test -z "$lines" || test "$lines" -eq 0; then break; fi rm -f conftest.s1 conftest.s2 --- 2701,2715 ---- $SEDDEFS EOF # Break up $SEDDEFS (now in conftest.sh) because some shells have a limit # on the size of here documents. + # Maximum number of lines to put in a single here document. + maxshlines=9 + while : do + # wc gives bogus results for an empty file on some systems. lines=`grep -c . conftest.sh` if test -z "$lines" || test "$lines" -eq 0; then break; fi rm -f conftest.s1 conftest.s2 *************** *** 2310,2316 **** rm -f conftest.h # Break up the sed commands because old seds have small limits. maxsedlines=20 ! cp $top_srcdir/$gen_config.in conftest.h1 while : do lines=`grep -c . conftest.sed` --- 2736,2742 ---- rm -f conftest.h # Break up the sed commands because old seds have small limits. maxsedlines=20 ! cp $top_srcdir/$file.in conftest.h1 while : do lines=`grep -c . conftest.sed` *************** *** 2324,2345 **** mv conftest.s2 conftest.sed done rm -f conftest.sed conftest.h ! echo "/* $gen_config. Generated automatically by configure. */" > conftest.h cat conftest.h1 >> conftest.h rm -f conftest.h1 ! if cmp -s $gen_config conftest.h 2>/dev/null; then # The file exists and we would not be changing it. rm -f conftest.h else ! rm -f $gen_config ! mv conftest.h $gen_config fi exit 0 EOF chmod +x config.status ! test -n "$no_create" || ./config.status # a hook for preserving undef directive in config.h --- 2750,2774 ---- mv conftest.s2 conftest.sed done rm -f conftest.sed conftest.h ! echo "/* $file. Generated automatically by configure. */" > conftest.h cat conftest.h1 >> conftest.h rm -f conftest.h1 ! if cmp -s $file conftest.h 2>/dev/null; then # The file exists and we would not be changing it. + echo "$file is unchanged" rm -f conftest.h else ! rm -f $file ! mv conftest.h $file fi + fi; done + exit 0 EOF chmod +x config.status ! ${CONFIG_SHELL-/bin/sh} config.status # a hook for preserving undef directive in config.h *** socket.c.orig Wed Jan 5 18:38:30 1994 --- socket.c Wed Jan 5 18:39:14 1994 *************** *** 1057,1063 **** Kill(m.m.attach.apid, SIG_BYE); break; } ! #if defined(ultrix) || defined(pyr) brktty(d_userfd); /* for some strange reason this must be done */ #endif #if defined(pyr) || defined(xelos) || defined(sequent) --- 1057,1063 ---- Kill(m.m.attach.apid, SIG_BYE); break; } ! #if defined(ultrix) || defined(pyr) || defined(NeXT) brktty(d_userfd); /* for some strange reason this must be done */ #endif #if defined(pyr) || defined(xelos) || defined(sequent) *** comm.sh.orig Sat Jan 15 17:32:53 1994 --- comm.sh Sat Jan 15 17:36:37 1994 *************** *** 65,72 **** $CC -E -I. -I${srcdir} ${srcdir}/comm.c > comm.cpp sed < comm.cpp \ -n \ ! -e '/^ { "/y/abcdefghijklmnopqrstuvwxyz/ABCDEFGHIJKLMNOPQRSTUVWXYZ/' \ ! -e '/^ { "/s/^ { "\([^"]*\)".*/\1/p' \ | $AWK ' /.*/ { printf "#define RC_%s %d\n",$0,i++; } --- 65,72 ---- $CC -E -I. -I${srcdir} ${srcdir}/comm.c > comm.cpp sed < comm.cpp \ -n \ ! -e '/^ *{ "/y/abcdefghijklmnopqrstuvwxyz/ABCDEFGHIJKLMNOPQRSTUVWXYZ/' \ ! -e '/^ *{ "/s/^ *{ "\([^"]*\)".*/\1/p' \ | $AWK ' /.*/ { printf "#define RC_%s %d\n",$0,i++; } *** os.h.orig Mon Aug 16 16:49:09 1993 --- os.h Sat Jan 15 17:52:15 1994 *************** *** 226,232 **** #ifndef SIG_T_DEFINED # ifdef SIGVOID ! # if defined(ultrix) # define sig_t void # else /* nice compilers: */ typedef void sig_t; --- 226,232 ---- #ifndef SIG_T_DEFINED # ifdef SIGVOID ! # if defined(ultrix) || defined(NeXT) # define sig_t void # else /* nice compilers: */ typedef void sig_t; *************** *** 240,246 **** # endif #endif /* SIG_T_DEFINED */ ! #if defined(SVR4) || (defined(SYSV) && defined(ISC)) || defined(_AIX) || defined(linux) || defined(ultrix) || defined(__386BSD__) || defined(BSDI) # define SIGPROTOARG (int) # define SIGDEFARG int sigsig # define SIGARG 0 --- 240,246 ---- # endif #endif /* SIG_T_DEFINED */ ! #if defined(SVR4) || (defined(SYSV) && defined(ISC)) || defined(_AIX) || defined(linux) || defined(ultrix) || defined(__386BSD__) || defined(BSDI) || defined(NeXT) # define SIGPROTOARG (int) # define SIGDEFARG int sigsig # define SIGARG 0 *************** *** 258,272 **** # define signal sigset #endif /* USESIGSET */ - #ifndef PID_T_DEFINED - typedef int pid_t; - #endif - - #ifndef UID_T_DEFINED - typedef int gid_t; - typedef int uid_t; - #endif - #ifdef GETUTENT typedef char *slot_t; #else --- 258,263 ---- *** loadav.c.orig Mon Aug 16 16:49:07 1993 --- loadav.c Sat Jan 15 16:51:39 1994 *************** *** 114,120 **** #if defined(NeXT) && !defined(LOADAV_DONE) #define LOADAV_DONE ! #include <mach.h> static processor_set_t default_set; --- 114,120 ---- #if defined(NeXT) && !defined(LOADAV_DONE) #define LOADAV_DONE ! #include <mach/mach.h> static processor_set_t default_set; -- Mike Andrews -- root@fragile.termfrost.org -- kramer@vax.wittenberg.edu "To defy the laws of tradition is a crusade only of the brave..." - Primus
From: sears@uh.edu (Paul S. Sears) Newsgroups: comp.sys.next.bugs,comp.sys.next.software,comp.sys.next.misc Subject: Updated GatorFTP Date: 18 May 1994 12:44:53 GMT Organization: University of Houston Message-ID: <2rd2k5$gmd@masala.cc.uh.edu> The newest release of GatorFTP dated 5/17/94 (still 1.5, but there were some minor fixes) is located at: uhoop.egr.uh.edu:/pub/misc/GatorFTP/GatorFTP_MAB_1.5.app.tar.Z -- Paul S. Sears * sears@uh.edu (NeXT Mail OK) The University of Houston * suggestions@tree.egr.uh.edu (NeXT Engineering Computing Center * comments, complaints, questions) NeXT System Administration * DoD#1967 '83 NightHawk 650SC >>> SSI Diving Certification #755020059 <<< "Programming is like sex: One mistake and you support it a lifetime."
Newsgroups: comp.sys.next.bugs From: cedman@princeton.edu (Carl Edman) Subject: Re: NS3.[12] emacs In-Reply-To: mcquill@next.duq.edu's message of 17 May 1994 19:11:49 -0400 To: mcquill@next.duq.edu (Tod McQuillin) Message-ID: <CEDMAN.94May18133253@capitalist.princeton.edu> Originator: news@nimaster Sender: news@Princeton.EDU (USENET News System) Organization: Princeton University References: <2r90fj$61l@keen.ccit.duq.edu> <2r93j3$8km@rosie.next.com> <2rbivl$2li@keen.ccit.duq.edu> Date: Wed, 18 May 1994 17:32:53 GMT In article <2rbivl$2li@keen.ccit.duq.edu> mcquill@next.duq.edu (Tod McQuillin) writes: In article <2r93j3$8km@rosie.next.com>, Erik Kay <erikkay@next.com> wrote: > In article <2r90fj$61l@keen.ccit.duq.edu> mcquill@next.duq.edu (Tod > McQuillin) writes: > > [...] Since they changed the system type from berkeley-unix to > > next-mach, paths.el defines manual-program as /usr/bin/man instead of > > /usr/ucb/man. What lossage. > > This bug has already been fixed for our next release. As says also Lennart Lovstrand. Thanks for the prompt response, guys. Sorry if I sounded a bit harsh. Any chance of having Carl Edman's "Emacs for NeXTstep 4.0" available in a future NeXTStep release? That was the plan. Unfortunately the person at NeXT with whom this was arranged went to another company and neglected to inform anyone else of it. Right now NeXT seems to be working under fairly strong deadlines with regard to the freezing of NS 3.3 and nobody any longer seems to have time to include Emacs for NS. Carl Edman
From: bobs@pth3.bu.edu (Robert Singleton) Newsgroups: comp.sys.next.bugs Subject: A Perplexing Problem Date: 20 May 1994 19:31:57 GMT Organization: Boston University Message-ID: <2rj37d$45r@news.bu.edu> Hello, Recently I asked about a "network" problem I'm having. I've taken my NeXT (NS 3.1) off the network and am presently using it at home. It was configured as pth3.bu.edu (as is the machine I'm currently using). When I reboot I get the following message when the daemons are started: autonfsmount pth3 can't get my address. It then takes about 5 min. or so before it gives up looking, after which it boots successfully. Apparently there is more to removing a machine from the network than simply unplugging it and taking it home. Someone suggested that I go into the Simple Network Starter and define myself as a stand alone node. It turns out that I've been a stand alone node all along. But while I was in the Simple Network Starter I changed my host name to "localhost" and configured the machine. Now the machine rebooted successfully, with no autonfsmount complaint - and the above problem seems to have disappeared. But now when I login, the Workspace manager is gone (although a shell window works). In about 5 minutes, the Workspace manager reappears and everything seems to be just fine. The next time I login, if I haven't rebooted, the Workspace manager starts up as it should. Any suggestions on this very perplexing problem?!? Thanks very much. Regards, -- bob singleton bobs@pth3.bu.edu
Newsgroups: comp.sys.next.bugs Organization: Antigone Press gateway, San Francisco Return-Path: <xystem!jsmit@relay.NL.net> Date: Sat, 21 May 94 13:57:13 +0100 From: Jack Smit <xystem!jsmit@relay.NL.net> Message-ID: <9405211257.AA00343@xystem> Subject: NS3.2 installation on a IBM Modell 95 I tried to install NeXTSTEP for INTEL on a IBM Modell 95 with the following configuration: - 500 MB scsi harddisk - 16 MB memory - 2.88 MB Floppy drive - XGA Adapter with 14" display I inserted the NEXTSTEP installation bootdisk and restarted the system. The IBM is booting from the disk but after the message "NEXTSTEP ......" I got the error "READ ERROR"!!!!!! Every stops after this messages. Does some have experiences with NS3.2 in combination with IBM PC or with the error message. FYI the bootdisk isn't damaged because on other not-IBM it works perfect. Because of other IBM units in the office I tried them ...... but after IBM sees the NEXTSTEP kernel it gives the same error. (Used IBM machines: IBM Thinkpad, M70, M90) -------------------------------------------------------- Jack Smit +31 46 529019 Xystem Consultancy BV JSmit@Xystem.NL The Netherlands NeXTmail welcome --------------------------------------------------------
From: hoff@darmstadt.gmd.de (Holger Hoffstaette) Newsgroups: comp.sys.next.bugs Subject: Suggestion: 'active' scrollbars Summary: GUI improvements, visual feedback Message-ID: <8054@darmstadt.gmd.de> Date: 22 May 94 13:41:41 GMT Sender: hoff@darmstadt.gmd.de Organization: German National Research Centre for Information Technology Dear GUI people @ NeXT ! I would like to make a rather trivial, but IMHO important suggestion about the behaviour of scrollbars: when you grab a scroller and shove it around, you do not _see_ that you have grabbed it. Of course you see that the scroller is moving when you start moving the mouse, but a sort-of 3D visual feedback would be very nice to have.. Come on folks, even OpenLook has this.. :-> Regards, H.Hoffstaette -- PANIC: /earth is full. Mankind dumped.
From: cgeyer@aol.com (Cgeyer) Newsgroups: comp.sys.next.bugs Subject: Re: NS3.2 installation on a IBM Modell 95 Date: 22 May 1994 18:07:02 -0400 Organization: America Online, Inc. (1-800-827-6364) Sender: news@search01.news.aol.com Message-ID: <2rol26$j2t@search01.news.aol.com> Doesn't the IBM Model 95 run on MCA? If it does I don't think you can use it. If I remember NeXTSTEP does not support MCA. If the 95 doesn't use MCA then I don't know what the problem is. Christopher Geyer
From: chris@nice.usergroup.ethz.ch (Christian Limpach) Newsgroups: comp.sys.next.bugs Subject: Re: Suggestion: 'active' scrollbars Date: 22 May 1994 17:14:21 GMT Organization: welcome to nowhere... Message-ID: <CHRIS.94May22191421arkin@nice.usergroup.ethz.ch> References: <8054@darmstadt.gmd.de> In-reply-to: hoff@darmstadt.gmd.de's message of 22 May 94 13:41:41 GMT Originator: ARKIN@nice In article <8054@darmstadt.gmd.de> hoff@darmstadt.gmd.de (Holger Hoffstaette) writes: > I would like to make a rather trivial, but IMHO important suggestion > about the behaviour of scrollbars: when you grab a scroller and shove it > around, you do not _see_ that you have grabbed it. Of course you see that > the scroller is moving when you start moving the mouse, but a sort-of > 3D visual feedback would be very nice to have.. I wouldn't like to have this. It would IMHO be quite useless. After all, real-world sliders don't give you a visual feedback if you use them, GUI buttons just like real-world buttons do... And what about the titlebar, the resizers ? Do you want to change these as well ? One of the things I most appreciate in NeXT's GUI is that not every little thing has some useless 3D feedback... christian
From: mow@marsu.tynet.sub.org (Markus Wenzel) Newsgroups: comp.sys.next.bugs Subject: Re: NS3.2 installation on a IBM Modell 95 Date: 22 May 1994 11:40:01 +0100 Organization: Palumbian Research Labs Message-ID: <2rncq1$lp1@marsu.tynet.sub.org> References: <9405211257.AA00343@xystem> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Jack Smit <xystem!jsmit@relay.NL.net> writes: >I tried to install NeXTSTEP for INTEL on a IBM Modell 95 with the following >configuration: > - 500 MB scsi harddisk > - 16 MB memory > - 2.88 MB Floppy drive > - XGA Adapter with 14" display >I inserted the NEXTSTEP installation bootdisk and restarted the system. The IBM >is booting from the disk but after the message "NEXTSTEP ......" I got the >error "READ ERROR"!!!!!! Every stops after this messages. I don't know if this is the cause, but 2.88 MB Floppy drives are not supported on NS/I, except SCSI Floppy drives. Markus. -- /dev/ Markus Wenzel /usr/spool/mail/ mow@marsu.tynet.sub.org /etc/zoneinfo/ University of Stuttgart /bin/ps System administration, Consulting, Networking
From: hoff@kylldarmstadt.gmd.de (Holger Hoffstaette) Newsgroups: comp.sys.next.bugs Subject: Re: Suggestion: 'active' scrollbars Date: 23 May 1994 12:48:05 GMT Organization: German National Research Centre for Information Technology Sender: hoff@kyll (Holger Hoffstaette) Distribution: world Message-ID: <2rq8m5$hrd@omega.gmd.de> References: <8054@darmstadt.gmd.de> <CHRIS.94May22191421arkin@nice.usergroup.ethz.ch> In article <CHRIS.94May22191421arkin@nice.usergroup.ethz.ch>, chris@nice.usergroup.ethz.ch (Christian Limpach) writes: >> I would like to make a rather trivial, but IMHO important suggestion >> about the behaviour of scrollbars: when you grab a scroller and shove it >> around, you do not _see_ that you have grabbed it. Of course you see that >> the scroller is moving when you start moving the mouse, but a sort-of >> 3D visual feedback would be very nice to have.. > >I wouldn't like to have this. It would IMHO be quite useless. After >all, real-world sliders don't give you a visual feedback if you use >them, GUI buttons just like real-world buttons do... And what about >the titlebar, the resizers ? Do you want to change these as well ? But that's the whole point of visual feedback: if the slider is real, I can feel it. Since I'm dealing with a 'virtual' slider when I use a GUI, I need some kind of 'virtual' feeling, too - and this is currently only possible by changing color, style, etc. This is especially important for GUIs which work in high resolutions, like NEXTSTEP usually does. As to the titlebar, resizers, etc. : buttons already show that they are being pressed, and resizing a window brings up the outline, so there's no need to change anything there.. >One of the things I most appreciate in NeXT's GUI is that not every >little thing has some useless 3D feedback... I agree - not _every_ little thing. But at least those which make up a standard (style guide conformant) GUI. Well..another possibility would be to assign sounds to each of the different GUI elements. If you press a button, it says 'ouch!'. ;-) bye, Holger --- "I'm gonna get myself connected.." - Stereo MC's
From: sanguish@digifix.com Newsgroups: comp.sys.next.announce,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.bugs,comp.sys.next.programmer,comp.sys.next.marketplace Subject: Quick Guide to NEXTSTEP information on the Internet Date: 23 May 1994 02:40:32 -0400 Organization: Next Announcements Message-ID: <2rpj50$lju@digifix.digifix.com> This post is made weekly, to help 'point' users to more NEXTSTEP information Topics include: Stepwise NEXTSTEP/OpenStep Information Server online comp.sys.next newsgroups related newsgroups comp.sys.next newsgroups mailing list ftp sites NeXTanswers New Information --------------- Stepwise NEXTSTEP/OpenStep Information Server online A product directory built around the World Wide Web system, this will allow full multimedia announcements by NEXTSTEP developers, as well as the ability to browse the available products for NEXTSTEP. This service is online now, and can be reached at http://digifix.digifix.com/ it can be reached using OmniWeb (available from ftp.omnigroup.com) or Mosaic. The entries are coming in quite quickly, and I'll be moving the entire NEXTSTEP Third Party Catalog contents in as soon as they become available. Additionally the NEXTSTEP Product Information Mail Server will be stocked full file files in the next week... you can get information on using the mail server at ns-products@digifix.com Suggestions or comments can be directed to me at sanguish@digifix.com comp.sys.next.* newsgroups -------------------------- Comp.Sys.Next.Advocacy This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. Comp.Sys.Next.Announce Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) The NEXTSTEP FAQs are posted here monthly as well. This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Comp.Sys.Next.Bugs A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT- specific groups as well. Comp.Sys.Next.Hardware Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. Comp.Sys.Next.Marketplace NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. Comp.Sys.Next.Misc For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! Comp.Sys.Next.Programmer Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. The NEXTSTEP programmer FAQs are posted here. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. Comp.Sys.Next.Software This is a place to talk about [third party] software products that run on NEXTSTEP systems. Comp.Sys.Next.Sysadmin Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. related Newsgroups ------------------ Comp.Soft-Sys.Nextstep Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. Comp.Lang.Objective-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. Comp.Object Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original Comp.Sys.Next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News ------------------------------------------- Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep-d next-advocacy-d next-announce-d next-bugs-d next-hardware-d next-marketplace-d next-misc-d next-programmer-d next-software-d next-sysadmin-d (For a full description, send mail saying LISTS to <digestif@antigone.com>). The subscription syntax is essentially the same as LISTSERV's. To subscribe, send a message to <digestif@antigone.com> saying: SUB Listname YourName Example: SUB next-hardware-d John Doe The ftp sites ------------- cs.orst.edu: The main site for North American submissions nova.cc.purdue.edu: Lots of older stuff, but very short on disk space ftp.informatik.uni-muenchen.de: In Germany. terra.stack.urc.tue.nl (Dutch NEXTSTEP User Group) and cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp.next.com: See the below ftp.next.com and NextAnswers@next.com ------------------------------------- From the document 1000_Help from ftp.next.com Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, or you can transfer them by anonymous ftp. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to NeXTanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO <your-address> If you have any problem with the system or suggestions for improvement, please send mail to NeXTanswers-request@NeXT.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to NeXTanswers-request@NeXT.com. Written by: Eric P. Scott eps@toaster.SFSU.EDU and Scott Anguish sanguish@digifix.com Additions from: Greg Anderson (Greg_Anderson@afs.com) and Michael Pizolato (Michael_Pizolato@afs.com)
From: bobs@pth3.bu.edu (Robert Singleton) Newsgroups: comp.sys.next.bugs Subject: configuring printer Date: 23 May 1994 15:02:09 GMT Organization: Boston University Message-ID: <2rqghh$t56@news.bu.edu> Hello, I'm trying to configure a printer to my NeXT. The printer (called ppr) is on the network on a host called pthpr, and my local NeXT is called pth3. In using the PrintManagter.app, the host for the printer ppr is being set to pth3.bu.edu - but it should be set to pthpr. I can't seem to change the host to pthpr. At one point I had another NeXT connected to this same printer, but now I've forgotten how I did it. I should also say that I'm not using NetInfo (if that's relevant). Thanks for any assistance. Regards, -- bob singleton bobs@pth3.bu.edu
From: bobs@pth3.bu.edu (Robert Singleton) Newsgroups: comp.sys.next.bugs Subject: annoying icon problem Date: 23 May 1994 18:11:30 GMT Organization: Boston University Message-ID: <2rqrki$3vr@news.bu.edu> I seem to have an annoying icon problem. When I'm logged on as root, the home directory bobs (which is my usual home directory) appears as a regular folder - it doesn't have the double house icon. Does anyone know what the problem is? Thanks very much. regards, -- bob singleton bobs@pth3.bu.edu
From: bobs@pth3.bu.edu (Robert Singleton) Newsgroups: comp.sys.next.bugs Subject: Re: configuring printer Date: 23 May 1994 18:18:08 GMT Organization: Boston University Message-ID: <2rqs10$41j@news.bu.edu> References: <2rqghh$t56@news.bu.edu> Just so you don't waste your time replying to this question: I now have it figured out. The problem was that I had to load the /etc/printcap file into NetInfo by hand using niload printcap . < /etc/printcap For some reason the PrintManager.app didn't do this (I've never had to use niload in configuring a printer to my NeXT before - maybe there is a but in PrintManager?) -- bob singleton bobs@pth3.bu.edu
From: mchrist@acpub.duke.edu (Margaret Christopher) Newsgroups: comp.sys.next.bugs Subject: Re: NS3.2 installation on a IBM Modell 95 Date: 23 May 1994 18:47:18 GMT Organization: Duke University; Durham, N.C., USA Message-ID: <2rqtnm$ig7@news.duke.edu> References: <9405211257.AA00343@xystem> Jack Smit (xystem!jsmit@relay.NL.net) wrote: : I tried to install NeXTSTEP for INTEL on a IBM Modell 95 with the following : configuration: The IBM PS/2 Model 95 is MicroChannel and NSFIP won't install on it. PERIOD.
From: chris@iastate.edu (Chris Wong) Newsgroups: comp.sys.next.bugs,comp.sys.next.software,comp.sys.next.admin Subject: Serial driver broken in NS 3.2 Date: 24 May 1994 04:54:19 GMT Organization: Iowa State University, Ames, IA Distribution: world Message-ID: <2rs19r$o2i@news.iastate.edu> Keywords: serial,driver,ns,NS,NeXTSTEP The serial driver in NS 3.2 has broken again. Config: Hayes OPTIMA 144 external modem connected to COM1 DELL 450DE/2 JAWS running NeXTSTEP 3.2 Proof: Running kermit under NS 3.2 via serial ports above 9600bps would cause characters to drop (even in interactive screen display), but this doesn't happen under OS/2 or DOS. Anyone got a fix? Does it imply that I cannot send FAX/DATA at higher than 9600bps? Thanks for any info. Chris -- NeXTMail super welcomed!!| Chris Wong | "Hardware is supposed to serve Software." chris@iastate.edu | Computer Engineering & Computer Science <:)>:)<:)>:)<:)>:)<:) | Iowa State University of Science and Technology
From: alfred@ca-risc.co.at Newsgroups: comp.sys.next.bugs Subject: BUG: PS-108.1 HELP! Date: 24 May 1994 13:39:22 GMT Organization: EUnet EDV-Dienstleistungsgesellschaft m.b.H Message-ID: <2rt02a$j1i@hp4at.eunet.co.at> I get that frequently and it crashes the system: May 24 13:55:32 knuddel WindowServer[1420]: WindowServer: Exception caught: PS-108.1 on host `knuddel', user `evelyne': Memory access exception on address 0x202dd4e8: invalid address (SIGSEGV) Any ideas what it is nad how to fix it? No one here seems to know it. Thanks a lot in advance Alfred -- ================================ Alfred H. Corrodi "Capitalism is based on the assumption Risk Manager that you can win the game. Communism Creditanstalt-Bankverein is based on the assumption that you can alfred@ca-risc.co.at (NeXT) break even. Mysticism is based on the Tel/43/1/531311967 Fax/5337341 assumption that you can leave the game."
From: kinch@jones.heart.rri.uwo.ca (Dave Kinchlea) Newsgroups: comp.sys.next.bugs Subject: Re: Suggestion: 'active' scrollbars Date: 24 May 1994 17:51:03 GMT Organization: Heart Valve Lab, Robarts Research Institute Message-ID: <2rteq7$loi@falcon.ccs.uwo.ca> References: <8054@darmstadt.gmd.de> <CHRIS.94May22191421arkin@nice.usergroup.ethz.ch> In article <CHRIS.94May22191421arkin@nice.usergroup.ethz.ch>, Christian Limpach <chris@nice.usergroup.ethz.ch> wrote: >In article <8054@darmstadt.gmd.de> hoff@darmstadt.gmd.de (Holger Hoffstaette) writes: > >> 3D visual feedback would be very nice to have.. > >I wouldn't like to have this. It would IMHO be quite useless. After >all, real-world sliders don't give you a visual feedback if you use No, real-world sliders provide a tactile feedback so you don't NEED a 3d feedback. >them, GUI buttons just like real-world buttons do... And what about >the titlebar, the resizers ? Do you want to change these as well ? > And why not? In what way would these hurt? >One of the things I most appreciate in NeXT's GUI is that not every >little thing has some useless 3D feedback... > One of the things I least appreciate in NeXT's GUI is that not any little thing has some customization possibilities... SO, make it configurable (like ALL interactions should be) then everyone could be happy! cheerskinch
From: sean@zapotec.math.byu.edu (Sean O. Luke) Newsgroups: comp.sys.next.bugs Subject: Re: Suggestion: 'active' scrollbars Date: 24 May 1994 18:14:05 GMT Organization: Brigham Young University Distribution: world Message-ID: <2rtg5d$6dg@hamblin.math.byu.edu> References: <8054@darmstadt.gmd.de> <CHRIS.94May22191421arkin@nice.usergroup.ethz.ch> <2rq8m5$hrd@omega.gmd.de> Holger Hoffstaette (hoff@kylldarmstadt.gmd.de) wrote: : As to the titlebar, resizers, etc. : buttons already show that they : are being pressed, and resizing a window brings up the outline, : so there's no need to change anything there.. I feel very strongly that feedback should be attached to _action_. When you press a button, you're performing an action--triggering something. When you press the little scroll-bar arrow buttons, you're performing an action--scrolling up or down. But when you press on a title bar of a window or a scroll bar, you're not doing anything. You only do something when you _move_ the scroll bar or the title bar. And that's when you get the feedback: the window moves. Or the ScrollView scrolls in real time. I think NeXT's feedback is excellent. Motif, which gives feedback for non- actions like just pressing the titlebar of a window, is poor. +---------------------------------------------------------+ | Sean Luke This signature no verb | | sean@digaudio.byu.edu ,,, | | sean@zapotec.math.byu.edu (o o) UCLA CS in September | +------------------------oOO--(_)--OOo--------------------+
From: bobs@pth3.bu.edu (Robert Singleton) Newsgroups: comp.sys.next.bugs Subject: Annoying Icon Problem II Date: 24 May 1994 19:07:19 GMT Organization: Boston University Message-ID: <2rtj97$6ev@news.bu.edu> Recently I wrote concerning an annoying icon problem I'm having. When I'm logged in as root, the home directory bobs (which is my usual home directory) appears as a regular folder - it doesn't have the double house icon. But if I'm logged in as bobs, the usual single house icon appears. Thanks very much to all those who replied to me. The most common suggestion was that I had a non-unique UID. I checked this and it was not the case. But I now have more information on the problem. Someone sent me a .dir.tiff file to try out (a yellow sun-like picture). I placed the .dir.tiff file in my home directory and of course my home folder icon became a yellow sun - and this happened logged in as myself *and* as root. So apparently when I'm logged in as root, it doesn't know which icon to use. There must be a directory with the appropriate icons, and a way to set a search path, but I've made no progress along these lines. Any suggestions would be greatly appreciated. regards, -- bob singleton bobs@pth3.bu.edu
From: chris@nice.usergroup.ethz.ch (Christian Limpach) Newsgroups: comp.sys.next.bugs Subject: Re: Suggestion: 'active' scrollbars Date: 24 May 1994 23:49:47 GMT Organization: welcome to nowhere... Distribution: world Message-ID: <CHRIS.94May25014947arkin@nice.usergroup.ethz.ch> References: <8054@darmstadt.gmd.de> <CHRIS.94May22191421arkin@nice.usergroup.ethz.ch> <2rq8m5$hrd@omega.gmd.de> In-reply-to: hoff@kylldarmstadt.gmd.de's message of 23 May 1994 12:48:05 GMT Originator: ARKIN@nice In article <2rq8m5$hrd@omega.gmd.de> hoff@kylldarmstadt.gmd.de (Holger Hoffstaette) writes: > But that's the whole point of visual feedback: if the slider is real, > I can feel it. Since I'm dealing with a 'virtual' slider when I use > a GUI, I need some kind of 'virtual' feeling, too - and this is > currently only possible by changing color, style, etc. > This is especially important for GUIs which work in high resolutions, > like NEXTSTEP usually does. well, I don't think that it's important, but that's not something one can argue about, I don't like 3D stuff and such like at all. > As to the titlebar, resizers, etc. : buttons already show that they > are being pressed, and resizing a window brings up the outline, > so there's no need to change anything there.. the resizer doesn't show you that it's being pressed until you start moving around (ok, there is a nearly invisible change) and the titlebar doesn't give you any feedback at all until you move the window... So why should the slider change, you can move it and see if it moves (just like the titlebar) The buttons have to give you some kind of visual feedback because there's nothing happening if you move around... > I agree - not _every_ little thing. But at least those which make up > a standard (style guide conformant) GUI. well, I consider the slider-feedback a little thing which doesn't need feedback... anyway, if NeXT should change it (which hopefully won't happen [yes, I am actually asking them no to do something :-)]) I would like it to be configurable... > Well..another possibility would be to assign sounds to each of the > different GUI elements. If you press a button, it says 'ouch!'. ;-) and me would remove the speaker if there was no other way to turn it off :-) [just like I have removed the Sounds from PB :-)] christian
From: gad@eclipse.its.rpi.edu (Garance A. Drosehn) Newsgroups: comp.sys.next.bugs Subject: Re: Suggestion: 'active' scrollbars Date: 25 May 1994 00:33:51 GMT Organization: Rensselaer Polytechnic Institute, Troy NY, USA Distribution: world Message-ID: <2ru6df$rg5@usenet.rpi.edu> References: <CHRIS.94May25014947arkin@nice.usergroup.ethz.ch> someone wrote: > > But that's the whole point of visual feedback: While we are arguing the whole point of visual feedback, please note that the whole point of comp.sys.next.bugs is for reporting bugs. It is not for debate on UI features. Please migrate this discussion to one of the other next-related newsgroups. -- Garance Alistair Drosehn = gad@eclipse.its.rpi.edu ITS Systems Programmer (handles NeXT-type mail) Rensselaer Polytechnic Institute; Troy NY USA
From: dillon@apollo.west.oic.com (Matthew Dillon) Newsgroups: comp.sys.next.bugs,comp.sys.next.software,comp.sys.next.admin Subject: Re: Serial driver broken in NS 3.2 Date: 24 May 1994 15:04:08 -0700 Organization: Obvious Implementations Corp Distribution: world Message-ID: <2rttko$2ia@apollo.west.oic.com> References: <2rs19r$o2i@news.iastate.edu> Keywords: serial,driver,ns,NS,NeXTSTEP In article <2rs19r$o2i@news.iastate.edu> chris@iastate.edu (Chris Wong) writes: :The serial driver in NS 3.2 has broken again. : :Config: : Hayes OPTIMA 144 external modem connected to COM1 : DELL 450DE/2 JAWS running NeXTSTEP 3.2 : :Proof: > Running kermit under NS 3.2 via serial ports above 9600bps would cause :characters to drop (even in interactive screen display), but this doesn't >happen under OS/2 or DOS. : :Anyone got a fix? : :Does it imply that I cannot send FAX/DATA at higher than 9600bps? : :Thanks for any info. : :Chris :-- :NeXTMail super welcomed!!| :Chris Wong | "Hardware is supposed to serve Software." :chris@iastate.edu | Computer Engineering & Computer Science :<:)>:)<:)>:)<:)>:)<:) | Iowa State University of Science and Technology * Make sure you are using the port with hardware handshaking ... /dev/cufa, /dev/ttyfa, etc... (I think). * Make sure you are using V.42[bis] connections... i.e. an error checked connection on your modem, otherwise there is no end-to-end hardware flow control even when there is computer<->modem hardware flow control. * Assuming NeXT supports the 16550's FIFOs, upgrade your serial ports to 16550's instead of 16450's (for PC/compatibles). If it is still screwing up, scream at NeXT ... their serial drivers have been broken from day one. I have heard, though, that reliable third party drivers are available... could someone post the company name? -Matt -- Matthew Dillon dillon@apollo.west.oic.com 1005 Apollo Way ham: KC6LVW (no mail drop) Incline Village, NV. 89451 Obvious Implementations Corporation USA Sandel-Avery Engineering [always include a portion of the original email in any response!]
From: bobs@pth3.bu.edu (Robert Singleton) Newsgroups: comp.sys.next.bugs Subject: Re: Annoying Icon Problem II Date: 25 May 1994 15:05:51 GMT Organization: Boston University Message-ID: <2rvpgf$2jq@news.bu.edu> References: <2rtj97$6ev@news.bu.edu> Thanks, to all those who gave me a hand with this. It turns out that I had "Large File System" selected in the Preferences.app in root. I unselected it and now the double house icon for my home directory shows up. regards, -- bob singleton bobs@pth3.bu.edu
From: eagle@catt.ncsu.edu (Daniel C. L'Hommedieu) Newsgroups: comp.sys.next.bugs,comp.sys.next.sysadmin Subject: System panic again! Date: 25 May 1994 16:06:10 GMT Organization: North Carolina State University Message-ID: <2rvt1i$ips@taco.cc.ncsu.edu> Okay, my system panicked again today. I was running nn this time, and I mailed an article to myself. When I selected "no .sig" and hit ENTER, BAM! System panic on "uncaught exception 15". That's what it is every time. It also says system dies due to Window Server death. To my knowledge, I have not changed anything on my system in quite some time (months). It has done this system panic thing 3 times in the last 48 hours. Any ideas or help? Thanks. Daniel -- Daniel "eagle" L'Hommedieu Daniel_LHommedieu@nest.catt.ncsu.edu eagle@nest.catt.ncsu.edu (NeXT Mail preferred)
Newsgroups: comp.sys.next.bugs From: samr@wri.com Subject: unknown #pragmas not ignored in cc Message-ID: <samr.769901353@dragonfly.wri.com> Keywords: pragma Sender: news@wri.com ( ) Organization: Wolfram Research, Inc. Date: Wed, 25 May 1994 21:29:13 GMT this code is ok with NeXT 2.x, but chokes 3.x cc: tst() { int a; #pragma unused( unused_cookie) int b; } main(){ tst(); } saipan [100] samr> cc tst.c tst.c:6: illegal expression, found `int' and... ANSI X3.159-1989 section 3.8.6 "... Any pragma not recognized by an implementation is ignored." -- Sam Roseman Wolfram Research, Inc. GamePro Amusements, Inc. samr@wri.com 100 Trade Center Drive P.O. Box 6173 Champaign, IL 61820 Champaign, IL 61826-6173 (217) 398-0700 (217) 355-6217
Newsgroups: comp.sys.next.advocacy,comp.sys.next.bugs From: rwyatt@netcom.com (Robert Wyatt) Subject: Bugs, bugs, bugs (system panic) Message-ID: <rwyattCqDqH8.1yF@netcom.com> Organization: Netcom Online Communications Services (408-241-9760 login: guest) Date: Wed, 25 May 1994 22:29:31 GMT I'm cross-posting this to both advocacy and bugs as I'm interested in both a "technical" and "user" perspective... I've been beta testing Virtuoso 2 from Altsys for a month or so now. It's a superb app except for one darn (very major) problem: system crashes! I have a DEC MTE system (from NeXT's "approved" list). Under V2, I get several "Kernal Traps" per day, causing me to have to shut down the whole system and reboot. In the process, files are sometimes lost. In speaking with Altsys, they can't reproduce my bugs. Their DELL system doesn't crash and neither do the systems of their other Intel beta testers. So, presumably there is a problem between the hardware, NEXTSTEP, and Virtuso 2. Has anyone else experienced such glaring problems with NS/Intel? All of my other apps work great. I really hate to think that NEXTSTEP is so finicky that an app can crash one computer and not another??? Just interested in your two cents... -Rob rwyatt@netcom.com (no NeXTmail)
From: erikkay@next.com (Erik Kay) Newsgroups: comp.sys.next.bugs Subject: Re: unknown #pragmas not ignored in cc Date: 26 May 1994 02:45:05 GMT Organization: NeXT, Inc. Message-ID: <2s12fh$80t@rosie.next.com> References: <samr.769901353@dragonfly.wri.com> In article <samr.769901353@dragonfly.wri.com> samr@wri.com writes: > this code is ok with NeXT 2.x, but chokes 3.x cc: > > tst() > { > int a; > #pragma unused( unused_cookie) > int b; > } > > main(){ tst(); } > > saipan [100] samr> cc tst.c > tst.c:6: illegal expression, found `int' > > > and... > > ANSI X3.159-1989 section 3.8.6 > "... Any pragma not recognized by an implementation is ignored." This appears to be a bug in cpp-precomp. To get around it for the time being, use -traditional-cpp on the file you're having a problem with. (note that any file you compile with this option on will not use precomps) So in this case: $ cc -o foo foo.c foo.c:5: illegal expression, found `int' $ cc -traditional-cpp -o foo foo.c $ Hope this helps for the time being. Erik std disclaimer: I don't speak for NeXT, and NeXT doesn't speak for me
Newsgroups: comp.sys.next.bugs From: mgoedel@muaddib.isar.muc.de (Maximilian Goedel) Subject: Re: Annoying Icon Problem II Message-ID: <CqE5vC.F8E@muaddib.isar.de> Sender: mgoedel@muaddib.isar.de (Maximilian Goedel) Organization: Michael Maximilian Goedel References: <2rtj97$6ev@news.bu.edu> Date: Thu, 26 May 1994 04:02:00 GMT In article <2rtj97$6ev@news.bu.edu> bobs@pth3.bu.edu (Robert Singleton) writes: | Recently I wrote concerning an annoying icon problem I'm | having. When I'm logged in as root, the home directory | bobs (which is my usual home directory) appears as a regular | folder - it doesn't have the double house icon. But if I'm | logged in as bobs, the usual single house icon appears. | | Thanks very much to all those who replied to me. The most | common suggestion was that I had a non-unique UID. I checked | this and it was not the case. But I now have more information | on the problem. Someone sent me a .dir.tiff file to try out | (a yellow sun-like picture). I placed the .dir.tiff file in my | home directory and of course my home folder icon became a yellow | sun - and this happened logged in as myself *and* as root. So | apparently when I'm logged in as root, it doesn't know which icon | to use. There must be a directory with the appropriate icons, and | a way to set a search path, but I've made no progress along these | lines. Any suggestions would be greatly appreciated. | | regards, | | | -- | bob singleton | bobs@pth3.bu.edu The icons for the workspace can be found on /usr/lib/NextStep/Workspace.app and deeper directories. Check also, if preferences are set correctly in the part UNIX, where you can set the Options in the filesystem. I got similar the same problem with an uncorrect setting in the root account. Nice Greetings from Munich -- Michael Maximilian Goedel email_____________________________ Gerhardstrasse 33 NeXT: mgoedel@muaddib.isar.muc.de 81543 Muenchen - GERMANY LINUX: max@funman.boss.sub.org Phone +49-89-652918 SGI: f11cs1@rz.unibw-muenchen.de
Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs From: andrew@stone.com (Andrew Stone) Subject: CANNOT setAccessoryView: in NXDataLinkPanel: WORKAROUND Message-ID: <1994May25.172338.2359@stone.com> Keywords: NXDataLinkPanel AccessoryView Good Boy Sender: andrew@stone.com Organization: Stone Design Corp Date: Wed, 25 May 1994 17:23:38 GMT This should save anyone trying to add an accessory view to the NXDataLinkPanel alot of time. Due to a bug in NS 3.2, when you call NXDataLinkPanel's setAccessoryView:, you get this assertion: Assertion failed: No box found in NXDataLinkPanel for setAccessoryView Here's a Workaround, which will "fix itself" when NeXT fixes the bug! Include this code somewhere in your project: @interface NXDataLinkPanel(cheaterMethods) - getOldBreakAllLinks; - resetBreakAllLinks:oldBreakAllLinks; @end @implementation NXDataLinkPanel(cheaterMethods) - getOldBreakAllLinks { id oldBreakAllLinks = breakAllLinks; breakAllLinks = lastUpdateTitle; return oldBreakAllLinks; } - resetBreakAllLinks:oldBreakAllLinks { breakAllLinks = oldBreakAllLinks; return self; } @end In your Application subclass overrride orderFrontDataLinkPanel: // SDLinkAccessory is the controller object for our custom accessory // the actual accessoryView is accessible via the accessoryView method - orderFrontDataLinkPanel:sender { static BOOL first = YES; if (first) { id dataLinkPanel = [NXDataLinkPanel new]; id acc = [SDLinkAccessory new]; if (![dataLinkPanel setAccessoryView:[acc accessoryView]]) { id oldBreakAllLinks = [dataLinkPanel getOldBreakAllLinks]; [dataLinkPanel setAccessoryView:[acc accessoryView]]; [dataLinkPanel resetBreakAllLinks:oldBreakAllLinks]; } first = NO; } [super orderFrontDataLinkPanel:sender]; return self; } That will be $5000 in cybercash, please. ;-) andrew -- ||<<->>||<<==>>|S<<++>>|T<<?>O<+>>N|<<-->>E|<<==>>||<<+>>|| !! Andrew Stone | (505) 345-4800 !! !! andrew@stone.com | Have Modem Why Travel !! ||<<->>|D<<==>>|E<<++>>|S<<?>|<+>>G|<<-->>N|<<==>>!|<<+>>||
From: mow@marsu.tynet.sub.org (Markus Wenzel) Newsgroups: comp.sys.next.bugs,comp.sys.next.software,comp.sys.next.admin Subject: Re: Serial driver broken in NS 3.2 Date: 25 May 1994 12:52:24 +0100 Organization: Palumbian Research Labs Message-ID: <2rve5o$bqk@marsu.tynet.sub.org> References: <2rs19r$o2i@news.iastate.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Keywords: serial,driver,ns,NS,NeXTSTEP chris@iastate.edu (Chris Wong) writes: >The serial driver in NS 3.2 has broken again. Why again? This is a known bug/problem since the release of 3.2. >Anyone got a fix? Get the Mux driver by ftp. >Does it imply that I cannot send FAX/DATA at higher than 9600bps? With the original NeXT serial driver: yes. With Mux, you can go for any rate you want, it just works fine for me at 38400/16800 -> 2000 cps effective throughput. Markus. -- /dev/ Markus Wenzel /usr/spool/mail/ mow@marsu.tynet.sub.org /etc/zoneinfo/ University of Stuttgart /bin/ps System administration, Consulting, Networking
From: chris@nice.usergroup.ethz.ch (Christian Limpach) Newsgroups: comp.sys.next.bugs,comp.sys.next.misc Subject: Re: Suggestion: 'active' scrollbars Followup-To: comp.sys.next.misc Date: 25 May 1994 23:21:30 GMT Organization: welcome to nowhere... Message-ID: <CHRIS.94May26012131arkin@nice.usergroup.ethz.ch> References: <8054@darmstadt.gmd.de> <CHRIS.94May22191421arkin@nice.usergroup.ethz.ch> <2rteq7$loi@falcon.ccs.uwo.ca> In-reply-to: kinch@jones.heart.rri.uwo.ca's message of 24 May 1994 17:51:03 GMT Originator: ARKIN@nice [Followup-To: set to comp.sys.next.misc] In article <2rteq7$loi@falcon.ccs.uwo.ca> kinch@jones.heart.rri.uwo.ca (Dave Kinchlea) writes: > >I wouldn't like to have this. It would IMHO be quite useless. After > >all, real-world sliders don't give you a visual feedback if you use > > No, real-world sliders provide a tactile feedback so you don't NEED a > 3d feedback. but you get a feedback as soon as you move the slider. And if you click on sliders and don't move them, then you have too much spare time and you're just playing around. I think that even then, moving the slider around, is more fun than just watching some boring 3D effect :-) > >them, GUI buttons just like real-world buttons do... And what about > >the titlebar, the resizers ? Do you want to change these as well ? > > And why not? In what way would these hurt? they would require NeXT to spend time on this, time which they could use to do some more important and more needed things... > One of the things I least appreciate in NeXT's GUI is that not any > little thing has some customization possibilities... yes indeed, I would like to have some more customization possibilities too. > SO, make it configurable (like ALL interactions should be) then > everyone could be happy! agreed. christian
Newsgroups: comp.sys.next.bugs From: arrouye@petole.imag.fr (Yves Arrouye) Subject: Re: Annoying Icon Problem II Message-ID: <1994May26.141957.13079@imag.fr> Sender: news@imag.fr Organization: Institut Imag, Grenoble, France References: <CqE5vC.F8E@muaddib.isar.de> Date: Thu, 26 May 1994 14:19:57 GMT In article <CqE5vC.F8E@muaddib.isar.de> mgoedel@muaddib.isar.muc.de (Maximilian Goedel) writes: > > The icons for the workspace can be found on > /usr/lib/NextStep/Workspace.app and deeper directories. > Check also, if preferences are set correctly in the part UNIX, > where you can set the Options in the filesystem. I got similar > the same problem with an uncorrect setting in the root account. > Eventually also check if the 'Large System File' is checked. Once I unchecked it (I still do not know what it does, anyway, so if someone can give me a clue...) everything worked fine on my machine where I experienced similar problems. Yves -- Advocates for the C++ school claim that a well designed Yves Arrouye program does not need the extra flexibility (a lie), Yves.Arrouye@imag.fr while advocates for the Objective-C school claim that (33) 76 57 48 64 the errors are no problem in practice (another lie). NeXT Mail
From: elitman@proxima.com (Eric A. Litman) Newsgroups: comp.sys.next.advocacy,comp.sys.next.bugs Subject: Re: Bugs, bugs, bugs (system panic) Date: 26 May 1994 09:45:33 -0500 Organization: Proxima, Inc. Sender: nobody@cs.utexas.edu Message-ID: <9405261445.AA08161@proxima.com> In article <rwyattCqDqH8.1yF@netcom.com> rwyatt@netcom.com writes: > Has anyone else experienced such glaring problems with NS/Intel? All of > my other apps work great. I really hate to think that NEXTSTEP is so > finicky that an app can crash one computer and not another??? Just > interested in your two cents... We have over 350 units installed throughout the country, and have experienced regular problems with kernel panics. More often than not, the crashed system needs to be rebuilt from CD as many of the low-numbered inodes are lost, rendering the system unusable. The systems we are using are Intel Pro/GX's w/ DPT 2022 EISA controllers, and Intel Premiere Pentium's with Adaptec 1542CF ISA controllers. We've gone through close to a dozen DPT firmware releases, none of which seem to be any more stable than the previous other. The Adaptec-based systems crash less often, but still crash nonetheless. Some of our developers have had problems with their personal systems (Gateway and DEC), which leads me to believe that the problems are not isolated to the Intel boxes we widely use. I have put a hold on all Intel architecture purchases (any systems, not just those manufactured by Intel) in our organization in hopes that NS running on HP will resolve these issues. It is extremely frustrating to provide custom application solutions for our customers with no stable system as a base. So the question now becomes, which system component(s) is to blame for this? Is it the firmware of the SCSI controller? I would certainly consider this possibility, but am discouraged to do so by the failures of more than one board type. Is it the drivers shipped by NeXT, or perhaps underlying DriverKit? Again, this is certainly plausible, but discomforting as the system crashes have occurred invariably in both 3.1 and 3.2. DPT is aware of the problems and insists that it is not their fault. I would be encouraged by a proactive response from NeXT, but have unfortunately not received this. Perhaps the new DPT driver will address the issue. -- Eric Litman Proxima, Inc. vox: (703) 506.1661 Director, ProxNet McLean, VA elitman+@proxima.com
From: dillon@apollo.west.oic.com (Matthew Dillon) Newsgroups: comp.sys.next.advocacy,comp.sys.next.bugs Subject: Re: Bugs, bugs, bugs (system panic) Date: 26 May 1994 08:57:23 -0700 Organization: Obvious Implementations Corp Distribution: world Message-ID: <2s2gt3$t70@apollo.west.oic.com> References: <9405261445.AA08161@proxima.com> In article <9405261445.AA08161@proxima.com> elitman@proxima.com (Eric A. Litman) writes: :In article <rwyattCqDqH8.1yF@netcom.com> rwyatt@netcom.com writes: : :> Has anyone else experienced such glaring problems with NS/Intel? All of :> my other apps work great. I really hate to think that NEXTSTEP is so :> finicky that an app can crash one computer and not another??? Just :> interested in your two cents... : :We have over 350 units installed throughout the country, and have :experienced regular problems with kernel panics. More often than :... :I have put a hold on all Intel architecture purchases (any systems, :not just those manufactured by Intel) in our organization in hopes that :... :So the question now becomes, which system component(s) is to blame for :this? Is it the firmware of the SCSI controller? I would certainly :... :-- :Eric Litman Proxima, Inc. vox: (703) 506.1661 :Director, ProxNet McLean, VA elitman+@proxima.com Every once in a while we get people on the Linux groups with crashing problems that nobody else has. It has turned out to be one (or a combination of): * Incorrect BIOS level configuration (too few wait states for cache, main memory VLBUS, or some idiot suping up the main bus speed for ISA machines). * Cache rams that are out of spec (i.e. too slow) * Overheating CPUs that work just fine under DOS, but die in 32 bit linear address mode. I suggest, before you start blaming the software, that you check your hardware: (1) Try increasing the number of wait states for the cache and for main memory (write down the old configuration!) (2) A 486 DX-33 gets pretty hot without a heat sink, but still generally works. A 486 DX/2-66 *REQUIRES* a heat sink. I strongly suggest a high quality heat sink *without* a fan (which only adds additional failure modes) for any high speed processor including a DX-33. It is amazing how many vendors, such as Micronics, think that their DX/2-66 motherboards will work without a heat sink. Simply put, you are asking for it if you do not have a heat sink. We also avoid 50MHz bus systems (DX-50) for several reasons. Not only do many 50 MHz bus motherboard manufacturers not have a clue about how to construct them, but most VL-Bus boards *cannot* take advantage of the higher bus speed and either must be strapped for additional wait states or not used at all. In many cases, main memory and/or cache must also be strapped for more wait states which kinda kills the whole reason for buying a 50 MHz system in the first place! The tolerances for 50MHz systems are simply too tight for our tastes. The hassle is not worth the incremental improvement. -Matt -- Matthew Dillon dillon@apollo.west.oic.com 1005 Apollo Way ham: KC6LVW (no mail drop) Incline Village, NV. 89451 Obvious Implementations Corporation USA Sandel-Avery Engineering [always include a portion of the original email in any response!]
Newsgroups: comp.sys.next.bugs From: kent@infoserv.com Subject: Re: Annoying Icon Problem II Message-ID: <CqE0uB.2JD@infoserv.com> Sender: kent@infoserv.com (Kent L. Shephard) Organization: K. L. Shephard Consulting References: <2rtj97$6ev@news.bu.edu> Distribution: na Date: Thu, 26 May 1994 02:13:23 GMT In article <2rtj97$6ev@news.bu.edu> bobs@pth3.bu.edu (Robert Singleton) writes: #Recently I wrote concerning an annoying icon problem I'm #having. When I'm logged in as root, the home directory #bobs (which is my usual home directory) appears as a regular #folder - it doesn't have the double house icon. But if I'm #logged in as bobs, the usual single house icon appears. # #Thanks very much to all those who replied to me. The most #common suggestion was that I had a non-unique UID. I checked #this and it was not the case. But I now have more information #on the problem. Someone sent me a .dir.tiff file to try out #(a yellow sun-like picture). I placed the .dir.tiff file in my #home directory and of course my home folder icon became a yellow #sun - and this happened logged in as myself *and* as root. So #apparently when I'm logged in as root, it doesn't know which icon #to use. There must be a directory with the appropriate icons, and #a way to set a search path, but I've made no progress along these #lines. Any suggestions would be greatly appreciated. Did you try checking preferences to see if large filesystem was checked? If it is you will not automatically see users directorys as little houses. Kent -- /* K.L. Shephard Consulting is my company. Infoserv only delivers my mail. */ /* Please direct mail to kent@infoserv.com other adresses may not work. */
From: jmack@skye.phys.ualberta.ca Newsgroups: comp.sys.next.bugs,comp.sys.next.programmer Subject: NS3.2 cpp-precomp failure for 'cc -E' Date: 29 May 1994 19:53:51 GMT Organization: Computer and Network Services, U of Alberta, Edmonton, Canada Message-ID: <2sarsf$qj6@quartz.ucs.ualberta.ca> Keywords: pre-compiler,cc 'cc -E' works fine under 3.0 but breaks under 3.2 (black) It looks as if the culprit is 'cpp-precomp -smart' Under 3.2, gcc 2.5.7 works just fine: -------------------- gcc -v -E Makefile.c > Makefile.typ Reading specs from /usr/local/lib/gcc-lib/m68k-next-bsd/2.5.7/specs gcc version 2.5.7 /usr/local/lib/gcc-lib/m68k-next-bsd/2.5.7/cpp -lang-c -v -undef -D__GNUC__=2 -D__GNUC_MINOR__=5 -Dmc68000 -Dm68k -DNeXT -Dunix -D__MACH__ -D__BIG_ENDIAN__ -D__ARCHITECTURE__="m68k" -D__mc68000__ -D__m68k__ -D__NeXT__ -D__unix__ -D__MACH__ -D__BIG_ENDIAN__ -D__ARCHITECTURE__="m68k" -D__mc68000 -D__m68k -D__NeXT -D__unix -D__MACH__ -D__BIG_ENDIAN__ -D__ARCHITECTURE__="m68k" -Asystem(unix) -Asystem(mach) -Acpu(m68k) -Amachine(m68k) -D__STDC__ -D_NEXT_SOURCE Makefile.c GNU CPP version 2.5.7 (68k, MIT syntax) #include "..." search starts here: #include <...> search starts here: /usr/local/include /usr/local/m68k-next-bsd/include /usr/local/lib/gcc-lib/m68k-next-bsd/2.5.7/include /usr/local/lib/gcc-lib/m68k-next-bsd/2.5.7/include/ansi /usr/local/lib/gcc-lib/m68k-next-bsd/2.5.7/include/bsd /NextDeveloper/Headers /NextDeveloper/Headers/ansi /NextDeveloper/Headers/bsd /LocalDeveloper/Headers /LocalDeveloper/Headers/ansi /LocalDeveloper/Headers/bsd /NextDeveloper/2.0CompatibleHeaders /usr/include /usr/include/bsd End of search list. BUT, with NeXT's version of cc, this just fails: ------------------------------ cc -v -E Makefile.c > Makefile.typ Reading specs from /lib/m68k/specs NeXT Computer, Inc. version cc-216.obj~13, gcc version 2.2.2 /lib/m68k/cpp-precomp -smart -lang-c -v -arch m68k -undef -D__GNUC__=2 -D__GNU__ -Dmc68000 -Dm68k -DNeXT -Dunix -DNX_COMPILER_RELEASE_3_0=300 -DNX_COMPILER_RELEASE_3_1=310 -DNX_CURRENT_COMPILER_RELEASE=310 -D__mc68000__ -D__m68k__ -D__NeXT__ -D__unix__ -D__MACH__ -D__BIG_ENDIAN__ -D__ARCHITECTURE__=\"m68k\" -D__NX_COMPILER_RELEASE_3_0__=300 -D__NX_COMPILER_RELEASE_3_1__=310 -D__NX_CURRENT_COMPILER_RELEASE__=310 -D__mc68000 -D__m68k -D__NeXT -D__unix -D__NX_COMPILER_RELEASE_3_0=300 -D__NX_COMPILER_RELEASE_3_1=310 -D__NX_CURRENT_COMPILER_RELEASE=310 -D__STDC__ -D_NEXT_SOURCE Makefile.c NeXT DevKit-based CPP 3.1 Makefile.c:89: illegal external declaration, missing `;' after `rv' Makefile.c:90: undefined type, found `ranlib' Makefile.c:91: illegal external declaration, missing `;' after `f77' Makefile.c:92: undefined type, found `cc' Makefile.c:93: illegal external declaration, missing `;' after `g' .. (lots more errors deleted) A workaraound is to forget about cpp-precomp and just use cpp (by hand): ------------- /lib/m68k/cpp -smart -lang-c -v -arch m68k -undef -D__GNUC__=2 -D__GNU__ -Dmc68000 -Dm68k -DNeXT -Dunix -DNX_COMPILER_RELEASE_3_0=300 - DNX_COMPILER_RELEASE_3_1=310 -DNX_CURRENT_COMPILER_RELEASE=310 -D__mc68000__ -D__m68k__ -D__NeXT__ -D__unix__ -D__MACH__ -D__BIG_ENDIAN__ -D__ARCHITECTURE__=\"m68k\" -D__NX_COMPILER_RELEASE_3_0__=300 -D__NX_COMPILER_RELEASE_3_1__=310 -D__NX_CURRENT_COMPILER_RELEASE__=310 -D__mc68000 -D__m68k -D__NeXT -D__unix -D__NX_COMPILER_RELEASE_3_0=300 -D__NX_COMPILER_RELEASE_3_1=310 -D__NX_CURRENT_COMPILER_RELEASE=310 -D__STDC__ -D_NEXT_SOURCE Makefile.c > Makefile.typ GNU CPP version 2.2.2 (68k, MIT syntax) .. which produces a work-able file. When are the cpp-precomp oddities going to be fixed? -- James S. MacKinnon Office: P-139 Avahd-Bhatia Physics Lab Computing/Networking Phone : (403) 492-8226 Department of Physics email : jmack@phys.ualberta.ca University of Alberta uucp : uofaphys!jmack iskye!jmack Edmonton, Canada T6G 2N5 bitnet: jmack@triumfcl jsm1@ualtamts
From: cnayak@crl.com (Chiraprakash Nayak) Newsgroups: comp.sys.next.bugs Subject: Dos file system and "undelete" Date: 30 May 1994 02:27:36 -0700 Organization: CRL Dialup Internet Access (415) 705-6060 [login: guest] Message-ID: <2scbi8$6np@crl.crl.com> Hi; In NeXTSTEP 3.2/Intel any dos file deleted ("DESTROY") from workspace does show up in the "undelete" list when restarted in dos. File is recoverable but contains only garbage - file size is ok but all the contents are lost. I don't run soft-pc, Does it happen on soft-pc too? I will appreciate if some body can clarify. May be while dos partition size limit is being fixed, NeXt will fix this anamoly too! Thanks a lot. Best Regards. --Chip (cnayak@crl.com)
Newsgroups: comp.sys.next.bugs From: tstuve@techfak.uni-bielefeld.de (Thilo Stuve) Subject: Backspace crashed? Sender: news@hermes.hrz.uni-bielefeld.de (News Administrator) Message-ID: <CqM42B.7z7@hermes.hrz.uni-bielefeld.de> Date: Mon, 30 May 1994 11:03:47 GMT Organization: Universitaet Bielefeld, Technische Fakultaet. I was logged in as root and had to leave my NeXTStation for a while. I thought it would be nice to my monitor if I run this screensaver[this time screen- locker]. When I returned after two hours, I couldn't unlock my screen, even the panel didn't appear. After a while I reached to power off my Station. What went wrong? Thilo
From: kimage@well.sf.ca.us (Kimbrough Bassett) Newsgroups: comp.sys.next.bugs Subject: White Sound vs. SCSI Date: 30 May 1994 20:11:31 GMT Organization: The Whole Earth 'Lectronic Link, Sausalito, CA Message-ID: <2sdh9j$c0p@nkosi.well.com> Summary: White (NeXT) sound hardware vs. scsi hardware bugs Keywords: next,bugs,sound,scsi Two questions regarding the documented bug of Pro Audio Spectrum sound card incompatability with Adaptec 1540 SCSI interfaces (first cited in Intel.NeXTstep 3.1): Q1: Has anybody had success (under 3.1 or 3.2, please specify) using another manufacturer's "ProAudioSpectrum-compatible" hardware card (such as the new generation of wave-modulators-or-whatever) with NeXTstep? Do NeXTstep's drivers drive those cards or is it a hopeless idea? If NeXTstep can use those cards, then can they be used alongside my Adaptec under 3.1 to circumvent the incompatability bug? Q2: Does 3.2 fix the incompatability bug? Thanks for helping me avoid expensive guesswork. - Kim Bassett
From: sanguish@digifix.com Newsgroups: comp.sys.next.announce,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.bugs,comp.sys.next.programmer,comp.sys.next.marketplace Subject: Quick Guide to NEXTSTEP Information on the Internet. Date: 30 May 1994 16:28:39 -0400 Organization: Next Announcements Message-ID: <2sdi9n$fer@digifix.digifix.com> This post is made weekly, to help 'point' users to more NEXTSTEP information Topics include: Stepwise NEXTSTEP/OpenStep Information Server online comp.sys.next newsgroups related newsgroups comp.sys.next newsgroups mailing list ftp sites NeXTanswers New Information --------------- Stepwise NEXTSTEP/OpenStep Information Server online A product directory built around the World Wide Web system, this will allow full multimedia announcements by NEXTSTEP developers, as well as the ability to browse the available products for NEXTSTEP. This service is online now, and can be reached at http://digifix.digifix.com/ it can be reached using OmniWeb (available from ftp.omnigroup.com) or Mosaic. The entries are coming in quite quickly, and currently consist of - NeXT Press Releases - OpenStep WhitePapers - Third Party Products Directory - NEXTSTEP Developer Directory - NEXTSTEP Community WhitePages - NEXTSTEP User Group Directory - comp.sys.next.announce archives (searchable) - searchable contents of Third Party compilation CDs Additionally the NEXTSTEP Product Information Mail Server will be stocked full file files in the next week... you can get information on using the mail server at ns-products@digifix.com Suggestions or comments can be directed to me at sanguish@digifix.com comp.sys.next.* newsgroups -------------------------- Comp.Sys.Next.Advocacy This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. Comp.Sys.Next.Announce Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) The NEXTSTEP FAQs are posted here monthly as well. This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Comp.Sys.Next.Bugs A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT- specific groups as well. Comp.Sys.Next.Hardware Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. Comp.Sys.Next.Marketplace NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. Comp.Sys.Next.Misc For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! Comp.Sys.Next.Programmer Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. The NEXTSTEP programmer FAQs are posted here. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. Comp.Sys.Next.Software This is a place to talk about [third party] software products that run on NEXTSTEP systems. Comp.Sys.Next.Sysadmin Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. related Newsgroups ------------------ Comp.Soft-Sys.Nextstep Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. Comp.Lang.Objective-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. Comp.Object Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original Comp.Sys.Next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News ------------------------------------------- Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep-d next-advocacy-d next-announce-d next-bugs-d next-hardware-d next-marketplace-d next-misc-d next-programmer-d next-software-d next-sysadmin-d (For a full description, send mail saying LISTS to <digestif@antigone.com>). The subscription syntax is essentially the same as LISTSERV's. To subscribe, send a message to <digestif@antigone.com> saying: SUB Listname YourName Example: SUB next-hardware-d John Doe The ftp sites ------------- cs.orst.edu: The main site for North American submissions nova.cc.purdue.edu: Lots of older stuff, but very short on disk space ftp.informatik.uni-muenchen.de: In Germany. terra.stack.urc.tue.nl (Dutch NEXTSTEP User Group) and cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp.next.com: See the below ftp.next.com and NextAnswers@next.com ------------------------------------- From the document 1000_Help from ftp.next.com Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, or you can transfer them by anonymous ftp. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to NeXTanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO <your-address> If you have any problem with the system or suggestions for improvement, please send mail to NeXTanswers-request@NeXT.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to NeXTanswers-request@NeXT.com. Written by: Eric P. Scott eps@toaster.SFSU.EDU and Scott Anguish sanguish@digifix.com Additions from: Greg Anderson (Greg_Anderson@afs.com) and Michael Pizolato (Michael_Pizolato@afs.com)
Newsgroups: comp.sys.next.bugs From: csaldanh@mae.carleton.ca (Chris Saldanha) Subject: Re: Dos file system and "undelete" Message-ID: <CqnFwK.BxK@cunews.carleton.ca> Sender: news@cunews.carleton.ca (News Administrator) Organization: Carleton University References: <2scbi8$6np@crl.crl.com> Date: Tue, 31 May 1994 04:17:08 GMT Chiraprakash Nayak (cnayak@crl.com) wrote: : In NeXTSTEP 3.2/Intel any dos file deleted ("DESTROY") from workspace : does show up in the "undelete" list when restarted in dos. File is : recoverable but contains only garbage - file size is ok but all the : contents are lost. I don't run soft-pc, Does it happen on soft-pc too? : I will appreciate if some body can clarify. May be while dos partition : size limit is being fixed, NeXt will fix this anamoly too! I remember reading once that the 'Destroy' menu option actually over-wrote the file with garbage deliberately, so as to be sure that the file was _really_ gone forever; as a security measure. I find no mention of this in any documentation, so maybe I heard wrong... --Chris Chris Saldanha -------------------------------------- Carleton University (Comp. Sci) |"The eternal silence of these infinite| chris@computerActive.on.ca (NeXTMail) | spaces terrifies me." -Blaise Pascal| csaldanh@mae.carleton.ca (NeXT/MIME) ------------------:-o-----------------
From: mow@marsu.tynet.sub.org (Markus Wenzel) Newsgroups: comp.sys.next.bugs Subject: Re: White Sound vs. SCSI Date: 31 May 1994 09:12:59 +0100 Organization: Palumbian Research Labs Message-ID: <2serib$2n8@marsu.tynet.sub.org> References: <2sdh9j$c0p@nkosi.well.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Keywords: next,bugs,sound,scsi kimage@well.sf.ca.us (Kimbrough Bassett) writes: >Two questions regarding the documented bug of Pro Audio >Spectrum sound card incompatability with Adaptec 1540 SCSI >interfaces (first cited in Intel.NeXTstep 3.1): >Q2: Does 3.2 fix the incompatability bug? No. I had crashes with PAS16 + 1542 as before. They disappeared as I upgraded to an EISA/VL board. Markus. -- /dev/ Markus Wenzel /usr/spool/mail/ mow@marsu.tynet.sub.org /etc/zoneinfo/ University of Stuttgart /bin/ps System administration, Consulting, Networking
From: erikkay@next.com (Erik Kay) Newsgroups: comp.sys.next.bugs Subject: Re: NS3.2 cpp-precomp failure for 'cc -E' Date: 31 May 1994 17:52:56 GMT Organization: NeXT, Inc. Message-ID: <2sftho$eog@rosie.next.com> References: <2sarsf$qj6@quartz.ucs.ualberta.ca> In article <2sarsf$qj6@quartz.ucs.ualberta.ca> jmack@skye.phys.ualberta.ca writes: ] 'cc -E' works fine under 3.0 but breaks under 3.2 (black) For us to be able to track bugs like this down, you'll need to include a source code example that can reproduce the problem. Erik
From: erikkay@next.com (Erik Kay) Newsgroups: comp.sys.next.bugs Subject: Re: Dos file system and "undelete" Date: 31 May 1994 18:11:27 GMT Organization: NeXT, Inc. Message-ID: <2sfukf$eov@rosie.next.com> References: <CqnFwK.BxK@cunews.carleton.ca> In article <CqnFwK.BxK@cunews.carleton.ca> c Chiraprakash Nayak (cnayak@crl.com) wrote: ] : I will appreciate if some body can clarify. May be while dos partition ] : size limit is being fixed, NeXt will fix this anamoly too! Actually, it won't. This is the first report we've had against this bug. I'll see if I can get to it in time for 3.3. saldanh@mae.carleton.ca (Chris Saldanha) writes: ] I remember reading once that the 'Destroy' menu option actually ] over-wrote the file with garbage deliberately, so as to be sure that the ] file was _really_ gone forever; as a security measure. ] ] I find no mention of this in any documentation, so maybe I heard wrong... You heard wrong... it just uses unlink and rmdir. Erik my opinions are my own
Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs From: Greg_Anderson@afs.com (Gregory H. Anderson) Subject: BUG: NXMapFile() doesn't release file descriptor on NXClose() Message-ID: <1994May31.151144.16049@afs.com> Sender: greg@afs.com Date: Tue, 31 May 1994 15:11:44 GMT Reply-To: Greg_Anderson@afs.com "Cannot eject disk: some application is using it" Look familiar? You may have puzzled over why removable media like floppies sometimes won't unmount until an application quits, even if all documents opened from that disk have been closed. Edit.app used to do this, but the latest version is fixed. FileMerge.app, on the other hand, still exhibits this annoying quirk. So did WriteUp and PasteUp, until today. I've found the culprit: NXMapFile(). When you're working with streams from files, NeXT recommends this code stream = NXMapFile(const char *name, int mode); but when you later NXClose() the stream, the file descriptor associated with it does _not_ get closed. And since the stream won't tell you what file descriptor it is opened on, there's no way to close it yourself. As a result, it becomes impossible for Workspace to unmount the device from which the file was retrieved, even if your app thinks it is closed. Workaround: End-users cannot workaround this problem for applications that are coded improperly. For programmers, instead of NXMapFile(), use this two-part construction: if ((fd = open(const char *name, int unixMode))) // O_RDONLY, etc. stream = NXOpenFile(int fd, int NeXTMode); // NX_READONLY, etc. and then at the bottom of your code, do both NXClose(stream); close(fd); Caveat for typedstreams: The composite method NXOpenTypedStreamForFile() seems to be implemented correctly; that is to say, the file descriptor allocated by this function does seem to get released by NXClose(). Floppies don't seem to get jammed is this case. -- Gregory H. Anderson | "Internet: a giant international network Stud Hombre Cybermuffin | of intelligent, informed computer Anderson Financial Systems | enthusiasts, by which I mean 'people greg@afs.com (NeXTmail OK) | without lives.'" -- Dave Barry, 2/6/94
From: Mark_Wagner@NeXT.COM (Mark Wagner) Newsgroups: comp.sys.next.bugs Subject: Re: NS3.2 cpp-precomp failure for 'cc -E' Date: 31 May 1994 22:46:45 GMT Organization: NeXT, Inc. Message-ID: <2sgeol$f0o@rosie.next.com> References: <2sftho$eog@rosie.next.com> In article <2sftho$eog@rosie.next.com> erikkay@next.com (Erik Kay) writes: > In article <2sarsf$qj6@quartz.ucs.ualberta.ca> jmack@skye.phys.ualberta.ca > writes: > ] 'cc -E' works fine under 3.0 but breaks under 3.2 (black) > > For us to be able to track bugs like this down, you'll need to include a > source code example that can reproduce the problem. > > Erik I didn't see the original post but... between 3.0 and 3.2 the semantics of cc -E were changed, slightly. Specifically, we found that more often than not when someone typed "cc -E" they wanted to see "what the compiler proper would be seeing if they had left off the -E". Thus, we changed the -E flag to mean "continue to run cpp-precomp but put the results to stdout, instead of to cc1obj". This will not work if the file you're running on is not valid code since the precompiler will not be able to process it. If you're running cc -E to get the effect of running a file through the preprocessor then you should run "cc -E -no-precomp" or "cc -E -traditional-cpp". I believe that there is a release note that also describes this. -- Mark
From: erikkay@next.com (Erik Kay) Newsgroups: comp.sys.next.bugs Subject: Re: BUG: NXMapFile() doesn't release file descriptor on NXClose() Date: 31 May 1994 23:42:27 GMT Organization: NeXT, Inc. Message-ID: <2sgi13$f2t@rosie.next.com> References: <1994May31.151144.16049@afs.com> In article <1994May31.151144.16049@afs.com> Greg_Anderson@afs.com (Gregory H. Anderson) writes: ] Reply-To: Greg_Anderson@afs.com ] ] "Cannot eject disk: some application is using it" ] ] Look familiar? You may have puzzled over why removable media like floppies ] sometimes won't unmount until an application quits, even if all documents ] opened from that disk have been closed. Edit.app used to do this, but the ] latest version is fixed. FileMerge.app, on the other hand, still exhibits ] this annoying quirk. So did WriteUp and PasteUp, until today. ] ] I've found the culprit: NXMapFile(). When you're working with streams from ] files, NeXT recommends this code ] ] stream = NXMapFile(const char *name, int mode); ] ] but when you later NXClose() the stream, the file descriptor associated ] with it does _not_ get closed. And since the stream won't tell you what ] file descriptor it is opened on, there's no way to close it yourself. As ] a result, it becomes impossible for Workspace to unmount the device from ] which the file was retrieved, even if your app thinks it is closed. Actually, this is not the problem. NXMapFile opens the file, does a map_fd, then closes the file descriptor. When you map_fd memory, you need to vm_dealloc it when you're done with it. Otherwise, the page will remain busy. The question is, is this a bug in our docs or in NXClose? You can work around this problem by using this instead: NXCloseMemory(stream, NX_FREEBUFFER); This will cause the vm that was allocated by NXMapFile to be deallocated. The problem is that NX_FREEBUFFER isn't documented. (therefore, I can't guarantee that we support it... I'll ask around if this should be documented or not) You could argue that NXClose should use NX_FREEBUFFER on mapped files, but I'm not sure. Again, I'm not a streams expert, but I'll ask around. ] Workaround: End-users cannot workaround this problem for applications that ] are coded improperly. For programmers, instead of NXMapFile(), use this ] two-part construction: ] ] if ((fd = open(const char *name, int unixMode))) // O_RDONLY, etc. ] stream = NXOpenFile(int fd, int NeXTMode); // NX_READONLY, etc. ] ] and then at the bottom of your code, do both ] ] NXClose(stream); ] close(fd); The reason this works isn't because you're freeing the file descriptor here, but because the memory isn't mapped in the same way as it is with NXMapFile. Erik
From: cnayak@crl.com (Chiraprakash Nayak) Newsgroups: comp.sys.next.bugs Subject: Re: Dos file system and "undelete" Date: 31 May 1994 21:21:04 -0700 Organization: CRL Dialup Internet Access (415) 705-6060 [login: guest] Message-ID: <2sh2bg$ofi@crl.crl.com> References: <CqnFwK.BxK@cunews.carleton.ca> <2sfukf$eov@rosie.next.com> In article <2sfukf$eov@rosie.next.com>, Erik Kay <erikkay@next.com> wrote: >In article <CqnFwK.BxK@cunews.carleton.ca> c >Chiraprakash Nayak (cnayak@crl.com) wrote: >] : I will appreciate if some body can clarify. May be while dos partition >] : size limit is being fixed, NeXt will fix this anamoly too! >Actually, it won't. This is the first report we've had against this bug. >I'll see if I can get to it in time for 3.3. > Thanks for the info. What is the current release schedule for 3.3? In a related note, is floppy access (macintosh fs or otherwise) is being fixed too? I end up with the following - # mount -n /dev/fd0a /floppy Please insert Floppy Disk 0 in drive 0. mount: /dev/fd0a on /floppy: Block device required. mount: givibg up on: /floppy. # fd0: Device Block Size: 512 bytes fd0: Device Capacity: 1440 KB fd0: Disk Label: Disk [these messages comes back after the failure message] # mount ...... (trying again) [ no message, apparantly successful] #cp .... (works) # any command here invaribly reuslts in this message unexpected kernel trap d eip 114188 Failed Instruction Exception (2.d.0) Waiting for ....... [system hangs here. I ended up reinstalling from CD because of serius file system corruption) Same thing happens if I try to access mac floppy from workspace invariably after copying files to floppy - no matter howlong I wait after copying. Is it my drive or ..... Thanks a lot. Best Regards. --Chip (cnayak@crl.com)
Newsgroups: comp.sys.next.bugs,comp.sys.next.programmer From: Atze (Alexander Spohr) Subject: DBModule saveChanges: last record not deleted (Solution) Message-ID: <1994Jun1.094530.1114@dart.de> Sender: aspohr@dart.de Organization: d'ART Software GmbH Date: Wed, 1 Jun 94 09:45:30 GMT If you have a DBRecordList which isModified because you deleted the last record, DBModules saveChanges: will not save this DBRecordList. The DBFetchGroup which is responsible for the DBRecordList "forgets" to tell it to saveModifications. Here is a workaround which should be implemented by a Controller for the DBModule. However, the delegates of DBModule and DBFetchGroup are informed about the normal stuff inside [dbmodule saveChanges:self]. All special cases (deletion of last record) will only inform the delegate of the DBRecordList. Sorry for that. The docuentation for saveModifications is wrong. This is why [aRecordList saveModifications] tests for 0 and not 1 (as the doc says). Atze PS. If anybody modifies this method to make it behave more like DBModules saveChanges: please send me your version. I don't need it right know so I will not invest more time on this bug-fix. I hope NeXT will fix it in 3.3 :-) - saveChanges:sender { id ret = self; id database = [dbmodule database]; ret = [dbmodule saveChanges:self]; if(ret && [[dbmodule rootFetchGroup] hasUnsavedChanges]) { int i; BOOL hasTransaction; BOOL wasConnected = [database isConnected]; List *fetchGroups = [[List alloc] init]; id aRecordList = nil; [dbmodule getFetchGroups:fetchGroups]; if(!wasConnected) [database connect]; hasTransaction = [database beginTransaction]; for(i=0; i<[fetchGroups count]; i++) { aRecordList = [[fetchGroups objectAt:i] recordList]; if([aRecordList isModified]) if([aRecordList saveModifications]) { ret = nil; break; } } if(hasTransaction) { if(ret) [database commitTransaction]; else [database rollbackTransaction]; } if(!wasConnected) [database disconnect]; [fetchGroups free]; } return ret; } -- Alexander Spohr, Hamburg, Germany Alexander_Spohr@DART.de (Mail ) Faces and faces (+49) 40 / 380 23-0 (Sound) See them and complain not (+49) 40 / 380 23-290 (Fax ) And am content with all
From: wjs@yucca.omnigroup.com (William Shipley) Newsgroups: comp.sys.next.bugs Subject: Re: BUG: NXMapFile() doesn't release file descriptor on NXClose() Date: 1 Jun 1994 03:20:06 -0700 Organization: Omni Development, Inc. Message-ID: <2shncm$6j5@yucca.omnigroup.com> References: <1994May31.151144.16049@afs.com> <2sgi13$f2t@rosie.next.com> Erik Kay writes: >This will cause the vm that was allocated by NXMapFile to be deallocated. >The problem is that NX_FREEBUFFER isn't documented. (therefore, I can't >guarantee that we support it... I'll ask around if this should be >documented or not) It is documented, just not in all the right places. For example, the the concepts guide pretty much says what you said: _______________________________________ When you're finished with a memory stream, close it by calling NXCloseMemory(): NXCloseMemory(stream, NX_FREEBUFFER); Usually, you'll use NX_FREEBUFFER as the second argument to free all storage used by the stream, but there are two other constants that can be used. If you've used the stream for writing, more memory may have been made available than was actually used; the constant NX_TRUNCATEBUFFER indicates that any unused pages of memory should be freed. (Calling NXClose() with a memory stream is equivalent to calling NXCloseMemory() and specifying NX_TRUNCATEBUFFER.) NX_SAVEBUFFER doesn't free the memory that had been made available. ________________________________________ -Wil Shipley
Newsgroups: comp.sys.next.bugs From: vlahos@cubism.portal.com (Harry Vlahos) Subject: Re: Bugs, bugs, bugs (system panic) Message-ID: <1994Jun1.031208.11062@cubism> Sender: vlahos@cubism Organization: HV Software, Inc. References: <2s2gt3$t70@apollo.west.oic.com> Date: Wed, 1 Jun 1994 03:12:08 GMT In article <2s2gt3$t70@apollo.west.oic.com> dillon@apollo.west.oic.com (Matthew Dillon) writes: > In article <9405261445.AA08161@proxima.com> elitman@proxima.com (Eric A. Litman) writes: > :In article <rwyattCqDqH8.1yF@netcom.com> rwyatt@netcom.com writes: > : > :> Has anyone else experienced such glaring problems with NS/Intel? All of > :> my other apps work great. I really hate to think that NEXTSTEP is so > :> finicky that an app can crash one computer and not another??? Just > :> interested in your two cents... > : > :We have over 350 units installed throughout the country, and have > :experienced regular problems with kernel panics. More often than > :... > :I have put a hold on all Intel architecture purchases (any systems, > :not just those manufactured by Intel) in our organization in hopes that > :... > :So the question now becomes, which system component(s) is to blame for > :this? Is it the firmware of the SCSI controller? I would certainly > :... > :-- > :Eric Litman Proxima, Inc. vox: (703) 506.1661 > :Director, ProxNet McLean, VA elitman+@proxima.com > > Every once in a while we get people on the Linux groups with crashing > problems that nobody else has. It has turned out to be one (or a > combination of): > > * Incorrect BIOS level configuration (too few wait states for cache, main > memory VLBUS, or some idiot suping up the main bus speed for ISA > machines). > > * Cache rams that are out of spec (i.e. too slow) > > * Overheating CPUs that work just fine under DOS, but die in 32 bit > linear address mode. > > > I suggest, before you start blaming the software, that you check your > hardware: > > (1) Try increasing the number of wait states for the cache and for main > memory (write down the old configuration!) > > (2) A 486 DX-33 gets pretty hot without a heat sink, but still generally > works. A 486 DX/2-66 *REQUIRES* a heat sink. I strongly suggest > a high quality heat sink *without* a fan (which only adds additional > failure modes) for any high speed processor including a DX-33. > > It is amazing how many vendors, such as Micronics, think that their DX/2-66 > motherboards will work without a heat sink. Simply put, you are asking for > it if you do not have a heat sink. > > We also avoid 50MHz bus systems (DX-50) for several reasons. Not only do > many 50 MHz bus motherboard manufacturers not have a clue about how to > construct them, but most VL-Bus boards *cannot* take advantage of the > higher bus speed and either must be strapped for additional wait states > or not used at all. In many cases, main memory and/or cache must also > be strapped for more wait states which kinda kills the whole reason for > buying a 50 MHz system in the first place! The tolerances for 50MHz > systems are simply too tight for our tastes. The hassle is not worth > the incremental improvement. > > -Matt > > -- > > Matthew Dillon dillon@apollo.west.oic.com > 1005 Apollo Way ham: KC6LVW (no mail drop) > Incline Village, NV. 89451 Obvious Implementations Corporation > USA Sandel-Avery Engineering > [always include a portion of the original email in any response!] > I have a Dell 466L (one that is recommended by Next on their approved list of tested configurations), with 32MB of RAM, no weird stuff, just standard configuration. I get system panics just about every day (until I stopped using it and went back to my mono-NextCube). So much for color.... HV
Newsgroups: comp.sys.next.bugs From: cedman@princeton.edu (Carl Edman) Subject: NX_NUMERICPAD bug Message-ID: <CEDMAN.94Jun1092641@capitalist.princeton.edu> Originator: news@nimaster Sender: news@Princeton.EDU (USENET News System) Organization: Princeton University Date: Wed, 1 Jun 1994 13:26:41 GMT There is a bug in the NeXTstep handling of numeric pad keys (including the cursor keys). You can demonstrate this in any application which distinguishes between a keystroke on the numeric pad and on the main keypad, such as Stuart.app with the "Emulated keypad" option set or Emacs.app. Try pressing any numeric pad key and before releasing it press one of the keys on the main keypad. The application will act as if you had struck the corresponding key on the numeric keypad and not on the main keypad. For example under Stuart.app, pressing the cursor up key and before releasing it (as can easily happen why editing) pressing the '=' key on the main keypad generates not an '=' sign but an escape sequence as if you had hit the '=' on the keypad. I investigated this some more and now I not only understand that the bug is there, but also how it happens (and a way to work around it) and why it is difficult for NeXT to fix it. Let me explain how the NS key handling works (I don't think this is documented anywhere). Striking or releasing a key (any key including modifier keys) on any keyboard generates a key code. This key code is extremely hardware dependent. ADB keyboards generate different keycodes than ISO keyboard and different keycodes than PC style keyboards. To translate this key code into something hardware independent an application can use (i.e. an ASCII character code), NeXT uses the keyboard mappings you can find in /NextLibrary/Keyboards. A keyboard mapping consists out of a number of parts but for us only two are of interest. Modifier keys are defined in a list like this: modifier-type, list of keys of that type ... For example in a NeXT ISO US keyboard you have: 1 (modifier-type shift), 0x52 (NeXT ISO keyboard keycode for the left shift key), 0x57 (keycode for the right shift key) 2 (modifier-type control), 0x51 (keycode for the control key) 3 (modifier-type alternate) ... 4 ... 5 (modifier-type numericpad), 0x09 (keycode for the left cursor), ..., 0x11 (keycode for the numeric keypad 1), ... Note that all the 22 numeric keypad keys are considered by NS to be modifier keys just like the shift or the control key ! The other part of interest of the keymapping file is the part which assigns character codes to all the live keys. It looks (in essence) like this: keycode, unmodified character, character with shift modifier, character with control modifier, character with shift _and_ control modifier ... For example, again from the NeXT ISO US keyboard: 0x39 (keycode for the 'a' key), 0x61 ('a' in ASCII), 0x41 ('A' in ASCII), 0x01 (ctrl-'a' in ASCII), ... ... So when you strike a key on the keyboard this happens: 1. NS checks if it is in the list of modifier keys. If so, it sets the modifier flag which corresponds to the key. 2. NS checks if it is in the list of live keys. If so, it uses the entry in the live key list and the current modifier flags to calculate the ascii code and sends the event to the application. Almost all keys are either modifier keys _or_ live keys so generally only 1. _or_ 2. happens but not both. The one exception are the numeric pad keys. They set both the numeric pad modifier _and_ send a keycode. When you release a key this happens: 1. NS checks if it is a modifier key. If so, it clears the modifier flag which corresponds to the key. Now you understand (I hope), why this bug occurs. Striking e.g. the cursor up key is to NS just like hitting a kind of shift key. Until you release it will modify all other keys which are struck. Changing this would require NeXT to rewrite a significant part of their keyboard code and to create a new keymapping format. As long as only a few apps like Emacs or Stuart are struck by this bug this is unlikely to happen. So how do you help yourself ? Just never trust the NX_NUMERICPAD bit in the event flags. It doesn't mean that the current keystroke is from the numeric pad any more than the NX_SHIFT bit means that the key currently struck is the shift key. It just means that any numeric pad key is being held down at the time. If you really need to know if the current keystroke came from the numeric pad in a portable manner, you'll have to actually get the keymapping from NS and check the list of numeric pad modifier keycodes. Only if the current keycode is part of that list can you be sure that the user is hitting a numeric pad key right now. For example code on how to do this look in the next release of Emacs (4.04). Carl Edman
From: "Thomas G. Mon" <tm31+@andrew.cmu.edu> Newsgroups: comp.sys.next.bugs Subject: Re: BUG: NXMapFile() doesn't release file descriptor on NXClose() Date: Thu, 2 Jun 1994 02:12:16 -0400 Organization: Senior, Math/Computer Science, Carnegie Mellon, Pittsburgh, PA Message-ID: <QhvLV0200iV348HAhq@andrew.cmu.edu> In-Reply-To: <2shncm$6j5@yucca.omnigroup.com> erikkay@next.com (Erik Kay) writes: > Actually, this is not the problem. NXMapFile opens the file, does > a map_fd, then closes the file descriptor. When you map_fd memory, > you need to vm_dealloc it when you're done with it. Otherwise, > the page will remain busy. The question is, is this a bug in our > docs or in NXClose? You can work around this problem by using this > instead: > > NXCloseMemory(stream, NX_FREEBUFFER); > > This will cause the vm that was allocated by NXMapFile to be > deallocated. The problem is that NX_FREEBUFFER isn't documented. > (therefore, I can't guarantee that we support it... I'll ask around > if this should be documented or not) > > You could argue that NXClose should use NX_FREEBUFFER on mapped > files, but I'm not sure. Again, I'm not a streams expert, but I'll > ask around. > I did find NX_FREEBUFFER documented, in a way in the Pasteboard class General Reference. It uses: NXCloseMemory(theStream, NX_FREEBUFFER). I was wondering what the difference is between NX_FREEBUFFER and NX_TRUNCATEBUFFER (since NX_FREEBUFFER is marked as a Pre3.0_Concept) and just for FYI, from DL you get: "Usually, you'll use NX_FREEBUFFER as the second argument to free all storage used by the stream, but there are two other constants that can be used. If you've used the stream for writing, more memory may have been made available than was actually used; the constant NX_TRUNCATEBUFFER indicates that any unused pages of memory should be freed. (Calling NXClose() with a memory stream is equivalent to calling NXCloseMemory() and specifying NX_TRUNCATEBUFFER.)" So should I be using NX_TRUNCATEBUFFER or NX_FREEBUFFER if I plan to open another memory stream again with the same identifier All of the documentation points to NX_FREEBUFFER when working with pasteboards, but what about other applications? Tom. tm31@andrew.cmu.edu [NeXTmail welcome]
From: robert@steffi.demon.co.uk (Robert Nicholson) Newsgroups: comp.sys.next.bugs,comp.sys.next.programmer Subject: Why does NXBrowserCell's drawInside composite using NX_COPY? Date: 02 Jun 1994 13:56:59 GMT Organization: me organized? That's a joke! Distribution: world Message-ID: <ROBERT.94Jun2145659@steffi.demon.co.uk> Tests seem to indicate that NX_SOVER isn't used? No big deal since I'm writing a subclass anyway and override drawInside:inView: -- "Real programmers don't create classes. They build hierarchies" (PGP key: send email with Subject: request pgp key) (ASCII for text only messages)
From: robert@steffi.demon.co.uk (Robert Nicholson) Newsgroups: comp.sys.next.bugs,comp.sys.next.programmer Subject: Re: Why does NXBrowserCell's drawInside composite using NX_COPY? Date: 02 Jun 1994 14:25:53 GMT Organization: me organized? That's a joke! Distribution: world Message-ID: <ROBERT.94Jun2152553@steffi.demon.co.uk> References: <ROBERT.94Jun2145659@steffi.demon.co.uk> <robert@steffi.demon.co.uk> writes: >Tests seem to indicate that NX_SOVER isn't used? >No big deal since I'm writing a subclass anyway and override >drawInside:inView: OK, so that assumption was correct. given the following output from NXShowPS NXBrowserCell 0 48 48 48 12 execuserobject 0 61 1 composite My ImageBrowserCell 0 48 48 48 12 execuserobject 0 0 2 composite -- Is this a deliberate decision because of the non-alt image highlighting behaviour? -- Oh and just for a laugh.... where does NXBrowserCell keep the image it uses? Currently, I'm doing good things like using accessor methods in my drawInside but it would be nice to know where this pointer is. -- "Real programmers don't create classes. They build hierarchies" (PGP key: send email with Subject: request pgp key) (ASCII for text only messages)
From: Ward_Travis@transarc.com Newsgroups: comp.sys.next.bugs Subject: floopy eject and /etc/mtab Date: Thu, 2 Jun 1994 12:47:15 -0400 Organization: Carnegie Mellon, Pittsburgh, PA Message-ID: <whvUoH6SMUQGNdFasC@transarc.com> Can anyone tell me if NS 3.2/m68k fixes the NS3.0/NS3.1/m68k bug in which the ejection of a floppy from Workspace causes /etc/mtab to get chmod'ed to 600 ? I usually find out about this when 'df' fails due to not being able to read mtab . -- Ward C. Travis Pittsburgh PA USA "The fact is the sweetest dream that travis@transarc.com (412) 338 4388 labor knows." - R. Frost
From: madler@cco.caltech.edu (Mark Adler) Newsgroups: comp.sys.next.bugs Subject: cpp bug Date: 3 Jun 1994 16:29:22 GMT Organization: California Institute of Technology, Pasadena Message-ID: <2snlp2$p49@gap.cco.caltech.edu> /* Bug in NeXT compiler (cpp actually), NeXTstep 3.2, 68040. Symptom: this source causes cpp to go into an (apparently) infinite loop. You can see where it hangs using "cc -v". */ typedef unsigned long bits; /* presumed to be 32 bits */ #define BLOG 5 /* log base 2 of bits in type bits */ #define BLEN (1<<BLOG) /* bits to use in type bits */ #if ((BLEN>>3) != sizeof(bits)) # error BLOG #define incorrect #endif
#################################################################### From: fischer-michael@cs.yale.edu (Michael Fischer) Newsgroups: comp.sys.next.bugs Subject: Re: floopy eject and /etc/mtab Date: 4 Jun 1994 23:56:29 GMT Organization: Yale University Computer Science Dept., New Haven, CT 06520-8285 Message-ID: <2sr4bd$3f4@babyblue.cs.yale.edu> References: <whvUoH6SMUQGNdFasC@transarc.com> Ward_Travis@transarc.com wrote: : Can anyone tell me if NS 3.2/m68k fixes the NS3.0/NS3.1/m68k bug in : which the ejection of a floppy from Workspace causes /etc/mtab to get : chmod'ed to 600 ? I'm running NS 3.2/m68k and it doesn't happen to me. But then maybe more is needed to show the bug than simply slipping in a floppy and then ejecting it. -- ================================================== | Michael Fischer <fischer-michael@cs.yale.edu> | ==================================================
From: damonc@hookup.net (Damon F. Cooper) Newsgroups: comp.sys.next.bugs Subject: TranSys SLIP Connect Problem Date: Mon, 6 Jun 1994 01:44:20 GMT Organization: HookUp Communication Corporation, Oakville, Ontario, CANADA Message-ID: <damonc.2.2DF27F73@hookup.net> I have been able to install TransSys PNI SLIP/CSLIP, but once connected, not much happens#################################################################### Xref: informatik.uni-muenchen.de comp.sys.next.software:10943 comp.sys.next.misc:13565 comp.sys.next.bugs:1804 Path: informatik.uni-muenchen.de!lrz-muenchen.de!fauern!xlink.net!howland.reston.ans.net!vixen.cso.uiuc.edu!usenet From: t-shawki@uiuc.edu (Tarek G. Shawki) Newsgroups: comp.sys.next.software,comp.sys.next.misc,comp.sys.next.bugs Subject: Font Problems with DVIPS & TeXview Date: 6 Jun 1994 09:22:32 GMT Organization: University of Illinois at Urbana Lines: 44 Message-ID: <2supso$hr@vixen.cso.uiuc.edu> Reply-To: t-shawki@uiuc.edu NNTP-Posting-Host: ruger-11.slip.uiuc.edu Keywords: fonts, dvips, texview, postscript, tex I do have a strange problem related to the printing of postscript CM fonts from within TeXview. More specifically, I have installed postscript versions of the computer modern fonts. I have also edited the /usr/lib/tex/ps/psfonts.map and built the fonts. In fact, the fonts are all available from the font panel to all applications. Furthermore, the fonts work correctly with TeXview as far as screen display is concerned. Now, here is the problematic scenario: (1) I draw a sketch using Diagram 2.0. I then annotate the sketch with equations produced by writing the TeX code in an Edit window and then invoking the service "TeXview -> TeX EQ->.EPS". I copy the resulting equation and paste it in Diagram. The equation looks great in Diagram and it can be sized without loss of resolution (since it is now in postscript). (2) I save the annotated sketch from Diagram in an "eps" format. (3) I now use the "epsf" macros to include the eps graphic into my TeX document. (4) Once the TeX file is compiled, it shows up "correctly" in TeXview. All the fonts appear to be there. As soon as you print (from TeXview), the output looks fine everywhere except that the annotations within the eps graphic come out in COURIER!!!!! (5) The amazing thing is that all TeX equations and text use the postscript fonts. More strange is the fact that I could print the eps file itself from Preview without a problem!! Hence, the problem reduces to that of the loss of font information from the included eps graphics when processed by dvips. I would greatly appreciate any help or ideas. This problem does not happen on my Black NeXT when I process and print the same TeX file. Thanks for your help... _______________________________________________________________ Tarek G. Shawki Associate Professor 134 Talbot Laboratory Theoretical & Applied Mechanics Tel: 217 333 7131 University of Illinois Fax: 217 333 3625 104 South Wright Street E-mail: t-shawki@uiuc.edu Urbana, Illinois 61801
From: flexus!rfschtkt@maze.ruca.ua.ac.be (Raf Schietekat) Newsgroups: comp.sys.next.bugs Subject: Re: floopy eject and /etc/mtab Date: 6 Jun 1994 13:10:11 -0500 Organization: UTexas Mail-to-News Gateway Sender: nobody@cs.utexas.edu Message-ID: <9406051446.AA01605@flexus> LS, This bug is fixed in 3.2, as stated in the ReleaseNotes somewhere, and in KBNS. Always read KBNS! Raf Schietekat, RfSchtkt@maze.ruca.ua.ac.be, Flanders, Belgium *** The designation BANRUC60.BITNET for this host will disappear by July 1 *** (real, i.e., with triangle in the Deliver button) NeXTmail preferred Addressing limitations: no !, % or .uucp (? I should test this again) Compiler designers rarely do it. --- Aho, Sethi, Ullman, ``Compilers: Principles, Techniques, and Tools'', 2nd ed., p. 173
From: flexus!rfschtkt@maze.ruca.ua.ac.be (Raf Schietekat) Newsgroups: comp.sys.next.bugs Subject: Re: (If you don't support it then why use it?) Was: Re: BUG: NXMapFile() Date: 6 Jun 1994 13:33:56 -0500 Organization: UTexas Mail-to-News Gateway Sender: nobody@cs.utexas.edu Message-ID: <9406061829.AA01801@flexus> LS, For those of you getting confused (Robert Nicholson...): this thread is going on in two newsgroups at once, as started by Greg Anderson (should he have done that, rhetorical question?). I quoted from a message by Erik Kay (@next.com), sent only to comp.sys.next.bugs. Robert and Erik, please read each other's statements in the other newsgroup. To Tom Mon (on comp.sys.next.bugs): in this case (-[pasteboard readTypeToStream:]), do use NX_FREEBUFFER (not NX_TRUNCATEBUFFER), or what the NXCloseMemory documentation now suggests (NXGetMemoryBuffer and vm_deallocate). So how about an official statement that changing the NXCloseMemory documentation was just a mistake, NeXT, and that both methods will be supported, so we can close this discussion? Well, with regard to Tom's question, there's a similar problem with how -[pasteboard readType:data:length:] data should be released. This used to be with vm_deallocate, but now ``you should *not* attempt to free the returned memory using vm_deallocate() or free()''. I suppose it's OK for old code (even recompiled under 3.X) to continue to use vm_deallocate as long as it doesn't use the new readTypeToStream: as well? Such things should be documented (it's not in the ReleaseNotes), if only to reassure us that NeXT knows what it is doing, in this case. Raf Schietekat, RfSchtkt@maze.ruca.ua.ac.be, Flanders, Belgium *** The designation BANRUC60.BITNET for this host will disappear by July 1 *** (real, i.e., with triangle in the Deliver button) NeXTmail preferred Addressing limitations: no !, % or .uucp (? I should test this again) Compiler designers rarely do it. --- Aho, Sethi, Ullman, ``Compilers: Principles, Techniques, and Tools'', 2nd ed., p. 173
From: scossio@csupomona.edu (Seth Cossio) Newsgroups: comp.sys.next.bugs Subject: NEXT Video Resolution Problems Date: 6 Jun 94 11:39:28 PST Organization: California State Polytechnic University, Pomona Distribution: world Message-ID: <1994Jun6.113928.1@clstac> Summary: Can only get 640x480 BW. Keywords: NEXT VIDEO -- I just got around to installing version 3.2 of NEXT on an ISA machine. I have an AST BRAVO MT 486DX2/66 with 24megs RAM and a 345meg IDE HD with local bus ATI video(2megs RAM). NEXT does not let me use a video resolution higher than 640 x 480 in B&W! Anyone have any ideas why??? --- Seth Cossio Email: scossio@vmsa.is.csupomona.edu
From: szallies@wickie.informatik.uni-dortmund.de (Constantin Szallies (PG234)) Newsgroups: comp.sys.next.bugs Subject: TransSys PNI BUG -- PNI crashes whole system Date: 7 Jun 1994 15:49:00 GMT Organization: CS Department, Dortmund University, Germany Sender: szallies@wickie (Constantin Szallies (PG234)) Distribution: world Message-ID: <2t24tc$3sa@fbi-news.informatik.uni-dortmund.de> Hardware: Intel 486/DX2 VLBus/ISA, 16 MB, serial Mouse. Software: NS FIP 3.2. and TransSys PNI 1.7BETA I used /dev/cufa in the PNI configfile. --- Too bad, because the next time I booted NS, the login window didn't show up and only the spinning disk showed up. I restarted the computer in single user mode and removed the stuff in rc.local -> everything works fine now. I guess I should use /dev/cufb instead???? Is /dev/cufa taken by the serial mouse? "more /dev/cufa" results in "/dev/cufa: device busy". Any surgestions?? ////////////////////////////////////////////////////////////////////// / Constantin Szallies is szallies@kunibert.informatik.uni-dortmund.de/ / Simrockstr.50 40235 Duesseldorf Tel. 49/211/682316/ / (__) / / (oo) Don't have a cow, man! / / /-------\/ / / / | || / / * ||----|| Sorry for wasting your disk space / / ^^ ^^ / //////////////////////////////////////////////////////////////////////
From: estraff@next.com (Ethan Straffin) Newsgroups: comp.sys.next.bugs Subject: Re: NEXT Video Resolution Problems Date: 7 Jun 1994 17:46:21 GMT Organization: NeXT, Inc. Distribution: world Message-ID: <2t2bpd$74e@rosie.next.com> References: <1994Jun6.113928.1@clstac> In article <1994Jun6.113928.1@clstac> scossio@csupomona.edu (Seth Cossio) writes: :-- :I just got around to installing version 3.2 of NEXT on an ISA machine. I :have an AST BRAVO MT 486DX2/66 with 24megs RAM and a 345meg IDE HD with :local bus ATI video(2megs RAM). : :NEXT does not let me use a video resolution higher than 640 x 480 in B&W! : :Anyone have any ideas why??? Not without a bit more information. Our ATI driver supports the ATI Ultra Pro (VL-Bus or EISA) with the ATI 68800 controller. Is this the card you have, is it some other card, or is the video on-board? What messages appear during startup when you boot with the ATI Ultra Pro driver? Ethan Straffin NEXTSTEP Technical Support
From: damonc@hookup.net (Damon F. Cooper) Newsgroups: comp.sys.next.bugs Subject: TransSys PNI Key Date: Mon, 6 Jun 1994 01:52:16 GMT Organization: HookUp Communication Corporation, Oakville, Ontario, CANADA Message-ID: <damonc.11.2DF2814F@hookup.net> Anybody have that TransSys PNI SLIP authorization key extention that was posted by TransSys a couple weeks ago? (and instructions on how to install it?) Thanks! Damon Cooper damonc@hookup.net
From: sean@zapotec.math.byu.edu (Sean O. Luke) Newsgroups: comp.sys.next.bugs Subject: Edit saves with "Incomplete Last Line" Date: 8 Jun 1994 16:19:47 GMT Organization: Brigham Young University Message-ID: <2t4r33$l41@hamblin.math.byu.edu> I've discovered a fun little bug in Edit.app. If I save a list of words, say, 5 lines long, wc tells me that the file only contains 4 lines! And vi reports that the file has an "Incomplete Last Line", meaning it's missing, what, a \0 at the end of the file? If I write the same file in vi, wc reports it properly and vi understands how to deal with it properly. This is an inexcusable bug in Edit. :-( Sean Luke sean@digaudio.byu.edu
From: doug@foxtrot.ccmrc.ucsb.edu (Douglas Scott) Newsgroups: comp.sys.next.bugs Subject: Re: Edit saves with "Incomplete Last Line" Date: 8 Jun 1994 17:29:44 GMT Organization: University of California, Santa Barbara Distribution: world Message-ID: <2t4v69$27g@hub.ucsb.edu> References: <2t4r33$l41@hamblin.math.byu.edu> In article <2t4r33$l41@hamblin.math.byu.edu> sean@zapotec.math.byu.edu (Sean O. Luke) writes: ] ] I've discovered a fun little bug in Edit.app. If I save a list of words, ] say, 5 lines long, wc tells me that the file only contains 4 lines! And vi ] reports that the file has an "Incomplete Last Line", meaning it's missing, ] what, a \0 at the end of the file? ] ] If I write the same file in vi, wc reports it properly and vi understands ] how to deal with it properly. This is an inexcusable bug in Edit. :-( ] ] Sean Luke ] sean@digaudio.byu.edu Just hit <return> at the bottom of your textfile before saving it. -- Douglas Scott | Senior Development Engineer Tel: (805) 893-8352 | Center for Computer Music Research and Composition Internet (NeXTMail ok): | Music Department <doug@ccmrc.ucsb.edu> | University of California, Santa Barbara
From: steve@tarski.math.nd.edu(Steven Buechler) Newsgroups: comp.sys.next.bugs Subject: Default app, suffix and icon problem Date: Thu, 9 Jun 1994 16:25:22 GMT Organization: University of Notre Dame Message-ID: <940609112522.1169AABpE.steve@tarski> Mime-Version: 1.0 (Generated by Eloquent) Content-Type: text/plain; charset=US-ASCII I'm getting inconsistent behavior with the choice of icon used to represent a file with a specified suffix in the browser and the default app chosen to open those files. For example, files with the suffix .tex I set to open under Emacs.app. However, if I have TeXEdit (or TeXView) in the application dock, it displays the file in the browser with the icon corresponding to, say, TeXEdit. If I double-click on the file it opens properly in Emacs. I checked in the file ~/.NeXT/.suffixes3_1m68k.wmd and the app listed next to tex is TeXEdit. I thought this meant the file should open in TeXEdit, not just that the icon representing the file should be taken from TeXEdit -- is this right? By the way, if I take both TeXEdit and TeXView out of the application dock the icon representing .tex files will be the Emacs one. This problem is not specific to .tex files and the apps mentioned above. I can reset the default app for .ps files to Edit, but in ~/.NeXT/.suffixes3_1m68k.wmd the corresponding app is still listed as Preview and the icon representing these files is the Preview one. If I double-click on the .ps file, however, it opens properly in Edit. So, any ideas as to what's wrong and how I might fix it? I'm running NS3.2 on black hardware. (This problem is relatively new and I've been using 3.2 for quite awhile.) Thanks, Steve __________________________________________ Steven Buechler |"But I thought I Department of Mathematics |fixed that!" Mail Distribution Center | University of Notre Dame | Notre Dame, IN 46556 | (219)631-6233 Internet: steve@tarski.math.nd.edu NeXTmail and MIME accepted
From: lloyd@max.tiac.net (Christopher Lloyd - not the actor) Newsgroups: comp.sys.next.bugs Subject: Re: Edit saves with "Incomplete Last Line" Date: 9 Jun 1994 17:21:27 GMT Organization: The Internet Access Company Message-ID: <2t7j2n$bl1@sundog.tiac.net> References: <Cr3rIu.2xx@arcane.calgary.ab.ca> <2t7ee7$g9m@gap.cco.caltech.edu> <2t7io1$epg@hamblin.math.byu.edu> In article <2t7io1$epg@hamblin.math.byu.edu> sean@zapotec.math.byu.edu (Sean O. Luke) writes: >This may be all fine and dandy, everyone, but the fact remains that a great >DEAL of UNIX utilities besides vi (wc? awk? etc.) count lines by the number >of returns. It's a dumb design, granted, but it's part and parcel of the >operating system. I could care less about vi complaining about an >incomplete last line. But it's a very big deal if wc can't deal with it. > >So what's the middle ground? Hit return. -- |: Christopher Lloyd :|: Yrrid Incorporated :|: lloyd@yrrid.com :| |: Fish Cannot Carry Guns :|
From: Christopher_Lane@Med.Stanford.EDU Newsgroups: comp.sys.next.bugs Subject: Re: Edit saves with "Incomplete Last Line" Date: 9 Jun 1994 17:58:12 GMT Organization: Stanford University Message-ID: <2t7l7k$dpa@nntp2.Stanford.EDU> References: <Cr3wtM.5rz@unixhub.SLAC.Stanford.EDU> Bob Cook <bobcook@slac.stanford.edu> writes: >Sean O. Luke writes >> >> I've discovered a fun little bug in Edit.app. If I save a list of words, >> say, 5 lines long, wc tells me that the file only contains 4 lines! And vi >> reports that the file has an "Incomplete Last Line", >> > Well, I consider it a restriction of vi that it FORCES a <CR> (not a \0) at > the end of the file. Edit makes no such requirement, which I consider a > better design, not a "fun little bug". In early NeXT release, Edit also forced the file to end with a <CR>, later it was changed not to. It should have been a user preference with appending a <CR> to be the default for naive users as without it *a lot* of things break or work incorrectly. Since Edit's preferences were redesigned to accomodate both naive and sophisticated Unix users, it's surprising that this was left out. - Christopher
Newsgroups: comp.sys.next.bugs From: bobcook@slac.stanford.edu (Bob Cook) Subject: Re: Edit saves with "Incomplete Last Line" Message-ID: <Cr3wtM.5rz@unixhub.SLAC.Stanford.EDU> Sender: news@unixhub.SLAC.Stanford.EDU Organization: Stanford Linear Accelerator Center References: <2t4r33$l41@hamblin.math.byu.edu> Date: Thu, 9 Jun 1994 01:44:09 GMT Sean O. Luke writes > >I've discovered a fun little bug in Edit.app. If I save a list of words, >say, 5 lines long, wc tells me that the file only contains 4 lines! And vi >reports that the file has an "Incomplete Last Line", meaning it's missing, >what, a \0 at the end of the file? > >If I write the same file in vi, wc reports it properly and vi understands >how to deal with it properly. This is an inexcusable bug in Edit. :-( Well, I consider it a restriction of vi that it FORCES a <CR> (not a \0) at the end of the file. Edit makes no such requirement, which I consider a better design, not a "fun little bug". Note that emacs also does not force such a requirement. -- Bob Cook bobcook@slac.stanford.edu Stanford Linear Accelerator Center NeXT mail okay
From: sanguish@digifix.com Newsgroups: comp.sys.next.announce,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.bugs,comp.sys.next.programmer,comp.sys.next.marketplace Subject: Quickguide to NEXTSTEP Information available on the Internet Date: 9 Jun 1994 01:07:57 -0400 Organization: Next Announcements Message-ID: <2t683d$2h4@digifix.digifix.com> Topics include: Stepwise NEXTSTEP/OpenStep Information Server online comp.sys.next newsgroups related newsgroups comp.sys.next newsgroups mailing list ftp sites NeXTanswers New Information --------------- Stepwise NEXTSTEP/OpenStep Information Server online A product directory built around the World Wide Web system, this will allow full multimedia announcements by NEXTSTEP developers, as well as the ability to browse the available products for NEXTSTEP. This service is online now, and can be reached at http://www.stepwise.com/ or (if you nameserver is out of date) http://digifix.digifix.com/ it can be reached using OmniWeb (available from ftp.omnigroup.com) or Mosaic. The entries are coming in quite quickly, and currently consist of - NeXT Press Releases - OpenStep WhitePapers - Third Party Products Directory - NEXTSTEP Developer Directory - NEXTSTEP Community WhitePages - NEXTSTEP User Group Directory - comp.sys.next.announce archives (searchable) - searchable contents of Third Party compilation CDs Additionally the NEXTSTEP Product Information Mail Server will be stocked full file files in the next week... you can get information on using the mail server at ns-products@digifix.com Suggestions or comments can be directed to me at sanguish@digifix.com comp.sys.next.* newsgroups -------------------------- Comp.Sys.Next.Advocacy This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. Comp.Sys.Next.Announce Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) The NEXTSTEP FAQs are posted here monthly as well. This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Comp.Sys.Next.Bugs A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT- specific groups as well. Comp.Sys.Next.Hardware Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. Comp.Sys.Next.Marketplace NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. Comp.Sys.Next.Misc For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! Comp.Sys.Next.Programmer Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. The NEXTSTEP programmer FAQs are posted here. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. Comp.Sys.Next.Software This is a place to talk about [third party] software products that run on NEXTSTEP systems. Comp.Sys.Next.Sysadmin Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. related Newsgroups ------------------ Comp.Soft-Sys.Nextstep Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. Comp.Lang.Objective-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. Comp.Object Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original Comp.Sys.Next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News ------------------------------------------- Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep-d next-advocacy-d next-announce-d next-bugs-d next-hardware-d next-marketplace-d next-misc-d next-programmer-d next-software-d next-sysadmin-d (For a full description, send mail saying LISTS to <digestif@antigone.com>). The subscription syntax is essentially the same as LISTSERV's. To subscribe, send a message to <digestif@antigone.com> saying: SUB Listname YourName Example: SUB next-hardware-d John Doe The ftp sites ------------- cs.orst.edu: The main site for North American submissions nova.cc.purdue.edu: Lots of older stuff, but very short on disk space ftp.informatik.uni-muenchen.de: In Germany. terra.stack.urc.tue.nl (Dutch NEXTSTEP User Group) and cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp.next.com: See the below ftp.next.com and NextAnswers@next.com ------------------------------------- From the document 1000_Help from ftp.next.com Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, or you can transfer them by anonymous ftp. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to NeXTanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO <your-address> If you have any problem with the system or suggestions for improvement, please send mail to NeXTanswers-request@NeXT.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to NeXTanswers-request@NeXT.com. Written by: Eric P. Scott eps@toaster.SFSU.EDU and Scott Anguish sanguish@digifix.com Additions from: Greg Anderson (Greg_Anderson@afs.com) and Michael Pizolato (Michael_Pizolato@afs.com)
Newsgroups: comp.sys.next.bugs From: alex@cs.umd.edu (Alex Blakemore) Subject: Re: Edit saves with "Incomplete Last Line" Message-ID: <Cr41s8.26E@genoa.com> Sender: alex@genoa.com (Alex Blakemore) Organization: Genoa Software Systems References: <2t4v69$27g@hub.ucsb.edu> Date: Thu, 9 Jun 1994 03:31:20 GMT In article <2t4r33$l41@hamblin.math.byu.edu> sean@zapotec.math.byu.edu (Sean O. Luke) writes: > I've discovered a fun little bug in Edit.app. If I save a list of words, > say, 5 lines long, wc tells me that the file only contains 4 lines! And vi > reports that the file has an "Incomplete Last Line", meaning it's missing, > what, a \0 at the end of the file? that's not a bug, that's definitely a feature in Edit, and a bug in vi that it can't read such files (or files with long lines) its missing a '\n' to end the last line not a '\0' emacs will let you do the same thing as Edit. there are occasions when you don't want to have a newline at the end. emacs is a little nicer than Edit in this regard in that it has an _option_ to append a newline to the last line if you forget to type one. -- Alex Blakemore alex@cs.umd.edu NeXT mail accepted
From: albert@proffa.cc.tut.fi (Ojala Pasi) Newsgroups: comp.sys.next.bugs Subject: Question: How to avoid the 64 MB/DOS bug? Date: 9 Jun 1994 12:06:55 +0300 Organization: Tampere University of Technology Distribution: inet Message-ID: <2t6m3f$bu4@proffa.cc.tut.fi> This is a forwarded message from: Juha Tuominen I'm developing a program running on NSi which is suppose to collect a database on a DOS partition. The database will be then written to a CDR (writable CD) using a program running on Windows. The problem is that the size of the database might be as big as 700 MB and the DOS partition reserved for the Windows, programs, DOS and database is 1.85 GB. It works fine and everything can be read from the NeXTstep side as well, but calculating the free space doesn't work (that's the known bug in NS3.2). That causes that when the disk is half full, NeXTstep hangs when writing on the DOS partition. To avoid problems I was thinking if it would be possible to split the whole 1.85 GB into 64 MB pieces? I'm not familiar with NeXTstep's dos partition handling procedures and I don't know if it can mount more that one partition at the time. So if someone has splitted a large DOS disk into 64 MB pieces and can use all of them at the same time without problems, please mail a reply. I'm not very fond of formatting the whole block without knowing for sure... I'm also interested in knowing when the fixed version of NSi is coming out. This might be dreaming but could there be a bugfix file available somewhere to heal this -=*VERY ANNOYING*=- bug...? -Juha
From: flight@mathi.uni-heidelberg.de (Gregor Hoffleit) Newsgroups: comp.sys.next.bugs Subject: Re: Question: How to avoid the 64 MB/DOS bug? Date: 9 Jun 1994 09:41:04 GMT Organization: University of Heidelberg, Germany Distribution: inet Message-ID: <2t6o3g$eqr@sun0.urz.uni-heidelberg.de> References: <2t6m3f$bu4@proffa.cc.tut.fi> Ojala Pasi (albert@proffa.cc.tut.fi) wrote: : To avoid problems I was thinking if it would be possible to split the whole : 1.85 GB into 64 MB pieces? I'm not familiar with NeXTstep's dos partition It's a _32_ MB bug!! You shouldn't work with partitions with more than 32 MB, or else they could be corrupted! : handling procedures and I don't know if it can mount more that one : partition at the time. So if someone has splitted a large DOS disk into : 64 MB pieces and can use all of them at the same time without problems, please : mail a reply. I'm not very fond of formatting the whole block without knowing : for sure... No, according to Erik Kay from NeXT this is impossible. In fact, NEXTSTEP accesses a DOS partition as e.g. /dev/rsd0h where '0' is the drive and 'h' the partition number. Since there's only one 'hook' for a DOS partition on every drive, only one DOS partition can be mounted from every drive ;-(. : I'm also interested in knowing when the fixed version of NSi is coming out. : This might be dreaming but could there be a bugfix file available somewhere : to heal this -=*VERY ANNOYING*=- bug...? : -Juha According to Erik, a bug fix for this problem is already finished and should be on ftp.next.com in a couple of days (weeks ?). Gregor -- | Gregor Hoffleit admin MATHInet / contact HeidelNeXT | | MAIL: Mathematisches Institut PHONE: (49)6221 56-5771 | | INF 288, 69120 Heidelberg / Germany FAX: 56-3812 | | EMAIL: flight@mathi.uni-heidelberg.de (NeXTmail) |
Newsgroups: comp.sys.next.bugs From: mdw@sitar.jazz.att.com (Mark Wuest) Subject: Re: Edit saves with "Incomplete Last Line" In-Reply-To: alex@cs.umd.edu's message of Thu, 9 Jun 1994 03:31:20 GMT Message-ID: <MDW.94Jun9082554@laura.jazz.att.com> Sender: news@nntpa.cb.att.com (Netnews Administration) Organization: AT&T Bell Labs, Red Hill References: <2t4v69$27g@hub.ucsb.edu> <Cr41s8.26E@genoa.com> Date: Thu, 9 Jun 1994 12:25:54 GMT In article <Cr41s8.26E@genoa.com> alex@cs.umd.edu (Alex Blakemore) writes: >emacs is a little nicer than Edit in this regard in that >it has an _option_ to append a newline to the last line >if you forget to type one. Well, I generally avoid Politics and Religion, but since Alex brought it up... It's also nicer than vi in this regard. This does the trick in the One True Editor: (setq require-final-newline t) -- Mark Wuest | Just a .sig. No advertisements for AT&T mdw@sitar.jazz.att.com | or anyone else for that matter.
Newsgroups: comp.sys.next.bugs From: terry@arcane.calgary.ab.ca (Terry Wilcox) Subject: Re: Edit saves with "Incomplete Last Line" Message-ID: <Cr3rIu.2xx@arcane.calgary.ab.ca> Sender: terry@arcane.calgary.ab.ca (Terry Wilcox) Organization: Arcane Systems Ltd. References: <2t4r33$l41@hamblin.math.byu.edu> Date: Wed, 8 Jun 1994 23:49:42 GMT Sean O. Luke writes >I've discovered a fun little bug in Edit.app. If I save a list of >words, say, 5 lines long, wc tells me that the file only contains >4 lines! And vi reports that the file has an "Incomplete Last >Line", meaning it's missing, what, a \0 at the end of the file? The same problem occurs in emacs. There's no return at the end of the last line. Edit and emacs don't insert one for you. Terry Wilcox -- Terry Wilcox (terry@arcane.calgary.ab.ca) `Some praise at morning what they blame at night; But always think the last opinion right.' - Alexander Pope
From: madler@cco.caltech.edu (Mark Adler) Newsgroups: comp.sys.next.bugs Subject: Re: Edit saves with "Incomplete Last Line" Date: 9 Jun 1994 16:02:15 GMT Organization: California Institute of Technology, Pasadena Message-ID: <2t7ee7$g9m@gap.cco.caltech.edu> References: <2t4r33$l41@hamblin.math.byu.edu> <Cr3rIu.2xx@arcane.calgary.ab.ca> One more minor note on this religious thread: I have used Edit.app on binaries (!), which works great, because Edit doesn't muck with anything you don't explicitly change just becuase it ought to "expect" a text file. Like not appending a newline at the end when there isn't one. In case you're wondering why anyone would use Edit in this way, it's the easiest way to modify Icon and Screen to use different shared libraries to get them to run under 3.x. mark
From: lones@lones.mit.edu (Lones A. Smith) Newsgroups: comp.sys.next.bugs Subject: A new bug with DOS partitions on NS3.2 [Intel] with EMM386.EXE Date: 9 Jun 1994 16:01:21 GMT Organization: Massachvsetts Institvte of Technology Message-ID: <2t7ech$s9n@senator-bedfellow.MIT.EDU> The Next Boot Manager is causing some sort of DOS memory management problem. I installed NeXTStep 3.2 on a DOS machine and I left a 100MB DOS 6.0 partition. I was able to boot into the DOS partition fine bnut Windows 3.1 would not run in 386 enhanced mode. I then discovered that loading DOS 6.0's EMM386.EXE caused the computer to hang when I tried to boot into DOS. My config.sys was as follows DEVICE=C:\DOS\HIMEM.SYS DEVICE=C:\DOS\EMM386.EXE DOS=HIGH,UMB FILESS=30 BUFFERS=20 and my autoexec.bat was .. @ECHO OFF PATH=C:\WINDOWS;C:\DOS DOSKEY To try and isolate the problem I booted to DOS from a DOS 6.0 floppy with the exact same config.sys and autoexec.bat. This fixed all the problems and windows was then up and running. There is clearly some sort of interference from the boot manager that causes memory manger's and related software to crash/hang. My hardware (all NeXT-approved) was as follows.. Intel DX/4 w/ VLB 512k Motherboard 32 MB RAM 1gig SCSI-2 Seagate drive Bustek SCSI-2 VLB controller DIamond Viper VLB 2 MB video 21" NEC Color Monitor I eventually had to give up on the whole DOS partition. I do hope this is fixed at least in 3.3, and a bug fix earlier would be MUCH appreciated. Sincerely, Lones A. Smith Department of Economics, MIT E52-252C, Cambridge MA 02139 voice: (617) 253-0914 [fax: (617) 253-6915] e-mail: lones@lones.mit.edu (NeXT mail okay)
From: lones@lones.mit.edu (Lones A. Smith) Newsgroups: comp.sys.next.bugs Subject: Data transfer with ISDN phone lines Date: 9 Jun 1994 16:27:57 GMT Organization: Massachvsetts Institvte of Technology Message-ID: <2t7fud$sse@senator-bedfellow.MIT.EDU> Hi, I have been unsuccessful at securing anything more than 2400 baud connection to the ISDN phone line in my phone --- even though MIT has given me a 9600 limit and my phone is set to 9600. Cruucially, non NeXT machines around me don't have this problem. The relevant lines from my /etc/ttys are #ttya "/usr/etc/getty std.9600" unknown off secure ttya "/usr/etc/getty std.9600" unknown off secure ttyb "/usr/etc/getty std.9600" unknown off secure ttyfa "/usr/etc/getty std.9600" unknown off secure ttyfb "/usr/etc/getty std.9600" unknown off secure ttyda "/usr/etc/getty D9600" unknown off ttydb "/usr/etc/getty D9600" unknown off ttydfa "/usr/etc/getty std.9600" unknown on secure ttydfb "/usr/etc/getty D9600" unknown off Any help is much appreciated, Lones A. Smith Department of Economics, MIT E52-252C, Cambridge MA 02139 voice: (617) 253-0914 [fax: (617) 253-6915]
From: sean@zapotec.math.byu.edu (Sean O. Luke) Newsgroups: comp.sys.next.bugs Subject: Re: Edit saves with "Incomplete Last Line" Date: 9 Jun 1994 17:15:45 GMT Organization: Brigham Young University Message-ID: <2t7io1$epg@hamblin.math.byu.edu> References: <2t4r33$l41@hamblin.math.byu.edu> <Cr3rIu.2xx@arcane.calgary.ab.ca> <2t7ee7$g9m@gap.cco.caltech.edu> This may be all fine and dandy, everyone, but the fact remains that a great DEAL of UNIX utilities besides vi (wc? awk? etc.) count lines by the number of returns. It's a dumb design, granted, but it's part and parcel of the operating system. I could care less about vi complaining about an incomplete last line. But it's a very big deal if wc can't deal with it. So what's the middle ground? +--------------------------------------------------------------+ | Sean Luke This signature no verb | | sean@digaudio.byu.edu ,,, <- finger for PGP key | | sean@zapotec.math.byu.edu (o o) | +------------------------oOO--(_)--OOo-------------------------+
From: Joe_Keenan@next.com (Joe Keenan) Newsgroups: comp.sys.next.bugs Subject: Re: A new bug with DOS partitions on NS3.2 [Intel] with EMM386.EXE Date: 9 Jun 1994 17:11:49 GMT Organization: NeXT, Inc. Message-ID: <2t7igl$a06@rosie.next.com> References: <2t7ech$s9n@senator-bedfellow.MIT.EDU> In article <2t7ech$s9n@senator-bedfellow.MIT.EDU> lones@lones.mit.edu (Lones A. Smith) writes: > The Next Boot Manager is causing some sort of DOS memory > management problem. That's incredibly unlikely. > I installed NeXTStep 3.2 on a DOS machine and I left > a 100MB DOS 6.0 partition. I was able to boot into the DOS > partition fine bnut Windows 3.1 would not run in 386 enhanced > mode. I then discovered that loading DOS 6.0's EMM386.EXE > caused the computer to hang when I tried to boot into DOS. When you say you "left a 100MB DOS partition", do you mean that there was originally a larger partition there? If so, did you reformat & reinstall on the DOS partition after you installed NS? If you didn't all bets are off on the usability of the partition. The easy way to check is to boot DOS and see if the size of the partition that FDISK reports is the same as what CHKDSK reports. If they don't, the partition is no good. joe
Newsgroups: comp.sys.next.bugs From: yong@iitmax.iit.edu (Yong Yoo) Subject: HELP:C compiler Message-ID: <1994Jun19.003014.8508@iitmax.iit.edu> Organization: Illinois Institute of Technology, Chicago Date: Sun, 19 Jun 94 00:30:14 GMT Hi, I've just upgrade my system from 486 to Pentium. Since Only thing I've changed on my system is MB and CPU, I though what ever I had on my drive will be ok and don't have to worry about it. But I was having some problems with c compiler and decided to reinstall the Developments Tools and Libraries. But when it reached the end of the each installation, it bombs and says error while trying to install. Do anyone knows whats going on? Am I suppose to reinall the NeXTStep when I upgraded the system? And response will be appreciated. thanks in advance Yong yong@iitmax.acc.iit.edu
From: jehu@jehu.async.vt.edu (john stanhope) Newsgroups: comp.sys.next.bugs Subject: Re: STANDALONE TRANSSYS SLIP POSSIBLE? Date: 18 Jun 1994 00:12:39 GMT Organization: Virginia Tech, Blacksburg, Virginia Distribution: world Message-ID: <2tte5n$hgn@solaris.cc.vt.edu> References: <2tt2k0$m6k@tar.wft.stack.urc.tue.nl> In article <2tt2k0$m6k@tar.wft.stack.urc.tue.nl> michael@wft.stack.urc.tue.nl (Michael Brouwer) writes: > In article <CrGrEt.48M@demon.co.uk>, > Stephen Rhodes <stephen@rhodes.demon.co.uk> wrote: > >In article <damonc.16.2DFB68FF@hookup.net> damonc@hookup.net (Damon F. Cooper) > >writes: > >[...stuff deleted...] To change the host name you are forced to hack the > >hostconfig file directly replacing -AUTOMATIC- by ypur selected host name - > >until this is done sendmail gets collisions with the local mail server. You > >also need to set up sendmail as a mail server > > Do not forget to add the hostname and IP address of your standalone machine > to NetInfo using Hostmanager.app if you edit /etc/hostconfig. If you do not > do this, the gethostbyname() call will fail because your hosts name is not > known to NetInfo thus breaking various programs. ---One of which is GNU emacs. > > Michael If you enter a host address without being on a network during boot I believe the system will hang, at least that is what mine did. When I set up PNI all I did was change my host name. John PS. Don't bother trying to email me, are school only lets you run slip for a short period before disconnecting so most likely your mail will bounce.
Newsgroups: comp.sys.next.bugs From: Jacques Garbi Subject: DOS patch didn't fix everything Message-ID: <CrI3tr.154@touga.vd.alphanet.ch> Sender: jacques@touga.vd.alphanet.ch (Jacques Garbi) Organization: Touga Management SA References: <corbyCrG7EG.CLJ@netcom.com> Date: Thu, 16 Jun 1994 17:41:51 GMT I'm still unable to format my SyQuest in DOS format !! Can anyone help ? Thanks Jacques -- ------------------------------------- Jacques GARBI Av. Davel 18 1004 Lausanne
From: lones@lones.mit.edu (Lones A. Smith) Newsgroups: comp.sys.next.bugs Subject: Re: A new bug with DOS partitions on NS3.2 [Intel] with EMM386.EXE Date: 10 Jun 1994 21:12:26 GMT Organization: Massachvsetts Institvte of Technology Message-ID: <2takvq$353@senator-bedfellow.MIT.EDU> References: <2t7igl$a06@rosie.next.com> In article <2t7igl$a06@rosie.next.com> Joe_Keenan@next.com (Joe Keenan) writes: > In article <2t7ech$s9n@senator-bedfellow.MIT.EDU> lones@lones.mit.edu (Lones A. > Smith) writes: > > The Next Boot Manager is causing some sort of DOS memory > > management problem. > > That's incredibly unlikely. > > > I installed NeXTStep 3.2 on a DOS machine and I left > > a 100MB DOS 6.0 partition. I was able to boot into the DOS > > partition fine bnut Windows 3.1 would not run in 386 enhanced > > mode. I then discovered that loading DOS 6.0's EMM386.EXE > > caused the computer to hang when I tried to boot into DOS. > > When you say you "left a 100MB DOS partition", do you mean that there was > originally a larger partition there? If so, did you reformat & reinstall on > the DOS partition after you installed NS? If you didn't all bets are off on > the usability of the partition. > > The easy way to check is to boot DOS and see if the size of the partition that > FDISK reports is the same as what CHKDSK reports. If they don't, the partition > is no good. > > joe > I find this response VERY annoying. The install was done from a raw disk thru the NeXTStep Intall process. In fact, the NeXTStep help over the phone were at a loss to explain what was happening, other than that we had found a new bug. I am posting this bug so that NeXT will take it seriously and FIX IT for 3.3 --- for me and for others. It took me time to do this, and I do hope that NeXT will give it the time it deserves. It has been unbelievable how many hurdles had to be surpassed to install 3.2 on entirely NeXT-approved white hardware. Everything required a special bug fix, and frequently the better hardware features of the NeXT-sanctioned hardware were simply not supported. (eg. the hard drive controller with a 10MG/sec limit could only be set to 5, etc. etc.). It thus is frustrating that (this particular) NeXT (support help) seeks to discount bug reports and blame it on the consumer. Sincerely, Lones A. Smith Department of Economics, MIT E52-252C, Cambridge MA 02139 voice: (617) 253-0914 [fax: (617) 253-6915] -- NewsGrazer, a NeXTstep(tm) news reader, posting -- M>UQR=&8P7&%N<VE[7&9O;G1T8FQ<9C!<9G-W:7-S($AE;'9E=&EC83M]"EQM M87)G;#$R,`I<;6%R9W(Q,C`*7'!A<F1<='@W,3%<='@Q-#(S7'1X,C$S-5QT M>#(X-#9<='@S-34X7'1X-#(W,%QT>#0Y.#%<='@U-CDS7'1X-C0P-5QT>#<Q M,39<9C!<8C!<:3!<=6QN;VYE7&9S,S)<9F,P7&-F,"!);B!A<G1I8VQE(#PR M=#=I9VPD83`V0')O<VEE+FYE>'0N8V]M/B!*;V5?2V5E;F%N0&YE>'0N8V]M M("A*;V4@2V5E;F%N*2!W<FET97,Z7`H^($EN(&%R=&EC;&4@/#)T-V5C:"1S M.6Y`<V5N871O<BUB961F96QL;W<N34E4+D5$53X@;&]N97-`;&]N97,N;6ET M+F5D=2`H3&]N97,@02X@(%P*/B!3;6ET:"D@=W)I=&5S.EP*/B`^(`E4:&4@ M3F5X="!";V]T($UA;F%G97(@:7,@8V%U<VEN9R!S;VUE('-O<G0@;V8@1$]3 M(&UE;6]R>5P*/B`^(&UA;F%G96UE;G0@<')O8FQE;2X@7`H^(%P*/B!4:&%T M)W,@:6YC<F5D:6)L>2!U;FQI:V5L>2Y<"CX@7`H^(#X@"0E)(&EN<W1A;&QE M9"!.95A44W1E<"`S+C(@;VX@82!$3U,@;6%C:&EN92!A;F0@22!L969T7`H^ M(#X@"6$@,3`P34(@1$]3(#8N,"!P87)T:71I;VXN($D@=V%S(&%B;&4@=&\@ M8F]O="!I;G1O('1H92!$3U-<"CX@/B`)<&%R=&ET:6]N(&9I;F4@8FYU="!7 M:6YD;W=S(#,N,2!W;W5L9"!N;W0@<G5N(&EN(#,X-B!E;FAA;F-E9%P*/B`^ M(`EM;V1E+B!)('1H96X@9&ES8V]V97)E9"!T:&%T(&QO861I;F<@1$]3(#8N M,"=S($5-33,X-BY%6$5<"CX@/B`)8V%U<V5D('1H92!C;VUP=71E<B!T;R!H M86YG('=H96X@22!T<FEE9"!T;R!B;V]T(&EN=&\@1$]3+EP*/B!<"CX@5VAE M;B!Y;W4@<V%Y('EO=2`B;&5F="!A(#$P,$U"($1/4R!P87)T:71I;VXB+"!D M;R!Y;W4@;65A;B!T:&%T('1H97)E('=A<R`@7`H^(&]R:6=I;F%L;'D@82!L M87)G97(@<&%R=&ET:6]N('1H97)E/R`@268@<V\L(&1I9"!Y;W4@<F5F;W)M M870@)B!R96EN<W1A;&P@;VX@(%P*/B!T:&4@1$]3('!A<G1I=&EO;B!A9G1E M<B!Y;W4@:6YS=&%L;&5D($Y3/R`@268@>6]U(&1I9&XG="!A;&P@8F5T<R!A M<F4@;V9F(&]N("!<"CX@=&AE('5S86)I;&ET>2!O9B!T:&4@<&%R=&ET:6]N M+EP*/B!<"CX@5&AE(&5A<WD@=V%Y('1O(&-H96-K(&ES('1O(&)O;W0@1$]3 M(&%N9"!S964@:68@=&AE('-I>F4@;V8@=&AE('!A<G1I=&EO;B!T:&%T("!< M"CX@1D1)4TL@<F5P;W)T<R!I<R!T:&4@<V%M92!A<R!W:&%T($-(2T132R!R M97!O<G1S+B`@268@=&AE>2!D;VXG="P@=&AE('!A<G1I=&EO;B`@7`H^(&ES M(&YO(&=O;V0N7`H^(%P*/B!J;V5<"CX@7`I<"@I<<&%R9%QT>#<P,%QT>#$T M,C!<='@R,3(P7'1X,C@T,%QT>#,U-#!<='@T,C8P7'1X-#DX,%QT>#4V.#!< M='@V-#`P7'1X-S$P,%QF8S!<8V8P(%P*22!F:6YD('1H:7,@<F5S<&]N<V4@ M5D5262!A;FYO>6EN9RX@"EQP87)D7'1X-S$Q7'1X,30R,UQT>#(Q,S5<='@R M.#0V7'1X,S4U.%QT>#0R-S!<='@T.3@Q7'1X-38Y,UQT>#8T,#5<='@W,3$V M7&9C,%QC9C`@7`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`I)="!H87,@8F5E;B!U;F)E;&EE=F%B;&4@:&]W(&UA;GD@:'5R9&QE M<R!H860@=&\@8F4@<W5R<&%S<V5D('1O(&EN<W1A;&P@,RXR(&]N(&5N=&ER M96QY($YE6%0M87!P<F]V960@=VAI=&4@:&%R9'=A<F4N($5V97)Y=&AI;F<@ M<F5Q=6ER960@82!S<&5C:6%L(&)U9R!F:7@L(&%N9"!F<F5Q=65N=&QY('1H M92!B971T97(@:&%R9'=A<F4@9F5A='5R97,@;V8@=&AE($YE6%0M<V%N8W1I M;VYE9"!H87)D=V%R92!W97)E('-I;7!L>2!N;W0@<W5P<&]R=&5D+B`H96<N M('1H92!H87)D(&1R:79E(&-O;G1R;VQL97(@=VET:"!A(#$P34<O<V5C(&QI M;6ET(&-O=6QD(&]N;'D@8F4@<V5T('1O(#4L(&5T8RX@971C+BDN7`I<"DET M('1H=7,@:7,@9G)U<W1R871I;F<@=&AA="`H=&AI<R!P87)T:6-U;&%R*2!. M95A4("AS=7!P;W)T(&AE;'`I('-E96MS('1O(&1I<V-O=6YT(&)U9R!R97!O M<G1S(&%N9"!B;&%M92!I="!O;B!T:&4@8V]N<W5M97(N"EP*7`H*4VEN8V5R M96QY+%P*7`I,;VYE<R!!+B!3;6ET:%P*1&5P87)T;65N="!O9B!%8V]N;VUI M8W,L($U)5"!<"D4U,BTR-3)#+"!#86UB<FED9V4@34$@,#(Q,SE<"G9O:6-E L.B`H-C$W*2`R-3,M,#DQ-"`@6V9A>#H@*#8Q-RD@,C4S+38Y,35=7`H*?0IE `
From: armes@pioneer.tds.com (Jim Armes) Newsgroups: comp.sys.next.bugs Subject: DOS filessystem bug fixes Date: 11 Jun 1994 18:42:26 GMT Organization: Trident Data Systems Distribution: world Message-ID: <2td0ii$77p@discovery.ectds.com> Keywords: DOS fs next I got the DOS fs patch last nite, it actually works! (not to be facetious or anything.) I mounted my second drive & secondary DOS partition (albiet a bit of terminal typing later). I was able to read & write with no problems so far. Get it!! Jim -- #--------------------------------------------------------# # Jim Armes | I've got a Pentium and you do not # # Matrix-Man | nahh nahh nah nah nahhhh... # # | # #Trident Data Systems| but IiiiiM a lumberjack and I'm ok# # armes@tds.com |I sleep all night & I work all day.#
From: sanguish@digifix.com Newsgroups: comp.sys.next.announce,comp.sys.next.misc,comp.sys.next.sysadmin,comp.sys.next.advocacy,comp.sys.next.hardware,comp.sys.next.software,comp.sys.next.bugs,comp.sys.next.programmer,comp.sys.next.marketplace Subject: Quick guide to NEXTSTEP Information on the Internet Date: 13 Jun 1994 01:49:51 -0400 Organization: Next Announcements Message-ID: <2tgs1v$qbe@digifix.digifix.com> Topics include: Stepwise NEXTSTEP/OpenStep Information Server online comp.sys.next newsgroups related newsgroups comp.sys.next newsgroups mailing list ftp sites NeXTanswers New Information --------------- Stepwise NEXTSTEP/OpenStep Information Server online A product directory built around the World Wide Web system, this will allow full multimedia announcements by NEXTSTEP developers, as well as the ability to browse the available products for NEXTSTEP. This service is online now, and can be reached at http://digifix.digifix.com/ it can be reached using OmniWeb (available from ftp.omnigroup.com) or Mosaic. The entries are coming in quite quickly, and currently consist of - NeXT Press Releases - OpenStep WhitePapers - Third Party Products Directory - NEXTSTEP Developer Directory - NEXTSTEP Community WhitePages - NEXTSTEP User Group Directory - comp.sys.next.announce archives (searchable) - searchable contents of Third Party compilation CDs Additionally the NEXTSTEP Product Information Mail Server will be stocked full file files in the next week... you can get information on using the mail server at ns-products@digifix.com Suggestions or comments can be directed to me at sanguish@digifix.com comp.sys.next.* newsgroups -------------------------- Comp.Sys.Next.Advocacy This is the "why NEXTSTEP is better (or worse) than anything else in the known universe" forum. It was created specifically to divert lengthy flame wars from .misc. Comp.Sys.Next.Announce Announcements of general interest to the NeXT community (new products, FTP submissions, user group meetings, commercial announcements etc.) The NEXTSTEP FAQs are posted here monthly as well. This is a moderated newsgroup, meaning that you can't post to it directly. Submissions should be e-mailed to next-announce@digifix.com where the moderator (Scott Anguish) will screen them for suitability. Comp.Sys.Next.Bugs A place to report verifiable bugs in NeXT-supplied software. Material e-mailed to Bug_NeXT@NeXT.COM is not published, so this is a place for the net community find out about problems when they're discovered. This newsgroup has a very poor signal/noise ratio--all too often bozos post stuff here that really belongs someplace else. It rarely makes sense to crosspost between this and other c.s.n.* newsgroups, but individual reports may be germane to certain non-NeXT- specific groups as well. Comp.Sys.Next.Hardware Discussions about NeXT-label hardware and compatible peripherals, and non-NeXT-produced hardware (e.g. Intel) that is compatible with NEXTSTEP. In most cases, questions about Intel hardware are better asked in comp.sys.ibm.pc.hardware. Questions about SCSI devices belong in comp.periphs.scsi. This isn't the place to buy or sell used NeXTs--that's what .marketplace is for. Comp.Sys.Next.Marketplace NeXT stuff for sale/wanted. Material posted here must not be crossposted to any other c.s.n.* newsgroup, but may be crossposted to misc.forsale.computers.workstation or appropriate regional newsgroups. Comp.Sys.Next.Misc For stuff that doesn't fit anywhere else. Anything you post here by definition doesn't belong anywhere else in c.s.n.*--i.e. no crossposting!!! Comp.Sys.Next.Programmer Questions and discussions of interest to NEXTSTEP programmers. This is primarily a forum for advanced technical material. The NEXTSTEP programmer FAQs are posted here. Generic UNIX questions belong elsewhere (comp.unix.questions), although specific questions about NeXT's implementation or porting issues are appropriate here. Note that there are several other more "horizontal" newsgroups (comp.lang.objective-c, comp.lang.postscript, comp.os.mach, comp.protocols.tcp-ip, etc.) that may also be of interest. Comp.Sys.Next.Software This is a place to talk about [third party] software products that run on NEXTSTEP systems. Comp.Sys.Next.Sysadmin Stuff relating to NeXT system administration issues; in rare cases this will spill over into .programmer or .software. related Newsgroups ------------------ Comp.Soft-Sys.Nextstep Like comp.sys.next.software and comp.sys.next.misc combined. Exists because NeXT is a software-only company now, and comp.soft-sys is for discussion of software systems with scope similar to NEXTSTEP. Comp.Lang.Objective-C Technical talk about the Objective-C language. Implemetations discussed include NeXT, Gnu, Stepstone, etc. Comp.Object Technical talk about OOP in general. Lots of C++ discussion, but NeXT and Objective-C get quite a bit of attention. At times gets almost philosophical about objects, but then again OOP allows one to be a programmer/philosopher. (The original Comp.Sys.Next no longer exists--do not attempt to post to it.) Exception to the crossposting restrictions: announcements of usenet RFDs or CFVs, when made by the news.announce.newgroups moderator, may be simultaneously crossposted to all c.s.n.* newsgroups. Getting the Newsgroups without getting News ------------------------------------------- Thanks to Michael Ross at antigone.com, the main NEXTSTEP groups are now available as a mailing list digest as well. next-nextstep-d next-advocacy-d next-announce-d next-bugs-d next-hardware-d next-marketplace-d next-misc-d next-programmer-d next-software-d next-sysadmin-d (For a full description, send mail saying LISTS to <digestif@antigone.com>). The subscription syntax is essentially the same as LISTSERV's. To subscribe, send a message to <digestif@antigone.com> saying: SUB Listname YourName Example: SUB next-hardware-d John Doe The ftp sites ------------- cs.orst.edu: The main site for North American submissions nova.cc.purdue.edu: Lots of older stuff, but very short on disk space ftp.informatik.uni-muenchen.de: In Germany. terra.stack.urc.tue.nl (Dutch NEXTSTEP User Group) and cube.sm.dsi.unimi.it (Italian NEXTSTEP User Group) ftp.next.com: See the below ftp.next.com and NextAnswers@next.com ------------------------------------- From the document 1000_Help from ftp.next.com Welcome to the NeXTanswers information retrieval system! This system allows you to request online technical documents, drivers, and other software, which are then sent to you automatically. You can request documents by fax or Internet electronic mail, or you can transfer them by anonymous ftp. NeXTanswers is an automated retrieval system. Requests sent to it are answered electronically, and are not read or handled by a human being. NeXTanswers does not answer your questions or forward your requests. USING NEXTANSWERS BY E-MAIL To use NeXTanswers by Internet e-mail, send requests to NeXTanswers@next.com. Files are sent as NeXTmail attachments by default; you can request they be sent as ASCII text files instead. To request a file, include that file's ID number in the Subject line or the body of the message. You can request several files in a single message. You can also include commands in the Subject line or the body of the message. These commands affect the way that files you request are sent: ASCII causes the requested files to be sent as ASCII text SPLIT splits large files into 95KB chunks, using the MIME Message/Partial specification These commands return information about the NeXTanswers system: HELP returns this help file INDEX returns the list of all available files INDEX BY DATE returns the list of files, sorted newest to oldest SEARCH keywords lists all files that contain all the keywords you list (ignoring capitalization) For example, a message with the following Subject line requests three files: Subject: 2101 2234 1109 A message with this body requests the same three files be sent as ASCII text files: 2101 2234 1109 ascii This message requests two lists of files, one for each search: Subject: SEARCH Dell SCSI SEARCH NetInfo domain NeXTanswers will reply to the address in your From: line. To use a different address either set your Reply-To: line, or use the NeXTanswers command REPLY-TO <your-address> If you have any problem with the system or suggestions for improvement, please send mail to NeXTanswers-request@NeXT.com. USING NEXTANSWERS BY FAX To use NeXTanswers by fax, call (415) 780-3990 from a touch-tone phone and follow the instructions. You'll be asked for your fax number, a number to identify your fax (like your phone extension or office number), and the ID numbers of the files you want. You can also request a list of available files. When you finish entering the file numbers, end the call and the files will be faxed to you. If you have problems using this fax system, please call Technical Support at 1-800-848-6398. You cannot use the fax system outside the U.S & Canada. USING NEXTANSWERS BY ANONYMOUS FTP To use NeXTanswers by Internet anonymous FTP, connect to FTP.NEXT.COM and read the help file pub/NeXTanswers/README. If you have problems using this, please send mail to NeXTanswers-request@NeXT.com. Written by: Eric P. Scott eps@toaster.SFSU.EDU and Scott Anguish sanguish@digifix.com Additions from: Greg Anderson (Greg_Anderson@afs.com) and Michael Pizolato (Michael_Pizolato@afs.com)
From: mow@marsu.tynet.sub.org (Markus Wenzel) Newsgroups: comp.sys.next.bugs Subject: Re: A new bug with DOS partitions on NS3.2 [Intel] with EMM386.EXE Date: 12 Jun 1994 21:24:13 +0100 Organization: Palumbian Research Labs Message-ID: <2tfqtd$fk@marsu.tynet.sub.org> References: <2t7igl$a06@rosie.next.com> <2takvq$353@senator-bedfellow.MIT.EDU> <2tc4p2$5om@news.CCIT.Arizona.EDU> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit mengpao@helium.Gas.UUG.Arizona.EDU (mengpao cheng) writes: > 1. If we set 1742 in standard mode which is compatible with > 1542, can 1742 still take advantage of 32bit EISA board. Or > said can 1742 still has the ability to access the memory more > than 16MB??? I must admit I don't know for sure, but I suppose that the 1742 acts exactly as a 1542 when set to standard mode. This implies all 1542/ISA restrictions. > 2. Supposely one has EISA motherboard and want to choose > either 1742 or 1542 controller. Since 1542 and 1742 are both > using the same driver in NS, why >would people pay more for using 1742 instead of using 1542 -- >assuming both of them make no difference to each others. Normally, people don't have NS and pay for the 1742, they have a 1742 and pay for NS... I think it is just because nobody really blamed NeXT for not supporting the 1742 with a driver that works in pure EISA mode. The one and only EISA controller supported is the DPT 2022/2122 family. > 3. I think a lot of fans know that PAS16 and Adaptec 1542 > will fight each other in ISA bus motherboard. Is this going > to be the same if one use both of them in EISA bus motherboard??? > In another hand, assuming both 1542 and 1742 are functioning > in the same way (since they are using the same driver), will > 1742 and 1542 fight each other in EISA bus motherboard??? No, the 1542 and the PAS worked fine for me in an EISA board, because the EISA architecture does not worry about more than one busmastering card. Thus, an EISA board is an advantage even if you use it with ISA cards only. :-) Markus. -- /dev/ Markus Wenzel /usr/spool/mail/ mow@marsu.tynet.sub.org /etc/zoneinfo/ University of Stuttgart /bin/ps System administration, Consulting, Networking
Newsgroups: comp.sys.next.bugs From: ctm@ardi.com (Clifford T. Matthews) Subject: Re: A new bug with DOS partitions on NS3.2 [Intel] with EMM386.EXE Message-ID: <CrCI1B.Iuw@cobra.cs.unm.edu> Sender: news@cobra.cs.unm.edu Organization: ARDI References: <2t7igl$a06@rosie.next.com> <2takvq$353@senator-bedfellow.MIT.EDU> Date: Mon, 13 Jun 1994 17:03:14 GMT If you do any writing to the 100 Mb DOS filesystem from within NEXTSTEP you run the risk of corrupting the DOS filesystem, and such corruption is not a rare thing, it's quite common. NeXT knows about this bug. --Cliff ctm@ardi.com In article <2takvq$353@senator-bedfellow.MIT.EDU> lones@lones.mit.edu (Lones A. Smith) writes: >In article <2t7igl$a06@rosie.next.com> Joe_Keenan@next.com (Joe >Keenan) writes: >> In article <2t7ech$s9n@senator-bedfellow.MIT.EDU> >lones@lones.mit.edu (Lones A. >> Smith) writes: >> > The Next Boot Manager is causing some sort of DOS memory >> > management problem. >> >> That's incredibly unlikely. >> >> > I installed NeXTStep 3.2 on a DOS machine and I left >> > a 100MB DOS 6.0 partition. I was able to boot into the DOS >> > partition fine bnut Windows 3.1 would not run in 386 enhanced >> > mode. I then discovered that loading DOS 6.0's EMM386.EXE >> > caused the computer to hang when I tried to boot into DOS. >> >> When you say you "left a 100MB DOS partition", do you mean that >there was >> originally a larger partition there? If so, did you reformat & >reinstall on >> the DOS partition after you installed NS? If you didn't all bets >are off on >> the usability of the partition. >> >> The easy way to check is to boot DOS and see if the size of the >partition that >> FDISK reports is the same as what CHKDSK reports. If they don't, >the partition >> is no good. >> >> joe >> > > >I find this response VERY annoying. > >The install was done from a raw disk thru the NeXTStep Intall >process. >In fact, the NeXTStep help over the phone were at a loss to explain >what was happening, other than that we had found a new bug. I am >posting this bug so that NeXT will take it seriously and FIX IT for >3.3 --- for me and for others. It took me time to do this, and I do >hope that NeXT will give it the time it deserves. > >It has been unbelievable how many hurdles had to be surpassed to >install 3.2 on entirely NeXT-approved white hardware. Everything >required a special bug fix, and frequently the better hardware >features of the NeXT-sanctioned hardware were simply not supported. >(eg. the hard drive controller with a 10MG/sec limit could only be >set to 5, etc. etc.). > >It thus is frustrating that (this particular) NeXT (support help) >seeks to discount bug reports and blame it on the consumer. > >Sincerely, > >Lones A. Smith >Department of Economics, MIT >E52-252C, Cambridge MA 02139 >voice: (617) 253-0914 [fax: (617) 253-6915]
From: jweiss@casbah.acns.nwu.edu (Jerry Weiss) Newsgroups: comp.sys.next.bugs Subject: Re: A new bug with DOS partitions on NS3.2 [Intel] with EMM386.EXE Date: 10 Jun 1994 23:30:08 GMT Organization: Northwestern University, Evanston IL USA Message-ID: <2tat20$9ah@news.acns.nwu.edu> References: <2t7igl$a06@rosie.next.com> <2takvq$353@senator-bedfellow.MIT.EDU> In article <2takvq$353@senator-bedfellow.MIT.EDU>, Lones A. Smith <lones@lones.mit.edu> wrote: >In article <2t7igl$a06@rosie.next.com> Joe_Keenan@next.com (Joe >Keenan) writes: >> In article <2t7ech$s9n@senator-bedfellow.MIT.EDU> >lones@lones.mit.edu (Lones A. >> Smith) writes: >> > The Next Boot Manager is causing some sort of DOS memory >> > management problem. >> >> That's incredibly unlikely. >> <Munch> > >I find this response VERY annoying. > >The install was done from a raw disk thru the NeXTStep Intall >process. >In fact, the NeXTStep help over the phone were at a loss to explain >what was happening, other than that we had found a new bug. I am >posting this bug so that NeXT will take it seriously and FIX IT for >3.3 --- for me and for others. It took me time to do this, and I do >hope that NeXT will give it the time it deserves. > >It thus is frustrating that (this particular) NeXT (support help) >seeks to discount bug reports and blame it on the consumer. I also find it hard to believe the Boot manager is somehow screwing up EMM386. Simple to check though, create a bootable floppy and see if you can recreate the problem. My bet would be a corrupt file in the DOS partition (emm386?) or screwed up partition. > > >It has been unbelievable how many hurdles had to be surpassed to >install 3.2 on entirely NeXT-approved white hardware. Everything >required a special bug fix, and frequently the better hardware >features of the NeXT-sanctioned hardware were simply not supported. >(eg. the hard drive controller with a 10MG/sec limit could only be >set to 5, etc. etc.). Welcome to Intel Hardware Hell. Most of my Dos/Windoze machines require special drivers and another patch every other day or so. NeXT on Intel isn't perfect, but I place the blame mostly on shoddy designs for PC hardware. Alot of problems involve ISA bus which was designed around the cheapest glue chips in the early 1980 era. All the hassles with bus-mastering and interrupts or lack thereof can be traced back to designs that are hard pressed to handle advanced O/S's. -- Jerry S. Weiss j-weiss@nwu.edu Dept. Medicine, Northwestern Univ. Medical School, Chicago, Illinois %SYSTEM-S-PHALOKTARG, Phasers Locked on Target, Ready to Fire
From: Charles William Swiger <infidel+@CMU.EDU> Newsgroups: comp.sys.next.bugs Subject: Re: A new bug with DOS partitions on NS3.2 [Intel] with EMM386.EXE Date: Mon, 13 Jun 1994 17:59:11 -0400 Organization: Carnegie Mellon, Pittsburgh, PA Message-ID: <ghzBOjy00iV5M5I7E=@andrew.cmu.edu> In-Reply-To: <CrCI1B.Iuw@cobra.cs.unm.edu> Excerpts from netnews.comp.sys.next.bugs: 13-Jun-94 Re: A new bug with DOS part.. by Clifford T. Matthews@ardi.com > If you do any writing to the 100 Mb DOS filesystem from within NEXTSTEP > you run the risk of corrupting the DOS filesystem, and such corruption is > not a rare thing, it's quite common. NeXT knows about this bug. NeXT also says they've fixed it. Read the post dated 6/12/94 on c.s.n.announce titled "DOS File System Patch - available from NextAnswers". A brief summary: This update corrects a number of problems in the way NEXTSTEP Release 3.2 works with DOS file systems, most notably: - You can now use DOS partitions larger than 32 megabytes, without risk of corrupting the files on that partition. - The File Viewer now accurately reports the space available on DOS partitions larger than 32 megabytes. - You can now use removable DOS hard disks (such as Syquest and Bernoulli disks). - You can now work with some of the DOS partitions that NEXTSTEP Release 3.2 was unable to "mount." ************** -Chuck Charles William Swiger - WhiteLight Systems | "All the world's a stage, and" --------------------------------------------+ "we are merely players...." AMS & normal mail: infidel@cmu.edu | NeXTmail: chuck@cswiger.slip.andrew.cmu.edu | "Semper ubi sub ubi."
Newsgroups: comp.sys.next.bugs From: thompson@netcom.com (Eric Thompson) Subject: Re: Question: How to avoid the 64 MB/DOS bug? Message-ID: <thompsonCrD2GL.Cp2@netcom.com> Organization: Saperstein, et al. References: <2t6m3f$bu4@proffa.cc.tut.fi> <2t6o3g$eqr@sun0.urz.uni-heidelberg.de> Date: Tue, 14 Jun 1994 00:24:20 GMT [Re: DOS partition 32mb limit] >According to Erik, a bug fix for this problem is already finished and should >be on ftp.next.com in a couple of days (weeks ?). It's there now. Haven't tried it yet, though. Eric
From: rueiwun@helium.Gas.UUG.Arizona.EDU (ruei-wun tu) Newsgroups: comp.sys.next.bugs Subject: [Question]:Floppy Controller error during boot NS/FIP Date: 14 Jun 1994 06:00:19 GMT Organization: University of Arizona, Tucson Message-ID: <2tjh1j$p7v@news.CCIT.Arizona.EDU> References: <2snlp2$p49@gap.cco.caltech.edu> Hello, First of all, thank you for your time to answer my question I finally got my NS3.2 installed on my TYAN motherboard. Here is what I did. I only put one SCSI HD, which is on ID 0, and SCSI CD ROM, which is on SCSI ID 1. And I also put on my AHA 1542C and ATI Ultra Graphic Pro 2M, and, of course the IDE controller, which is used for my Floppy. It works without any problem. However, I found two serious problems after I finished the installation. 1. When I add my IDE HD latter on, the NS could not recognize the device. I could not mount it or use "fdisk" to format the IDE. 2. When booting the system. if I didn't insert the floopy disk in floppy driver, it shows the following warning message several times: FloppyCntIo:flushIntMsgs:Stray Interrupt After all, the system still boot successfully. And the floppy disk functionarity seems working O.K. Anyway, I am wondering if you have any ideals could solve those annoying problems. Thanks again for your help.
From: albert@proffa.cc.tut.fi (Ojala Pasi) Newsgroups: comp.sys.next.bugs Subject: Re: Question: How to avoid the 64 MB/DOS bug? Date: 14 Jun 1994 12:00:08 +0300 Organization: Tampere University of Technology Distribution: inet Message-ID: <2tjrio$km4@proffa.cc.tut.fi> References: <2t6m3f$bu4@proffa.cc.tut.fi> <2t6o3g$eqr@sun0.urz.uni-heidelberg.de> <thompsonCrD2GL.Cp2@netcom.com> This is a forwarded message from: Juha Tuominen >>According to Erik, a bug fix for this problem is already finished and should >>be on ftp.next.com in a couple of days (weeks ?). > > >It's there now. Haven't tried it yet, though. I have. Seems to work OK with 1.85 GB DOS partition which is formatted using DOS 6.2. Thanks NeXT, I was just about to loose my hope. -Juha
Newsgroups: comp.sys.next.bugs From: corby@netcom.com (Corby Anderson) Subject: Re: DOS filessystem bug fixes Message-ID: <corbyCrG7EG.CLJ@netcom.com> Keywords: DOS fs next Organization: Fight organized crime. Abolish the IRS. References: <2td0ii$77p@discovery.ectds.com> Date: Wed, 15 Jun 1994 17:03:51 GMT In article <2td0ii$77p@discovery.ectds.com>, Jim Armes <armes@tds.com> wrote: >I got the DOS fs patch last nite, it actually works! (not to be facetious or >anything.) >I mounted my second drive & secondary DOS partition (albiet a bit of terminal >typing later). I was able to read & write with no problems so far. > >Get it!! > >Jim Is there anything special that you have to do in order to get your second DOS partition recognized by NS? It didn't "just work." ;-) Can you tell us what your "bit of terminal typing" consisted of? Corby
From: darren@inherent.com (Inherent Network Admin Center) Newsgroups: comp.sys.next.bugs,comp.sys.next.sysadmin,comp.sys.next.software Subject: Re: Memory Leaks: What are they? Date: 15 Jun 1994 17:30:55 GMT Organization: Internetworks, Inc. Distribution: USA Message-ID: <2tndsg$dps@pdx1.world.net> Keywords: memory leaks Gentle Readers, I am rather new to the business of System Administration and have a question concerning a recent issue I have faced: memory leaks. I have just started to get deep into the technical issues for both Unix and NEXTSTEP and do not fully understand the concept of a memory leak. I lack a firm definition of what one is and how they work. My understanding does not go far beyond the beleif that it is linked with sporadic and untamed malloc calls. One of the systems that I am monitoring is apparently being effected by this type of bug. Within five days of a reboot the swap file has jumped from the normal 16.7 to over 40 megabytes. It is believed to be linked to the Front Desk application, but we have no way to detect this for sure. Does anyone know of a way to detect memory leaks? Has anyone else ever witness this problem either with Front Desk or with other applications? I would appreciate any information, especially further documentation that I may research concerning detection and prevention of memory leaks. Thanks in advance -- Darren McKee - Technical Staff | darren@inherent.com Network Administration Center | Tel.: 503-224-6751 Inherent Technologies, Inc. | Fax: 503-224-8872 2140 SW Jefferson St. Suite 130 | sysadmin@inherent.com Portland, Oregon USA 97201 |
Newsgroups: comp.sys.next.bugs From: stephen@rhodes.demon.co.uk (Stephen Rhodes) Subject: Re: STANDALONE TRANSSYS SLIP POSSIBLE? Message-ID: <CrGrEt.48M@demon.co.uk> Sender: news@demon.co.uk (Usenet Administration) Organization: Demon Internet References: <damonc.16.2DFB68FF@hookup.net> Date: Thu, 16 Jun 1994 00:16:04 GMT In article <damonc.16.2DFB68FF@hookup.net> damonc@hookup.net (Damon F. Cooper) writes: > Is there ANYONE out there that is running a STANDALONE NS box that > is successfully running TransSys PNI SLIP?? I posed a similar question > last week, and I got 31 (THIRTY ONE!!)responses from people looking for help > as well... > > Can anyone help? (Please!) > > Given a "fresh", standalone machine, install NEXTSTEP 3.2 on it, > what are the steps to follow? Anybody have this working that could > post/upload the config files/etc? > > THANK-YOU VERY MUCH IN ADVANCE!! > > Damon Cooper > damonc@hookup.net Yes I'm running it standalone and it seems to be working very well. It's rather tricky, however, to get it all right and I had to experiment quite a bit. I've now built a front end program to manage its starting and stopping. The basic trick is to ignore the auto starting configuration and the associated tcl scripts and launch the pnid program itself under one's own control. Setting up NS to handle mail etc. is also a lot of fun as there seem to be no explicit instructions. To change the host name you are forced to hack the hostconfig file directly replacing -AUTOMATIC- by ypur selected host name - until this is done sendmail gets collisions with the local mail server. You also need to set up sendmail as a mail server Stephen
Newsgroups: comp.sys.next.bugs From: massimo.ping.de!max (Max Boehm) Subject: copying files to DOS partition results in files of zero size Message-ID: <CrGnMs.C6@massimo.ping.de> Sender: max@massimo.ping.de (Max Boehm) Date: Wed, 15 Jun 1994 22:54:28 GMT I am using NS3.2/Intel. When I copy files to the DOS partition this results in files of zero length if the file size was a multiple of 8192 bytes. I have to log out and log in to see this. The same thing happens after installing the DOS fs patch from ftp.next.com. cd /tmp foreach f (256 512 1k 2k 4k 8k 16k 32k 64k 128k 256k) mkfile $f $f cp $f /ms-dos_5 end log out and log in again. ls -l /ms-dos_5 -rw-rw-rw- 1 max 0 Jun 15 23:07 128k -rw-rw-rw- 1 max 0 Jun 15 23:07 16k -rw-rw-rw- 1 max 1024 Jun 15 23:07 1k -rw-rw-rw- 1 max 256 Jun 15 23:07 256 -rw-rw-rw- 1 max 0 Jun 15 23:07 256k -rw-rw-rw- 1 max 2048 Jun 15 23:07 2k -rw-rw-rw- 1 max 0 Jun 15 23:07 32k -rw-rw-rw- 1 max 4096 Jun 15 23:07 4k -rw-rw-rw- 1 max 512 Jun 15 23:07 512 -rw-rw-rw- 1 max 0 Jun 15 23:07 64k -rw-rw-rw- 1 max 0 Jun 15 23:07 8k Any idea what has happend? --------------------------- Max Boehm boehm@informatik.uni-koeln.de max@massimo.ping.de
From: jimra@aol.com (JimRA) Newsgroups: comp.sys.next.bugs Subject: Re: DOS filessystem bug fixes Date: 16 Jun 1994 01:07:06 -0400 Organization: America Online, Inc. (1-800-827-6364) Sender: news@search01.news.aol.com Message-ID: <2tomlq$kim@search01.news.aol.com> References: <corbyCrG7EG.CLJ@netcom.com> In article <corbyCrG7EG.CLJ@netcom.com>, corby@netcom.com (Corby Anderson) writes: >Is there anything special that you have to do in order to get your >second DOS partition recognized by NS? It didn't "just work." ;-) >Can you tell us what your "bit of terminal typing" consisted of? It didn't immediately mount the drive,and the console gave me messages about the second drive being 'busy' (which is strange, since its a full disk DOS partition :> ) Anyway, the past couple of times I've booted it seems to mount them correctly without any intervention on my part (which is a little suprising, since I kind thought I'd have to add them to /etc/fstab.) SoftPC (which I played with a bit tonite) seems very confused and unhappy when I try to mount the partitions, so I think they will have to do some update to reset whatever they look at. Not to mention, it unmounted my second partition (very unsportsmanlike, I should think.) ;) All in all, fairly painless to get it working -- kinda makes me think they seriously tested it for this version :) Jim
From: schmo1@info.isbiel.ch (Olivier Schmid) Newsgroups: comp.sys.next.bugs Subject: Viper driver Date: 16 Jun 1994 10:07:32 GMT Organization: Biel School of Engineering, CH-2501 Biel, Switzerland Distribution: world Message-ID: <2tp895$opn@vega.info.isbiel.ch> The Viper - driver works great (but I have expected more speed with this card). There is a little problem when booting NS: The screen becomes black and you don't see anymore what's the status of the boot. Is there a solution for that ? Thanks Oli
From: michael@wft.stack.urc.tue.nl (Michael Brouwer) Newsgroups: comp.sys.next.bugs Subject: Re: STANDALONE TRANSSYS SLIP POSSIBLE? Date: 17 Jun 1994 22:55:28 +0200 Organization: WayForward Technologies Message-ID: <2tt2k0$m6k@tar.wft.stack.urc.tue.nl> References: <damonc.16.2DFB68FF@hookup.net> <CrGrEt.48M@demon.co.uk> In article <CrGrEt.48M@demon.co.uk>, Stephen Rhodes <stephen@rhodes.demon.co.uk> wrote: >In article <damonc.16.2DFB68FF@hookup.net> damonc@hookup.net (Damon F. Cooper) >writes: >[...stuff deleted...] To change the host name you are forced to hack the >hostconfig file directly replacing -AUTOMATIC- by ypur selected host name - >until this is done sendmail gets collisions with the local mail server. You >also need to set up sendmail as a mail server Do not forget to add the hostname and IP address of your standalone machine to NetInfo using Hostmanager.app if you edit /etc/hostconfig. If you do not do this, the gethostbyname() call will fail because your hosts name is not known to NetInfo thus breaking various programs. ---One of which is GNU emacs. Michael
From: "mmalcolm Crawford" <malc@dcs.shef.ac.uk> Newsgroups: comp.sys.next.bugs,comp.sys.next.programmer Subject: Re: Bug? Playing 8-bit 11kHz sound files double-speed? Date: Mon, 20 Jun 1994 18:18:38 GMT Organization: Department of Computer Science, University of Sheffield Message-ID: <940620191838.9321AACUE.malc@jeeves> References: <940618151222.4649AACUE.malc@jeeves> <2u4dtp$oof@hamblin.math.byu.edu> Mime-Version: 1.0 (Generated by Eloquent) Content-Type: text/plain; charset=US-ASCII > NeXT hardware doesn't have an option for 11kHZ sound files. It provides: > > 44.1 > 22.05 > 8.x (for mu-law) > > I imagine it's trying to play your sound at 8.x. > Ummm, I know that's what is usually *said*, but, unless my ears very much deceive me, there does appear to be support for other sampling rates. 11kHz 16-bit sounds seem fine... Anyway -- I've added an isPlayable line and conversion options, so, anybody want to try out Workspace Inspectors for AIFF and WAV sounds? Have fun, mmalcolm. SHeffield Auditory Group | Vox : (+44) 742 768555 ext 5569 Dept. Computer Science | direct : 825569 Sheffield University | Fax : (+44) 742 780972 Regent Court | Email: malc@dcs.shef.ac.uk 211 Portobello Street | (NeXTMail, SunMail, MIME welcome) Sheffield S1 4DP, UK. | (Read-Receipts discouraged :-)
From: swekla@ee.ualberta.ca (Brent Swekla) Newsgroups: comp.sys.next.bugs Subject: Re: STANDALONE TRANSSYS SLIP POSSIBLE? Date: 18 Jun 1994 09:18:39 GMT Organization: Computer and Network Services, U of Alberta, Edmonton, Canada Distribution: world Message-ID: <2tue5f$sov@quartz.ucs.ualberta.ca> References: <2tte5n$hgn@solaris.cc.vt.edu> In article <2tte5n$hgn@solaris.cc.vt.edu> jehu@jehu.async.vt.edu (john stanhope) writes: > In article <2tt2k0$m6k@tar.wft.stack.urc.tue.nl> > michael@wft.stack.urc.tue.nl (Michael Brouwer) writes: > > > > Do not forget to add the hostname and IP address of your standalone > > machine to NetInfo using Hostmanager.app if you edit /etc/hostconfig. > > If you do not do this, the gethostbyname() call will fail because > > your host's name is not known to NetInfo thus breaking various > > programs. ---One of which is GNU emacs. > > If you enter a host address without being on a network during boot I > believe the system will hang, at least that is what mine did. When I > set up PNI all I did was change my host name. To clarify: if you enter an ip address in the Local... panel of HostManager AND you are not "on a network during boot" your machine WILL hang - mine did :(. BTW, editing the "HOSTNAME=" line in /etc/hostconfig is exactly the same as changing the hostname using the Local... panel of HostManager (in NS 3.2, at least). However, I'm pretty sure what Michael is suggesting is using the Open menu of HostManager to edit host localhost in the root NetInfo domain, giving it your slip admin-supplied ip address and adding your hostname as an alias. I think you can also sort of get this to work by adding a new host, but this is not what you really want to do. The distinction between these two operations is subtle, yet critical; it had my head spinning for weeks. BTW, if you don't do this properly you not only break things (like the newsreader) in GNU emacs, you break talk as well. So if you can't get talk to work over a SLIP link, now you know why (give it a try John ;^). -- Brent Swekla "Never put off until tomorrow what you can do swekla@ee.ualberta.ca the day after tomorrow" - Mark Twain
From: damonc@hookup.net (Damon F. Cooper) Newsgroups: comp.sys.next.bugs Subject: HERE'S HOW: Standalone SLIP in 11 Easy Steps Date: Sat, 18 Jun 1994 13:27:23 GMT Organization: HookUp Communication Corporation, Oakville, Ontario, CANADA Message-ID: <damonc.32.2E02F63A@hookup.net> I got TransSys PNI 1.9 working on my standalone NS Intel 3.2 machines!!! Thanks to everyone who responded (WOW! Talk about a helpfull community!!) Every one of the replies put a piece in the puzzle. What I was doing wrong was a combination of things: 1) I had MTU in pni0.config left at 1006 when my provider had an MTU of 2000+ 2) I may have had "bad" chars in some of my config files (I replaced them with ones supplied/suggested and did #1 above and it "just worked"!!) Anyway, many thanks to all who helped out. If you are in a similar situation, and are still struggling to get a SLIP connection going, here are the basic steps (as I remember them anyway.) If I've missed anything, or misrepresented something, someone please correct me. Do all steps logged in as root. 1) Install NEXTSTEP. 2) Get an account with Internet provider and get info (IP addresses, etc) 3) Get and install TransSys PNI 1.9 (available @ cs.orst.edu). When installing, answer "YES" to the "Make Non-NetInfo Network". 4) Restart the machine. 5) In HostManager, do Hosts->Open->Local and: a) Set the IP address to the one given you by the provider. b) Add an alias. The alias should be the "Host Name" given you by the provider (in my case "damonc"). 6) Open /etc/hostconfig and change the HOSTNAME field from the default value of -AUTOMATIC- to the value of the "Host Name" you set in step 5(b) above. When finished, it should look similar to this (your HOSTNAME will be different, of course): # /etc/hostconfig # Local host configuration file # HOSTNAME=damonc INETADDR=-AUTOMATIC- ROUTER=-NO- IPNETMASK= IPBROADCAST=-AUTOMATIC- NETMASTER=-NO- YPDOMAIN=-NO- TIME=-NO- 7) Open /etc/resolv.conf (or create one if it doesn't exist so that your machine can find a nameserver. You need a) the IP address of the nearest Name Server and b) your Domain Suffix (your local domain) from your Internet provider. When finished, it should look similar to this (your addresses will be different, of course): # /etc/resolv.conf # Domain name resolver configuration file # domain tor.hookup.net nameserver 165.154.1.1 8) Restart the machine 9) Set up the TransSys PNI config files. a) Save the following as pni0.config in /etc/pni/config. CHECK WITH YOU INTERNET PROVIDER FOR THE CORRECT "MTU" TO USE. (Your addresses, etc will be different, of course): # # Pretty close to minimal configuration file for just plain SLIP using # # $Header: /local/SRCS/slip2/pnid/support/config.slip,v 1.4 1993/10/06 02:42:48 louie Exp $ ## PART I. set Config(pni:INTERFACE) pni0 set Config(pni:MTU) 2000 set Config(pni:ADDRESS) local-address set Config(pni:REMOTEADDRESS) remote-address set Config(pni:NETWORKMASK) 255.255.255.0 set Config(pni:DEFAULT) 1 set Config(slip:MODE) slip set Config(tty:DEVICE) /dev/cufa set Config(tty:SPEED) 38400 set Config(tty:MODEMTYPE) zyxel set Config(tty:SERVERTYPE) unix set Config(tty:SERVERNAME) SlipServerName set Config(tty:NUMBER) 416-630-0970 set Config(SlipServerName:DIALTYPE) TONE set Config(SlipServerName:USERNAME) username set Config(SlipServerName:PASSWORD) password # PART II. stack PNI pni stack SLIP slip stack TTY tty # PART III. proc LINK_start { encap } { log "LINK $encap connected" } proc LINK_stop { encap } { log "LINK $encap disconnected" } # # END OF FILE b) Save the following as login-unix.tcl in /etc/pni/support (Your login script will be different, of course): # Copyright 1993 by TransSys, Inc. # All rights reserved. # # $Header: /local/SRCS/slip2/pnid/support/login-unix.tcl,v 1.7 1994/05/14 20:15:32 louie Exp $ # # # Login script for use when remote end is some sort of UNIX box that will # drop into SLIP mode. # proc login-hookup { DIALER cfg } { global Config syslog LOG_INFO "Start Hookup login script" if {[info exists Config($cfg:SECRETFILE)] && [file exists $Config($cfg:SECRETFILE)]} { if {[file readable $Config($cfg:SECRETFILE)]} { source $Config($cfg:SECRETFILE) } file stat $Config($cfg:SECRETFILE) stat if {[expr $stat(mode)&044]} { error "File with password, $Config($cfg:SECRETFILE), is readable by non-owner!" } } else { if {[info exists Config($cfg:USERNAME)]} { set username $Config($cfg:USERNAME) } if {[info exists Config($cfg:PASSWORD)]} { set password $Config($cfg:PASSWORD) } } set timeout 20 # In case autobaud # $DIALER xmit "\r" # $DIALER sleep 1 $DIALER xmit "\r" $DIALER sleep 10 # Look for unix login prompt $DIALER expect timeout {error "waiting for login prompt"} "*ogin:*" # Send login poop $DIALER xmit "$username\r" $DIALER expect timeout {error "waiting for password"} "*assword:*" $DIALER sleep 1 $DIALER xmit "$password\r" # Wait for connection set timeout 60 $DIALER expect timeout {error "waiting for SLIP startup"} "*choice:*" $DIALER xmit "2\r" $DIALER sleep 5 # set timeout 20 syslog LOG_INFO "Entering SLIP mode" # That's all return 1 } c) Copy the Zyxel modem file (dial-zyxel.tcl) from the /etc/pni/support or /etc/pni/SAMPLES (can't remember which one?) to /etc/pni/config. This file name must match the value for "Config(tty:MODEMTYPE)" line in the /etc/pni/config/pni0.config file. 10) If you want to bring up the SLIP link when you start the machine, rename the /etc/pni/config/pni0.config to /etc/pni/config/pni0.config-auto. 11) Restart the machine. If you chose to set up the link on startup, you should now be connected. If you chose not to bring up the link on startup, enter the following in a terminal window to manually bring up the link: /etc/pni/pnid -t -d -c -f /etc/pni/config/pni0.config or /etc/pni/pnid -t -d -c -f /etc/pni/config/pni0.config-auto (if you are bringing up the link at startup but it failed) NOTE: this command line will work, but it shows test and debugging info. Lookup /etc/pni/pnirun in the TransSys doco for a "proper" command. THANKS AGAIN TO EVERYONE THAT HELPED OUT!! SEE YOU AT THE EXPO!! Regards, Damon F. Cooper damonc@hookup.net
From: robert@steffi.demon.co.uk (Robert Nicholson) Newsgroups: comp.sys.next.bugs Subject: Re: STANDALONE TRANSSYS SLIP POSSIBLE? Date: 18 Jun 1994 08:12:15 GMT Organization: me organized? That's a joke! Message-ID: <ROBERT.94Jun18091215@steffi.demon.co.uk> References: <damonc.16.2DFB68FF@hookup.net> <CrGrEt.48M@demon.co.uk> <2tt2k0$m6k@tar.wft.stack.urc.tue.nl> To: michael@wft.stack.urc.tue.nl (Michael Brouwer) In-reply-to: michael@wft.stack.urc.tue.nl's message of 17 Jun 1994 22:55:28 +0200 <michael@wft.stack.urc.tue.nl> writes: >In article <CrGrEt.48M@demon.co.uk>, >Stephen Rhodes <stephen@rhodes.demon.co.uk> wrote: >>In article <damonc.16.2DFB68FF@hookup.net> damonc@hookup.net (Damon F. Cooper) >>writes: >>[...stuff deleted...] To change the host name you are forced to hack the >>hostconfig file directly replacing -AUTOMATIC- by ypur selected host name - >>until this is done sendmail gets collisions with the local mail server. You >>also need to set up sendmail as a mail server >Do not forget to add the hostname and IP address of your standalone machine >to NetInfo using Hostmanager.app if you edit /etc/hostconfig. If you do not >do this, the gethostbyname() call will fail because your hosts name is not >known to NetInfo thus breaking various programs. ---One of which is GNU emacs. Can you explain this further please? Standalone machines... typically slip based. Do _not_ require an IP address in the INETADDR= field. It is actually wrong to do this and gethostbyname will not return a fully qualified domain name if you do this. HOSTNAME=steffi INETADDR= ROUTER=-NO- IPNETMASK= IPBROADCAST=-AUTOMATIC- NETMASTER=-NO- YPDOMAIN=-NO- TIME=-AUTOMATIC- -- "Real programmers don't create classes. They build hierarchies" (PGP key: send email with Subject: request pgp key) (ASCII for text only messages)
From: srosie@umich.edu (Steven S. Rosenblum) Newsgroups: comp.sys.next.bugs Subject: Strange boot behavior Date: 20 Jun 1994 19:45:00 GMT Organization: University of Michigan -- Ann Arbor I'm wondering if anyone else has observed the following problem. Message-ID: <2u4rjs$emn@lastactionhero.rs.itd.umich.edu> Originator: srosie@euripides.ccs.itd.umich.edu The really strange part of this behavior is that I _can_ get the drive to work if, after I attempt a boot, I shut the machine/external drives down and switch sd0's SCSI ID from 0 to 1 (or vice versa). My SCSI devices are: SCSI ID DEVICE 0/1 200 meg hard drive (with internal termination) (last device on scsi chain) 2 NeXT CD-ROM drive 4 PLI superfloppy 6 (?) Internal (NeXT supplied) 40 Meg hard drive 7(?) Mother board. -- NeXT Magneto-optical drive. Note that this was an upgraded '030 cube. Also, note that this boot behavior is _completely_ reproduceable. If the SCSI ID of sd0 is 1, then setting it to 0 after my first boot attempt will allow the system to boot normally. Similarly, if the SCSI ID of sd0 is 0, then switching it to 1 after my first boot attempt will allow the system to boot normally. You may wonder why I don't leave the system up all the time. The NeXT is not networked and sits in my bedroom, so I don't like leaving it on overnight. Please, if anyone else has seen this problem or has any idea why it occurs, please respond to me or the net. thanks in advance, rosie srosie@umich.edu (no NeXTMail, please)
From: "mmalcolm Crawford" <malc@dcs.shef.ac.uk> Newsgroups: comp.sys.next.bugs,comp.sys.next.programmer Subject: Bug? Playing 8-bit 11kHz sound files double-speed? Date: Sat, 18 Jun 1994 14:12:22 GMT Organization: Department of Computer Science, University of Sheffield Message-ID: <940618151222.4649AACUE.malc@jeeves> Mime-Version: 1.0 (Generated by Eloquent) Content-Type: text/plain; charset=US-ASCII WAAAAAAAAAARGH! Can somebody confirm or otherwise that NeXT h/w running NS 3.2 plays 8-bit, 11kHz sound files at double speed...? No wonder I couldn't find the bug in my program.... :-((( Malcolm. SHeffield Auditory Group | Vox : (+44) 742 768555 ext 5569 Dept. Computer Science | direct : 825569 Sheffield University | Fax : (+44) 742 780972 Regent Court | Email: malc@dcs.shef.ac.uk 211 Portobello Street | (NeXTMail, SunMail, MIME welcome) Sheffield S1 4DP, UK. | (Read-Receipts discouraged :-)
From: mengpao@helium.Gas.UUG.Arizona.EDU (mengpao cheng) Newsgroups: comp.sys.next.bugs Subject: [Question]: "NeXT Clock" differ than "System Clock"??? Date: 18 Jun 1994 14:27:31 GMT Organization: University of Arizona, Tucson Message-ID: <2tv08j$dp6@news.CCIT.Arizona.EDU> References: <2tte5n$hgn@solaris.cc.vt.edu> <2tue5f$sov@quartz.ucs.ualberta.ca> Hello, I found that my "system clock" which set by BIOS differ than the "clock" shown on the preference icon. I try to correct the "system clock", however, the NS seemsly can't read it. For example, if I set "system clock" on correct time at 7:30:00, I will get my "preference clock" on 12:30:00. It almost about 5 hours faster than "system clock". Althought when I set the "preference clock" at correct time and ignore the "system clock", it seemsly works fine, I still want to know if there is anyway I can fix this "bug"... Any suggestions welcome.... Rueiwun Tu
From: starksm@genesis.mcs.com (Scott M. Stark) Newsgroups: comp.sys.next.software,comp.sys.next.bugs Subject: Archie bug fix Date: 18 Jun 1994 16:01:06 GMT Organization: MCSNet Services Distribution: world Message-ID: <2tv5o2$l42@News1.mcs.com> Keywords: Archie Archie version 2.15, June 18 1994 -rw-rw-r-- 1 me 492440 Jun 18 10:37 Archie_215.tar.gz sum Archie_215.tar.gz -> 34778 481 Location: cs.orst.edu Path: pub/next/submissions or pub/next/binaries/wide-area-info Bug fixes: 1. Crash during ftp retrieval if session ivoked as service. Scott Stark starksm@genesis.mcs.com -- Scott Stark Stark International Software 4950 N. Marine Dr. #102 Chicago, IL 60640 starksm@genesis.mcs.com (NeXT mail accepted)
Newsgroups: comp.sys.next.bugs From: jbright@stimpy.carleton.ca (Jason Bright) Subject: Re: [Question]: "NeXT Clock" differ than "System Clock"??? Message-ID: <CrLp7A.BsI@cunews.carleton.ca> Sender: news@cunews.carleton.ca (News Administrator) Organization: Carleton University References: <2tv08j$dp6@news.CCIT.Arizona.EDU> Date: Sat, 18 Jun 1994 16:16:22 GMT In article <2tv08j$dp6@news.CCIT.Arizona.EDU> mengpao@helium.Gas.UUG.Arizona.EDU (mengpao cheng) writes: > Hello, > > I found that my "system clock" which set by BIOS differ than the "clock" shown on the preference icon. I try to correct the "system clock", however, the NS seemsly can't read it. For example, if I set "system clock" on correct time at 7:30:00, I will get my "preference clock" on 12:30:00. It almost about 5 hours faster than "system clock". Althought when I set the "preference clock" at correct time and ignore the "system clock", it seemsly works fine, I still want to know if there is anyway > This isn't a bug..it just depends on what you consider a clock to be. The 'system clock' maintains standard time with respect to Grenwich. (0 Longitude). The 'preferences clock' consists of the 'system clock' plus whatever time zone you have set yourself in. Thus, when you set the system clock to 7:30 (grenwich time), your preferences clock correctly displays 12:30 (arizona time)-- its not five hours faster, its the correct 7 hours behind. later jay jbright@stimpy.carleton.ca (NeXTMail happy!!)
From: michael@wft.stack.urc.tue.nl (Michael Brouwer) Newsgroups: comp.sys.next.bugs Subject: Re: STANDALONE TRANSSYS SLIP POSSIBLE? Date: 18 Jun 1994 11:29:58 +0200 Organization: WayForward Technologies Message-ID: <2tueqm$nat@tar.wft.stack.urc.tue.nl> References: <2tt2k0$m6k@tar.wft.stack.urc.tue.nl> <2tte5n$hgn@solaris.cc.vt.edu> In article <2tte5n$hgn@solaris.cc.vt.edu>, john stanhope <jehu@jehu.async.vt.edu> wrote: >In article <2tt2k0$m6k@tar.wft.stack.urc.tue.nl> >michael@wft.stack.urc.tue.nl (Michael Brouwer) writes: >> >> Do not forget to add the hostname and IP address of your standalone >> machine to NetInfo using Hostmanager.app if you edit /etc/hostconfig. If >> you do not do this, the gethostbyname() call will fail because your hosts >> name is not known to NetInfo thus breaking various programs. ---One of >> which is GNU emacs. > >If you enter a host address without being on a network during boot I >believe the system will hang, at least that is what mine did. When I >set up PNI all I did was change my host name. The right thing to do would then be to create an alias for localhost using HostManager.app with the new name of your machine. If your machine is not connected to a network, you still have a lo0 network device. By doing this, the hostname you enter in /etc/hostconfig can be resolved to the IP address of localhost. You should remove this alias as soon as you plan on connecting to a network though. (The above has not been tested, but I think this is the easyist and cleanest way to solve the problem). Another way to solve the hang during boot problem is to remove the `serves' property in machines/broadcasthost using NetInfoManager.app. This will stop nibindd from looking for a machine to serve the `../network' netinfo domain. You can then add a machine with HostManager.app with the name of your machine. And an arbitary IP address (not that of your SLIP interface). Michael
Newsgroups: comp.sys.next.bugs,comp.sys.next.software From: brad@instep.wimsey.bc.ca Subject: Edit.app : when is RTF NOT RTF? Message-ID: <1994Jun17.193237.8173@instep.wimsey.bc.ca> Sender: usenet@instep.wimsey.bc.ca Organization: InStep Mobile Communications Inc. Date: Fri, 17 Jun 1994 19:32:37 GMT I'm trying to send some rich text mail to someone who can't receive NeXTmail. Since I know RTF is "just text", I decide to fool Edit.app into opening the rtf file by removing the ".rtf" file extension. To my amazement and dismany, this did not work -- Edit still (incorrectly imho) opened the document and displayed it as Rich Text! So I ended up cat'ing the file in Terminal.app and copying and pasting the text into mail in order to achieve my goal. I should have been able to use Edit.app. For example if I have a file named HelloThere.rtf and inside is "Hello there!" in 64 point Helvetica. I should be able to remove the extension ".rtf", and have Edit open it as if it were just plain old ascii: {\rtf0\ansi{\fonttbl\f0\fswiss Helvetica;\f1\fmodern Ohlfs;} \paperw13040 \paperh11900 \margl120 \margr120 \pard\tx1340\tx2680\tx4020\tx5360\tx6720\tx8060\tx9400\tx10740\tx12080\tx13440\ f0\b0\i0\ulnone\fs128\fc0\cf0 Hello there! } Instead when I open the file HelloThere what Edit shows me is Rich text! Wrong! Wrong! Wrong! What's the point in having file name extensions?! Brad Head <brad@instep.wimsey.bc.ca> Software Developer, InStep Mobile Communications Inc. Vancouver, British Columbia CANADA
From: srosie@umich.edu (Steven S. Rosenblum) Newsgroups: comp.sys.next.bugs Subject: Re: Strange boot behavior Date: 20 Jun 1994 20:55:19 GMT Organization: University of Michigan -- Ann Arbor Message-ID: <2u4vnn$f59@lastactionhero.rs.itd.umich.edu> References: <2u4rjs$emn@lastactionhero.rs.itd.umich.edu> Originator: srosie@dante.ccs.itd.umich.edu (It forgot to describe the problem). Basically, whenever I boot my '040 cube (NS 3.2), the first attempt stops at the console because the internal 40Meg hard drive can't transfer data. If I shut the system down, reset my main drive (an external 200 meg drive) SCSI from 0 to 1 (or 1 to 0) the system boots fine -- until I shut down and try again. (I'll include some of the original post so people know what I'm talking about). In article <2u4rjs$emn@lastactionhero.rs.itd.umich.edu> srosie@umich.edu writes: > The really strange part of this behavior is that I _can_ get the >drive to work if, after I attempt a boot, I shut the machine/external drives >down and switch sd0's SCSI ID from 0 to 1 (or vice versa). My SCSI devices >are: > > SCSI ID DEVICE > 0/1 200 meg hard drive (with internal termination) > (last device on scsi chain) > 2 NeXT CD-ROM drive > 4 PLI superfloppy > 6 (?) Internal (NeXT supplied) 40 Meg hard drive > 7(?) Mother board. > -- NeXT Magneto-optical drive. > -sorry about the mis-post, -rosie
From: wampler@me10.lbl.gov (Russ Wampler) Newsgroups: comp.sys.next.bugs Subject: Bug in 68040 processor? Date: 20 Jun 1994 21:06:32 GMT Organization: Lawrence Berkeley Laboratory Distribution: world Message-ID: <2u50co$ha6@overload.lbl.gov> I was wondering if anyone knows what the problem was with the first version of the MC68040 processor that was used in the first NeXT workstations that were released. Apparently NeXT sent out replacement processors to anyone that had the computers with the first mask, but I'm wondering if the old processors still work and what errors would be expected from one.
From: giddings@sadie.chem.wisc.edu (Michael Giddings) Newsgroups: comp.sys.next.bugs,comp.sys.next.hardware Subject: [Help]: System won't boot Date: 18 Jun 1994 22:49:44 GMT Organization: Division of Information Technology Distribution: world Message-ID: <2tvtm8$ppu@news.doit.wisc.edu> This is posted for a friend, but please send any e-mail replies to me at giddings@whitewater.chem.wisc.edu and I will forward the info. The system: Lucky Goldstar '486 The prelude: Disk was filled to within 200k of maximum (500 MB SCSI) The final mitigating factor: Rebooting the system The disease: The system won't boot. -It pauses for several minutes at the "mounting remote filesystems" step, and then hangs completely at the "Starting File service daemons:" step. Other Symptoms: An unusually high number of "RDB connection" hangups during boot, about 50% of the time. Things already attempted to solve problem: - config=Default boot - single user boot to free up 10 MB of space - unplugging from the network - fsck Any help would be appreciated. Michael Giddings
From: andym@hq.af.mil (Andy McConnell) Newsgroups: comp.sys.next.bugs Subject: Re: [Question]: "NeXT Clock" differ than "System Clock"??? Date: 20 Jun 1994 21:19:08 GMT Organization: 7th Communications Group, Wash D.C. Message-ID: <2u514c$hn7@hq.hq.af.mil> References: <CrLp7A.BsI@cunews.carleton.ca> In article <CrLp7A.BsI@cunews.carleton.ca> jbright@stimpy.carleton.ca (Jason Bright) writes: > In article <2tv08j$dp6@news.CCIT.Arizona.EDU> > mengpao@helium.Gas.UUG.Arizona.EDU (mengpao cheng) writes: > > Hello, > > > > I found that my "system clock" which set by BIOS differ than the > "clock" shown on the preference icon. I try to correct the "system > clock", however, the NS seemsly can't read it. For example, if I set > "system clock" on correct time at 7:30:00, I will get my "preference > clock" on 12:30:00. It almost about 5 hours faster than "system clock". > Althought when I set the "preference clock" at correct time and ignore the > "system clock", it seemsly works fine, I still want to know if there is > anyway > > > > This isn't a bug..it just depends on what you consider a clock to be. The > 'system clock' maintains standard time with respect to Grenwich. (0 > Longitude). The 'preferences clock' consists of the 'system clock' plus > whatever time zone you have set yourself in. Thus, when you set the system > clock to 7:30 (grenwich time), your preferences clock correctly displays > 12:30 (arizona time)-- its not five hours faster, its the correct 7 hours > behind. > > later > jay > > jbright@stimpy.carleton.ca (NeXTMail happy!!) We've found this "feature" to be a problem when dual booting back and forth with DOS. We use a network-time clock setting procedure, which writes the LOCAL time to the system clock. NeXT interprets the system clock as UTC (Greenwich mean time). So the clock is 5 hours behind every time we reboot into NeXTstep. I think the problem can be solved by using one NeXT-only machine in the NetInfo domain as a time server; each NeXT looks to that one for the correct time on boot. Good luck! -- Lt Andy McConnell Communications Engineer andym@hq.af.mil 7th Communications Group, USAF, Pentagon ---NeXTMail accepted and encouraged--- (Opinions expressed are solely those of the author and not the USAF)
From: dcl@panix.com (David Lambert) Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs Subject: hideOtherApplications == nothing? Date: 11 Jun 1994 22:27:20 -0400 Organization: PANIX Public Access Internet and Unix, NYC Message-ID: <2tdrq8$c7h@panix.com> Keywords: workspace hideOtherApplications Has anyone else found that [[Application workspace] hideOtherApplications]; doesn't seem to do anything at all (at least under NS3.2/Intel)? Does it work on black machines? Has anyone else made it work on Intel machines? Is this just another piece of documented non-functionality? David Lambert dcl@homer.uu.panix.com <- please use this address to reply.
From: captain@arsenal.com (Andrew T. Foster) Newsgroups: comp.sys.next.bugs,comp.sys.next.hardware Subject: DOS.fs Patch File - A Question Date: 20 Jun 1994 23:30:03 GMT Organization: Division of Information Technology Message-ID: <2u58pr$dq@news.doit.wisc.edu> Howdy, Well, I grabbed this new patch and I have a question. 1. I've got a 2.0 gig Fujitsu with a 500meg NeXTSTEP partition and multiple extended DOS partitions. -- I desperatly need to mount these other partitions. NeXTSTEP still only recognizes the drive C partition. I've tried a host of mount command syntax's, but I still can't get it to work. Are there any kind souls that have gotten this new patch and successfully mounted logical extended DOS partitions? Thanks! - C -- _/_/_/ _/_/_/ Knight Enterprises - A Film/Video Production Company _/_/_/ C. Knight - captain@arsenal.com _/_/_/ (608) 251-5522 _/_/_/ (608) 251-5727 FAX _/_/_/
From: flexus!rfschtkt@maze.ruca.ua.ac.be (Raf Schietekat) Newsgroups: comp.sys.next.bugs Subject: Re: Default app, suffix and icon problem Date: 11 Jun 1994 02:30:40 -0500 Organization: UTexas Mail-to-News Gateway Sender: nobody@cs.utexas.edu Message-ID: <9406101751.AA03595@flexus> LS, (No specific sound bite to quote, read original message.) There is nothing wrong with this, and it's perfectly predictable. The system does some scanning to get to know the icons that applications vend for file extensions, and it caches this information, apparently in the file named by Steven. There is a shortcoming: instead of giving the user control of which application wins, this is decided by the path hierarchy (Dock, ~/Apps, /LocalLibrary/Apps etc.) and luck (within each element of that hierarchy). Then there is an independent mechanism that decides what application gets to open files with this extension by default, and this is controlled by the user from the WM's Tools Inspector (I don't know where that's stored, but that is all rather irrelevant). Except for the named shortcoming, it's a feature. :-) Raf Schietekat, RfSchtkt@maze.ruca.ua.ac.be, Flanders, Belgium *** The designation BANRUC60.BITNET for this host will disappear by July 1 *** (real, i.e., with triangle in the Deliver button) NeXTmail preferred Addressing limitations: no !, % or .uucp (? I should test this again) Compiler designers rarely do it. --- Aho, Sethi, Ullman, ``Compilers: Principles, Techniques, and Tools'', 2nd ed., p. 173
From: mengpao@helium.Gas.UUG.Arizona.EDU (mengpao cheng) Newsgroups: comp.sys.next.bugs Subject: Re: A new bug with DOS partitions on NS3.2 [Intel] with EMM386.EXE Date: 11 Jun 1994 10:48:02 GMT Organization: University of Arizona, Tucson Message-ID: <2tc4p2$5om@news.CCIT.Arizona.EDU> References: <2t7igl$a06@rosie.next.com> <2takvq$353@senator-bedfellow.MIT.EDU> Hello NeXT Fans, I have a couple of questions really annoying me. Hopefully I can gather some answers from all of you. On "NeXTAnswers #1475", it said that the Adaptec 1742 SCSI controller must be set in "Standard Mode" in order to run NS. It also mentioned that when you set Adaptec 1742 in "Standard Mode", it is compitable with Adaptec 1542. My questions are: 1. If we set 1742 in standard mode which is compatible with 1542, can 1742 still take advantage of 32bit EISA board. Or said can 1742 still has the ability to access the memory more than 16MB??? 2. Supposely one has EISA motherboard and want to choose either 1742 or 1542 controller. Since 1542 and 1742 are both using the same driver in NS, why would people pay more for using 1742 instead of using 1542 -- assuming both of them make no difference to each others. 3. I think a lot of fans know that PAS16 and Adaptec 1542 will fight each other in ISA bus motherboard. Is this going to be the same if one use both of them in EISA bus motherboard??? In another hand, assuming both 1542 and 1742 are functioning in the same way (since they are using the same driver), will 1742 and 1542 fight each other in EISA bus motherboard??? Any suggestions or answers will be so appreciated.... Rueiwun Tu
From: flexus!rfschtkt@maze.ruca.ua.ac.be (Raf Schietekat) Newsgroups: comp.sys.next.bugs Subject: Re: Edit.app : when is RTF NOT RTF? Date: 19 Jun 1994 14:07:00 -0500 Organization: UTexas Mail-to-News Gateway Sender: nobody@cs.utexas.edu Message-ID: <9406191818.AA02762@flexus> LS, No bug. Info>Preferences...Temporary Settings>Rich Text Options Raf Schietekat, RfSchtkt@maze.ruca.ua.ac.be, Flanders, Belgium *** The designation BANRUC60.BITNET for this host will disappear by July 1 *** (real, i.e., with triangle in the Deliver button) NeXTmail preferred Addressing limitations: no !, % or .uucp (? I should test this again)
Newsgroups: comp.sys.next.bugs From: peter@corsica (Peter Eisch) Subject: Re: Bug in 68040 processor? Message-ID: <Crq7Ao.Hzq@news.cis.umn.edu> Sender: news@news.cis.umn.edu (Usenet News Administration) Organization: University of Minnesota Hospital and Clinic, Laboratory Information Services References: <2u50co$ha6@overload.lbl.gov> Date: Tue, 21 Jun 1994 02:36:47 GMT I know that the ones I came across that had "SAMPLE" stamped on them had flakey serial port behavior. They were both color and mono slabs. On one there was a bad ethernet controller, but there may not be a correlation in this case (that may have been a turbo as I think about it, so it wouldn't apply). peter -- Always looking for a handy place to nap... peter@tahiti.umhc.umn.edu (Peter Eisch) peisch@cfa.org
From: rueiwun@helium.Gas.UUG.Arizona.EDU (ruei-wun tu) Newsgroups: comp.sys.next.bugs,comp.sys.next.programmer Subject: Questions about Adaptec 1742 and 1542??? Date: 11 Jun 1994 12:36:54 GMT Organization: University of Arizona, Tucson Message-ID: <2tcb56$6sg@news.CCIT.Arizona.EDU> References: <1994Jun1.094530.1114@dart.de> Hello NeXT Fans, I have a couple of questions really annoying me. Hopefully I can gather some answers from all of you. On "NeXTAnswers #1475", it said that the Adaptec 1742 SCSI controller must be set in "Standard Mode" in order to run NS. It also mentioned that when you set Adaptec 1742 in "Standard Mode", it is compitable with Adaptec 1542. My questions are: 1. If we set 1742 in standard mode which is compatible with 1542, can 1742 still take advantage of 32bit EISA board. Or said can 1742 still has the ability to access the memory more than 16MB??? 2. Supposely one has EISA motherboard and want to choose either 1742 or 1542 controller. Since 1542 and 1742 are both using the same driver in NS, why would people pay more for using 1742 instead of using 1542 -- assuming both of them make no difference to each others. 3. I think a lot of fans know that PAS16 and Adaptec 1542 will fight each other in ISA bus motherboard. Is this going to be the same if one use both of them in EISA bus motherboard??? In another hand, assuming both 1542 and 1742 are functioning in the same way (since they are using the same driver), will 1742 and 1542 fight each other in EISA bus motherboard??? Any suggestions or answers will be so appreciated.... Rueiwun Tu
Newsgroups: comp.sys.next.bugs From: paul@seer.demon.co.uk (Paul Lynch) Subject: Re: Default app, suffix and icon problem Message-ID: <1994Jun12.085758.1660@seer.demon.co.uk> Sender: news@seer.demon.co.uk Organization: P & L Systems References: <9406101751.AA03595@flexus> Date: Sun, 12 Jun 1994 08:57:58 GMT In article <9406101751.AA03595@flexus> flexus!rfschtkt@maze.ruca.ua.ac.be (Raf Schietekat) writes: > LS, > > (No specific sound bite to quote, read original message.) > > There is nothing wrong with this, and it's perfectly predictable. The system > does some scanning to get to know the icons that applications vend for file > extensions, and it caches this information, apparently in the file named by > Steven. There is a shortcoming: instead of giving the user control of which > application wins, this is decided by the path hierarchy (Dock, ~/Apps, > /LocalLibrary/Apps etc.) and luck (within each element of that hierarchy). This is controlled by the WorkspaceSearchPath dwrite; I haven't tried changing the sequence, but I suspect it might work. IMHO, the bug isn't this; it is that the icon offered by the Workspace is not related to which app will open the document (if you have used the Tools inspector to override the default app). Paul -- Paul Lynch P & L Systems (NeXTmail) paul@seer.demon.co.uk Tel: (0494)671501 9 Stable Lane, Seer Green, Fax: (0494)680228 Bucks, HP9 2YT, UK
From: jaked@well.sf.ca.us (Jake Donham) Newsgroups: comp.sys.next.bugs Subject: Terminal problems telnet'ing to a NeXT Date: 15 Jun 1994 07:37:50 GMT Organization: The Whole Earth 'Lectronic Link, Sausalito, CA Distribution: world Message-ID: <JAKED.94Jun15003750@well.well.sf.ca.us> When I telnet to a NeXT (I've seen this on a few machines) there are (sometimes) linefeeds without carriage returns. I've twiddled stty options to no avail. How do I fix this? It's really annoying. (I looked on rtfm.mit.edu for a FAQ but there seems to be no comp.sys.next hierarchy there.) Thanks, Jake
From: nathan@laplace.csb.yale.edu (Nathan F. Janette) Newsgroups: comp.sys.next.bugs Subject: Re: A new bug with DOS partitions on NS3.2 [Intel] with EMM386.EXE Date: 12 Jun 1994 19:46:39 GMT Organization: Yale University, Department of Computer Science, New Haven, CT Distribution: world Message-ID: <2tfomv$hce@babyblue.cs.yale.edu> References: <2tc4p2$5om@news.CCIT.Arizona.EDU> In article <2tc4p2$5om@news.CCIT.Arizona.EDU> mengpao@helium.Gas.UUG.Arizona.EDU (mengpao cheng) writes: > 2. Supposely one has EISA motherboard and want to choose > either 1742 or 1542 controller. Since 1542 and 1742 are > both using the same driver in NS, why would people pay more > for using 1742 instead of using 1542 -- assuming both of > them make no difference to each others. Some people, such as myself, were fooled by one release of the NeXT hardware compatibility guide that listed the 1740 without mentioning they would only be supporting it in bullshit mode. Most annoying! -- Nathan "USENET" Janette Systems Manager, Brunger Lab, Dept MB&B, Yale Univ/HHMI, New Haven, CT PPP link from hilbert.csb.yale.edu Please reply to: nathan@laplace.csb.yale.edu (NeXT)
From: damonc@hookup.net (Damon F. Cooper) Newsgroups: comp.sys.next.bugs Subject: STANDALONE TRANSSYS SLIP POSSIBLE? Date: Sun, 12 Jun 1994 19:58:55 GMT Organization: HookUp Communication Corporation, Oakville, Ontario, CANADA Message-ID: <damonc.16.2DFB68FF@hookup.net> Is there ANYONE out there that is running a STANDALONE NS box that is successfully running TransSys PNI SLIP?? I posed a similar question last week, and I got 31 (THIRTY ONE!!)responses from people looking for help as well... Can anyone help? (Please!) Given a "fresh", standalone machine, install NEXTSTEP 3.2 on it, what are the steps to follow? Anybody have this working that could post/upload the config files/etc? THANK-YOU VERY MUCH IN ADVANCE!! Damon Cooper damonc@hookup.net
From: droege@informatik.uni-koblenz.de (Detlev Droege) Newsgroups: comp.sys.next.bugs Subject: Re: Edit.app : when is RTF NOT RTF? Date: 20 Jun 1994 12:33:51 GMT Organization: University Koblenz / Germany Distribution: world Message-ID: <2u42bf$op1@newshost.uni-koblenz.de> References: <1994Jun17.193237.8173@instep.wimsey.bc.ca> In article <1994Jun17.193237.8173@instep.wimsey.bc.ca> brad@instep.wimsey.bc.ca writes: > > I'm trying to send some rich text mail to someone who can't > receive NeXTmail. Since I know RTF is "just text", I decide to fool Edit.app > into opening the rtf file by removing the ".rtf" file extension. > > To my amazement and dismany, this did not work -- Edit still (incorrectly imho) > opened the document and displayed it as Rich Text! I'd consider that behaviour ok. E.g. you couldn't have Rich Text source code otherwise. But if you aim is to get RTF uninterpreted into the pasteboard, open a Terminal window and type copy < theFile.rtf See "man copy" for more info. If you really want to edit the RTF "code" see Raf Schietekat's answer: Info>Preferences...Temporary Settings>Rich Text Options -- Detlev Droege, Uni Koblenz, FB Informatik, Rheinau 1, D-56075 Koblenz, Germany Tel: +49 261 9119-421, Fax: -497, NeXT/Email: droege@informatik.uni-koblenz.de
From: "mmalcolm Crawford" <malc@dcs.shef.ac.uk> Newsgroups: comp.sys.next.bugs,comp.sys.next.programmer Subject: Re: Bug? Playing 8-bit 11kHz sound files double-speed? Date: Tue, 21 Jun 1994 22:14:17 GMT Organization: Department of Computer Science, University of Sheffield Message-ID: <940621231417.13592AACUH.malc@jeeves> References: <940618151222.4649AACUE.malc@jeeves> <2u4dtp$oof@hamblin.math.byu.edu> <940620191838.9321AACUE.malc@jeeves> <2u7a0m$p1a@hamblin.math.byu.edu> Mime-Version: 1.0 (Generated by Eloquent) Content-Type: text/plain; charset=US-ASCII > Wait a minute... I just remembered... with 3.1, didn't they add code to the > sound mixer to arbitrarily push any sampling rate to 44.1? It's somewhere > in the release notes if I recall. > Yes, it is... I've tried various sampling rates now: 17kHz seems to play fine for example... but maybe I'm imagining it... 8-) What is very frustrating is that I'm now checking with [aSound isplayable] (from the docs) Returns YES if the Sound can be played, otherwise returns NO. Some unplayable Sounds just need to be converted to another format, sampling rate, or number of channels; others are inherently unplayable, such as those whose format is SND_FORMAT_DISPLAY. To play a Sound that's just been recorded from the DSP, you must change its format from SND_FORMAT_DSP_DATA_16 to SND_FORMAT_LINEAR_16. and for 11kHz Sounds this returns YES. This means that my Inspector doesn't change the format, so plays out double speed. If possible I'd rather avoid either (a) simply converting all sounds to 22.05kHz 16-bit stereo, as this takes time and memory, or (b) trying to exhaustively list any formats which will fail an converting them... any suggetions? I presume I am correct and this is a bug? Have fun, mmalcolm. SHeffield Auditory Group | Vox : (+44) 742 768555 ext 5569 Dept. Computer Science | direct : 825569 Sheffield University | Fax : (+44) 742 780972 Regent Court | Email: malc@dcs.shef.ac.uk 211 Portobello Street | (NeXTMail, SunMail, MIME welcome) Sheffield S1 4DP, UK. | (Read-Receipts discouraged :-)
Newsgroups: comp.sys.next.programmer,comp.sys.next.bugs From: Jayson Adams Subject: Re: hideOtherApplications == nothing? Message-ID: <1994Jun12.203108.749@millennium.com> Sender: jayson@millennium.com Organization: Millennium Software Labs, Inc. References: <2tdrq8$c7h@panix.com> Date: Sun, 12 Jun 1994 20:31:08 GMT In article <2tdrq8$c7h@panix.com>, David Lambert writes: > Has anyone else found that > > [[Application workspace] hideOtherApplications]; > > doesn't seem to do anything at all (at least under > NS3.2/Intel)? I'm pretty sure this doesn't work (at least that was the case in 3.1). __jayson
Newsgroups: comp.sys.next.bugs From: sdavenpo@vaxc.stevens-tech.edu Subject: PAS Sound and Adaptec 1542C Message-ID: <1994Jun22.153550.1@vaxc.stevens-tech.edu> Sender: news@dmi.stevens-tech.edu (USENET News System) Organization: Stevens Institute Of Technology Date: Wed, 22 Jun 1994 20:35:50 GMT I've just upgraded to 3.2 (Intel) and installed a PAS sound card. Unfortunately, sounds longer than one second start to chop apart (an example: ni.snd sounds find, but rooster.snd gets broken and chopped). I have an Adaptec SCSI card in the machine at the defaults (port 330, irq 11, dma 5) and the PAS is at dma 7, port 338, irq 5. I've read the NeXTanswers and release notes - which basically said try using different combinations of dma's and irq's. Well - i tried a bunch and had no luck. I then tried using the PASPlus driver at ftp.next.com inthe 3_2Drivers area and still had no luck. Has anyone out there got these two boards working together and have no problems with sound playback? Scott Davenport sdavenpo@vaxc.stevens-tech.edu
Newsgroups: comp.sys.next.bugs From: Kjell.Borg@um.erisoft.se (Kjell Borg) Subject: Re: Viper driver (What about VideoBlitz?) Message-ID: <1994Jun22.093438.22033@um.erisoft.se> Sender: usenet@um.erisoft.se (For NNTP posting) Organization: Erisoft AB Umea Sweden References: <2tp895$opn@vega.info.isbiel.ch> Date: Wed, 22 Jun 1994 09:34:38 GMT >The Viper - driver works great (but I have expected more speed with this card). >There is a little problem when booting NS: The screen becomes black and you don't >see anymore what's the status of the boot. >Is there a solution for that ? > >Thanks Oli > Well, that sound nice. Unfortunately, I have a Genoa Videoblitz 9200VL, but I can't get the driver to work! Has anyone else had the problem that the machine hangs at boot when it is supposed to sense the VideoBlitz card (Just after the "Beep" has been installed)? I get an image when the Standard VGA driver is used, so the hardware works, but not via the Genoa VideoBlitz driver. Anyone else using the Genoa hardware? Please. Kjell --- -- --------------------------------------------------------------------- || Kjell Borg | Email: Kjell.Borg@um.erisoft.se || || Erisoft Umea | MEMO: ECOM.EPLKBG || || Box 4205 | Tel: +46 90-153059 || || S-904 06 Umea | Fax: +46 90-153099 || || Sweden | || ---------------------------------------------------------------------
From: wln@news.cs.columbia.edu (William Lee Nussbaum) Newsgroups: comp.sys.next.bugs,comp.sys.next.misc Subject: PS printer errors with NXImages? Anyone else experience this?... Date: 23 Jun 1994 01:37:39 -0400 Organization: Columbia University Department of Computer Science Message-ID: <2ub733$da4@pizza.cs.columbia.edu> I have, consistently, for several months, been getting the following type of error when printing most (not all) documents with NeXTSTEP-generated bitmaps to an HP LaserJet 4: ERROR: rangecheck OFFENDING COMMAND: get STACK: -1022 ( !!"#$%&()*+,,--../0011223344566778899:;;;<<==>>??@@AABBCCDDEFFGGHHIIJJ ) (where the line with the serial text continues off the side of the page..., and the number is usually approximately -1K, + or - (usu -) 0 to 2) The NeXT is connecting to the printer via an ethernet JetDirect card, through a pty-to-tcp... Incoming faxes print properly (have received up to 20p in some). Bitmaps in jobs generated on other platforms (Sun Frame, etc.) print properly. Long, 80+ page documents print properly. Bitmaps that use the appkit to produce their PostScript will usually die partway through printing (some of the image will usually print), taking the rest of the job with them... The above consistent success and failures lead me to believe that the problem is specifically in the NXImage class: problems in anything relating to the printer connection would affect other types of jobs. These jobs render on screen; the same files that crash the printer preview happily in Preview.app... If someone can test on something besides a LaserJet 4 or a NeXTprinter (I seem to recall the NeXTprinter being happy with these jobs, though that was a couple months ago) your help would be appreciated... Does anyone have any clue what NeXT is doing here that's not kosher, is tolerated by NeXT's Adobe Level 2 PS server, but upsets HP's Adobe Level 2 PS priners? And any idea what one might do about it? (besides getting NeXT aware of it...) Thanks, hope someone can help.. - Lee
Newsgroups: comp.sys.next.bugs From: rpomeroy@aunext1.amd.com (Ron Pomeroy x(Coop)) Subject: Re: Still no DOS SyQuest access. Message-ID: <CsHLuo.n4n@txnews.amd.com> Sender: news@txnews.amd.com Organization: Advanced Micro Devices, Austin TX References: <2vceva$9hj@rosie.next.com> Distribution: usa Date: Tue, 5 Jul 1994 21:47:11 GMT In article <2vceva$9hj@rosie.next.com> erikkay@next.com (Erik Kay) writes: >>In article <1994Jul1.152859.1@vaxc.stevens-tech.edu> >>sdavenpo@vaxc.stevens-tech.edu writes: >>] I just patched up my machine with the patch to fix up access to DOS >>harddrives, >>] which also stated that dos formatted SyQuest drives would be readable >>under NS. >>] >>] Yet - I still can mount a dos formatted SyQuest - I always get a SCSI >>disk not >>] readable, Eject/Initialize message. >>] >>] Anyone have any clues? >> >>Are you running on Intel or m68k? If it's m68k, then unfortunately, it >>still doesn't work even with the patch. Sorry. It's a simple >>byte-swapping bug that's been fixed for 3.3. >> >>(Thus are the perils of not having black hardware on my desk any more.) >> >>Erik Speaking of bug fixes in 3.3, will NeXT publish a comprehensive list of bugs that have been fixed in 3.3 ? -- Ronald Pomeroy Advanced Micro Devices CIM Applications Group rpomeroy@aunext1.amd.com
From: erikkay@next.com (Erik Kay) Newsgroups: comp.sys.next.bugs Subject: Re: Still no DOS SyQuest access. Date: 5 Jul 1994 22:39:02 GMT Organization: NeXT, Inc. Distribution: usa Message-ID: <2vcne7$9m3@rosie.next.com> References: <CsHLuo.n4n@txnews.amd.com> In article <CsHLuo.n4n@txnews.amd.com> rpomeroy@aunext1.amd.com (Ron Pomeroy x(Coop)) writes: ] Speaking of bug fixes in 3.3, will NeXT publish a comprehensive list of bugs ] that have been fixed in 3.3 ? I don't know if there will be a single comprehensive list. However, we do try to include a list of bug fixes and enhancements since the previous release in the release notes for each release. I admit that often we're not great here, but that's where to look. Erik
Newsgroups: comp.sys.next.bugs From: Jacques Garbi Subject: Re: Still no DOS SyQuest access. Message-ID: <CsDvwC.B0@touga.vd.alphanet.ch> Sender: jacques@touga.vd.alphanet.ch (Jacques Garbi) Organization: Touga Management SA References: <1994Jul1.152859.1@vaxc.stevens-tech.edu> Date: Sun, 3 Jul 1994 21:33:47 GMT In article <1994Jul1.152859.1@vaxc.stevens-tech.edu> sdavenpo@vaxc.stevens-tech.edu writes: > I just patched up my machine with the patch to fix up access to DOS harddrives, > which also stated that dos formatted SyQuest drives would be readable under NS. > > Yet - I still can mount a dos formatted SyQuest - I always get a SCSI disk not > readable, Eject/Initialize message. > > Anyone have any clues? > Well, I don't know what's wrong with your SyQuest ? I'm using one myself and it works perfectly. The only thing is I can't initialize a SyQuest in DOS format from NS. I can't initialize an already DOS formated cartridge in NeXT format. To do so, I'm obliged to first initialize it as MacIntosh (still from NS) and then initialize it NeXT format. Keep smiling, it works. Jacques GARBI
From: work@dannug.dk (Michael Hallin) Newsgroups: comp.sys.next.bugs Subject: Re: Probs with Seagate & NS = Panic's Date: 4 Jul 1994 22:24:56 GMT Organization: Danish NeXT User Group Message-ID: <2va27o$136@machthenext.dannug.dk> References: <Cs8zyL.8BG@softpac.com.au> In article <Cs8zyL.8BG@softpac.com.au> Dave THOMAS <dave@softpac.com.au> writes: |>We are having problems with Seagate drives, which we now suspect may be |>related to the NeXTSTEP operating environment. It seems only to occur (stuff deleted) Hi, I have only used black hardware, but I have tried these two and still use one the drives, ST11200N and ST31200N, and I never experienced any problems. More likely your problems are controller-related.... but I couldn t be sure, as I use black hardware... Best Michael --- _____________________________________________ Michael Hallin Copenhagen, Denmark NeXTMail: work@dannug.dk NonNeXTMail: mh.xeroxvang@rxdk.xerox.com Voice: Int + 45 43 53 34 33 Fax: Int + 45 43 53 34 33 Work: Int + 45 43 43 43 03 Fax: Int + 45 43 43 59 70 _____________________________________________
From: erikkay@next.com (Erik Kay) Newsgroups: comp.sys.next.bugs Subject: Re: Still no DOS SyQuest access. Date: 5 Jul 1994 20:14:33 GMT Organization: NeXT, Inc. Message-ID: <2vceva$9hj@rosie.next.com> References: <1994Jul1.152859.1@vaxc.stevens-tech.edu> In article <1994Jul1.152859.1@vaxc.stevens-tech.edu> sdavenpo@vaxc.stevens-tech.edu writes: ] I just patched up my machine with the patch to fix up access to DOS harddrives, ] which also stated that dos formatted SyQuest drives would be readable under NS. ] ] Yet - I still can mount a dos formatted SyQuest - I always get a SCSI disk not ] readable, Eject/Initialize message. ] ] Anyone have any clues? Are you running on Intel or m68k? If it's m68k, then unfortunately, it still doesn't work even with the patch. Sorry. It's a simple byte-swapping bug that's been fixed for 3.3. (Thus are the perils of not having black hardware on my desk any more.) Erik
From: gmk@pegasos.ccsr.uiuc.edu (Gottfried Mayer-Kress) Newsgroups: comp.sys.next.bugs Subject: Mailboxes won't show up in NeXT Mail Date: 2 Jul 1994 17:57:42 GMT Organization: University of Illinois at Urbana Message-ID: <2v49qm$h30@vixen.cso.uiuc.edu> It happened just recently that the Command-M (Mailboxes) command would not open the Mailboxes in my Mail application on my NeXT cube (version 3.2(v100)) I have the impression that it might have something to do with Bill Romine's MailSearch program. Any help would be appreciated. Thanks, -- Gottfried Mayer-Kress Center for Complex Systems Research, Beckman Institute, 405 N Mathews, Urbana, Il 61801,(217)-244-5877(voice/fax1),x8371(fax2), x1994 (msg) gmk@pegasos.ccsr.uiuc.edu, gmk@igc.apc.org, gmk@santafe.edu URLs: http://www.ccsr.uiuc.edu/People/gmk/gmk-b.html file://ftp.santafe.edu/pub/Users/gmk
From: klingler@unm.edu (Dave "CIRT Boy" Klingler) Newsgroups: comp.sys.next.hardware,comp.sys.next.bugs,comp.sys.next.sysadmin Subject: Re: Probs with Seagate & NS = Panic's Date: 4 Jul 1994 13:40:00 -0600 Organization: University of New Mexico, Albuquerque Message-ID: <2v9oig$1qr5@argo.unm.edu> References: <Cs8zyL.8BG@softpac.com.au> Keywords: seagate, problem,panic In article <Cs8zyL.8BG@softpac.com.au>, Dave THOMAS <dave@softpac.com.au> wrote: >We are having problems with Seagate drives, which we now suspect may be >related to the NeXTSTEP operating environment. It seems only to occur with >certain drives, not all. We have checked out different revisions of >firmware, different controllers and are unable to find any problem with >these or the drives, except when we load NeXTSTEP. Is there a known >problem with Seagate Drives? What do the error messages actually mean? > I've been tearing my hair out trying to get an ST11200N working on my cube. I've tried all three termination options, active termination, and disabling parity. I'd be happy to get some answers, too. The error messages basically mean that the drive is not working on your system. You have some corrupted files that the system needs in order to boot. The file system on my drive (and I think you have the same problem) gets corrupted on the fly. I don't have any great suggestions, except that you can't trust those systems until you get the problem licked. Please let me know if you do. Dave
These are the contents of the former NiCE NeXT User Group NeXTSTEP/OpenStep software archive, currently hosted by Marcel Waldvogel and Netfuture.ch.