Received: from vmb.brl.mil by VMB.BRL.MIL id aa01438; 10 May 90 1:59 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa01298; 10 May 90 1:38 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa01245; 10 May 90 1:25 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa00675; 10 May 90 1:14 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA06378; Wed, 9 May 90 22:02:07 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 10 May 90 04:54:34 GMT From: Matthew A Machlis Organization: Massachusetts Institute of Technology Subject: IRIS Video Recording Message-Id: <1990May10.045434.25681@athena.mit.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Can anyone tell me if there is a way of recording the video signal from an IRIS 3020? I would like to record either the direct 30Hz output from the computer or the RGB signal output when the computer is put into NTSC mode. Any info would be appreciated.   Received: from vmb.brl.mil by VMB.BRL.MIL id aa03085; 10 May 90 7:42 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa03020; 10 May 90 7:31 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa02979; 10 May 90 7:15 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa00848; 10 May 90 7:04 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA21818; Thu, 10 May 90 03:58:08 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 10 May 90 06:52:47 GMT From: mcsun!cernvax!chx400!urz.unibas.ch!doelz@uunet.uu.net Organization: University of Basel, Switzerland Subject: Re: CG2 information needed Message-Id: <1990May10.075247.719@urz.unibas.ch> References: <9005091811.AA26683@> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <9005091811.AA26683@>, lmo@lsr-vax.UUCP ("Lance M. Optican - LMO") writes: > Subject: GC2 Genlock board. > > I would like to know how to turn on my genlock board, > and what sequence of setvideo() calls I need to make it all work: modl [/usr/people/4Dgifts/examples/video] % ls Makefile cg0 cg1 cg2 cg3 cgmode de3 video README cg0.c cg1.c cg2.c cg3.c Try one of these, and don't be disappointed about the 'quality'. (flame on) Especially the scan conversion of a one-line grid aligned on the screen is insufficient; some lines are there, others perfectly absent. Overlay mode does not even work. I showed the 'performance' to both local and international SGI people and they were 'impressed'. Colors and stability of the picture are beyond any discussion. Though having admitted the problem (which makes it impossible to use the equipment for tecting purposes as intended) SGI failed to get the CG2 hardware work out properly. We are complaining for nearly 2 years now (!) without getting it done - ... (flame off, irony on.) On our 4D/120 we needed to reboot the machine, unplug the board, wait for a better moon phase, and promptly something came out the right port. Our SONY equipment was even capable to catch the signal provided that nno OVERLAY mode was used. It's like in bad science fiction movies: You look at something you believed that it could be realistic some time in the future but you don't recognize what it's meant to be... (irony off) You should forget the CG2 and get a different equipment provided that you convince SGI to pay for it. - Reinhard ************************************************************************ Dr. Reinhard Doelz * EAN doelz@urz.unibas.ch Biocomputing * DECNET 48130::doelz Biozentrum der Universitaet * X25 psi%46211142::embnet Klingelbergstrasse 70 * FAX x41 61 256760 CH 4056 Basel * TEL x41 61 253880 ext 888 ************************************************************************   Received: from vmb.brl.mil by VMB.BRL.MIL id aa05565; 10 May 90 9:50 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa04910; 10 May 90 9:28 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa04798; 10 May 90 9:06 EDT Received: from ugw.utcs.utoronto.ca by VGR.BRL.MIL id aa01193; 10 May 90 8:48 EDT Received: from KRDC.INT.Alcan.CA (stdin) by ugw.utcs.utoronto.ca with BSMTP id 57475; Thu, 10 May 90 08:45:48 EDT Date: Thu, 10 May 90 08:46:00 EDT From: Shawn Allin - Alcan KRDC Computer Services Subject: VTxxx terminal emulators for 4D's To: info-iris@BRL.MIL Message-id: <98A58D45A37FA18897@KRDC.INT.Alcan.CA> X-Envelope-to: info-iris@brl.MIL X-VMS-To: IN%"info-iris@brl.mil" Date sent: 10-MAY-1990 08:37:39 Hi, I think I saw some discussion on the list earlier about this, but here goes again: Does anyone know of a VT200 or VT300-series terminal emulator that will run on a 4D-series machine? The need for the more "modern" emulation is to satisfy the requirement of displaying 8-bit characters such as French characters on the screen. We have TGRAF07 here, which as part of its emulation suite does a reasonable job of VT100, but that is not enough. Thanks in advance for any info. Cheers, Shawn Allin Alcan International Ltd., P.O. Box 8400, Kingston, Ont., Canada K7L 5L9 (613) 541-2178 ACCESS@KRDC.INT.Alcan.CA Bitnet: ACCESS@ALCANKTN   Received: from vmb.brl.mil by VMB.BRL.MIL id aa09413; 10 May 90 12:06 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa08854; 10 May 90 11:55 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa08716; 10 May 90 11:37 EDT Received: from uunet.UU.NET by VGR.BRL.MIL id aa01941; 10 May 90 10:47 EDT Received: from lsr-vax.UUCP by uunet.uu.net (5.61/1.14) with UUCP id AA05834; Thu, 10 May 90 09:59:52 -0400 Received: by (4.1/SMI-4.0) id AA25360; Thu, 10 May 90 09:14:11 EDT Date: Thu, 10 May 90 09:14:11 EDT From: "Lance M. Optican - LMO" Message-Id: <9005101314.AA25360@> To: uunet!brl.mil!info-iris@uunet.uu.net To: Info-Iris Subject: Stereographics query I have had a borrowed Stereographics liquid crystal shutter system (Crystal Eyes) on my Iris 4D120 for several months now. It works remarkably well, and is easy to program (examples of stereo perspective routines available on request). A crossed-polarizer version is also available. For details, contact the company: Stereographics Corp. 2171-H East Francisco Blvd. San Rafael, CA 94901 phone: 415-459-4500 fax: 415-459-3020 Lance M. Optican uunet!lsr-vax!lmo National Eye Institute, NIH, Bethesda, MD (301) 496-3549   Received: from vmb.brl.mil by VMB.BRL.MIL id aa10685; 10 May 90 12:58 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa08031; 10 May 90 11:32 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa07989; 10 May 90 11:12 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa01944; 10 May 90 10:48 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA03821; Thu, 10 May 90 07:34:03 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 10 May 90 07:42:38 GMT From: Ian Hoyle Organization: none Subject: Toolchests under NeWS on SGI PIs. Message-Id: <1546@merlin.bhpmrl.oz.au> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL This is posted on behalf of a friend in another division of this company ... hopefully the problem isn't too obscure ;-) =-=-=-===-=-=-=-=-=- SUBJECT - Toolchests under NeWS on SGI PIs. The problem that we are trying to overcome is to provide a two layer menu from a toolchest to control a system that we have developed on SGI PIs. The toolchest is to be called 'Pits', after the system. The initial menu - displayed when the toolchest is picked would have the following choices : Pits => Create area Grant access => Revoke access => Backup area => Delete area => where the options marked with a => can be dragged to display a submenu of available areas. The list of available areas for any user account is stored in the file ~/.pits. I have managed to construct the first menu layer successfully by adding the options to the toolchest definition script, but despite the help of the remote toolchest in 4D gifts, I cannot get the second layer of menus to work. In other words I can't seem to be able to read in from a _file_ the menu choices that I want to 'pop up'. Having spent a futile two days in cranial wallbeating, i would appreciate some assistance Mike Horton =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- -- Ian Hoyle /\/\ / / /\ BHP Melbourne Research Laboratories / / / \ 245 Wellington Rd, Mulgrave, 3170 / / / /\ \ AUSTRALIA \ \/ / / / \ / / / Phone : +61-3-560-7066 \/\/\/ ACSnet : ianh@bhpmrl.oz.au Internet: ianh%bhpmrl.oz.au@uunet.uu.net   Received: from vmb.brl.mil by VMB.BRL.MIL id aa12916; 10 May 90 15:11 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa11140; 10 May 90 13:27 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa11113; 10 May 90 13:18 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa02721; 10 May 90 12:46 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA01292; Thu, 10 May 90 09:41:38 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 10 May 90 15:41:13 GMT From: George Elkins Organization: Rutgers Univ., New Brunswick, N.J. Subject: Invert greyscale of black and white image Message-Id: Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Is there a command or simple technique to invert the greyscale of a black and white image? In the man page for tobw there is mention of an Appendix H, "Using the Image Library" of the Users' Guide. Where is this? George Elkins   Received: from vmb.brl.mil by VMB.BRL.MIL id aa17186; 10 May 90 21:21 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa16763; 10 May 90 19:38 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa16732; 10 May 90 19:28 EDT Received: from umrvma.umr.edu by VGR.BRL.MIL id aa05174; 10 May 90 19:02 EDT Received: from UMRVMA.UMR.EDU by UMRVMA.UMR.EDU (IBM VM SMTP R1.2.1MX) with BSMTP id 9330; Thu, 10 May 90 17:57:32 CDT Received: by UMRVMA (Mailer R2.07) id 9329; Thu, 10 May 90 17:57:26 CDT Date: Thu, 10 May 90 17:53:42 CDT From: Bob Funchess Subject: Key bindings To: info-iris@BRL.MIL Message-ID: <9005101902.aa05174@VGR.BRL.MIL> My wonderful tn3270 doesn't work anymore, and neither does vi... We got a new Iris, with the new (3.2.1) system software... and something weird has been done to the F1 - F4 keys, and the Page Up / Page Down keys. They used to work fine but now they seem to be internally mapped to something else. I'd like for vi and tn3270 to work on our new iris the same way they do on the old one. How can I unbreak this? I don't *care* why it was done... I'm much more likely to use vi and tn3270 than whatever nifty program is the reason for "fixing" them. < Bob | S090726@UMRVMA.UMR.EDU | Funchess > University of Missouri - Rolla "Never trust a man in a blue trench coat, Never drive a car when you're dead" -- T. Waits   Received: from vmb.brl.mil by VMB.BRL.MIL id aa17332; 10 May 90 22:13 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa17057; 10 May 90 21:00 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa17039; 10 May 90 20:44 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa05510; 10 May 90 20:20 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA07596; Thu, 10 May 90 17:15:31 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 10 May 90 21:14:54 GMT From: "Prof. Steven H. Izen" Organization: NASA Ames Research Center, Moffett Field, CA Subject: Re: dvipage on Silicon Graphics Message-Id: <6082@amelia.nas.nasa.gov> References: <3242@usceast.UUCP> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <3242@usceast.UUCP> sharp@usceast.UUCP (Harry Sharp) writes: >We have brought TeX up on the Irises and would like to be able to >preview files. xdvi does not work on these machines. The systems >group at SGI doesn't quite know what to make of this and we have >given up on this approach. I raqn into the same problem on my 4D/25. I dug through the code and found that for some unknown reason the server was barfing on an Xputimage call. After wasting countless hours on this, I switched to dvix, which works fine. It seems to be a more primitive version of the same program which does not use the X toolkit. I occasionally have trouble telling it where to find finds, but for the most part it works well. I forgot where I got it dvix, but it was some anonymous ftp site. -- Steve Izen: {sun,uunet}!cwjcc!skybridge!izen386!steve / Quote corner: or steve@izen386.math.cwru.edu / or izen@cwru.cwru.edu /-------------------------/ My second bike is a car. | Klein bottle for sale - Inquire within.   Received: from vmb.brl.mil by VMB.BRL.MIL id aa19596; 11 May 90 7:20 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa19449; 11 May 90 7:09 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa19346; 11 May 90 6:45 EDT Received: from funet.fi by VGR.BRL.MIL id aa06954; 11 May 90 5:37 EDT Received: by funet.fi; id AA27063; Fri, 11 May 90 12:22:36 +0300 Received: from geeni.bio.vtt.fi by router.funet.fi; Fri, 11 May 90 10:50 EET Received: by geeni.bio.vtt.fi (5.52/1.1.geeni) id AA06465; Fri, 11 May 90 10:49:47 GMT Date: Fri, 11 May 90 10:49:47 GMT From: laaksone@bio.vtt.fi To: info-iris@BRL.MIL Message-Id: <9005111049.AA06465@geeni.bio.vtt.fi> We are running a SGI 4D/70GTB with a lot of problems. Most of the problems disappeared when we (finally) got IRIX 3.2.2 Now we have a new problem. The service personnel installed a new Fujitsu M2263E disk (ordered in August 89 arrived May 90, really fast service from SGI). There is even the option for that disk in fx. The installation went nice apart from one problem: all disk io to that disk is now about 20-30 slower than that to the system disk. There are also no messages in the SYSLOG. Is there something special I should know about when I install a Fujitsu M2263E? All help is appreciated. This battle with the Fujitsu disk has gone on for a week now and I would be happy to do some real work again. -leif laaksonen -- **************************************************************************** Leif Laaksonen I "Learn from yesterday, Phone: 358-0-4565105 I live for today, Fax: 358-0-4552028 I look to tomorrow, I rest this afternoon." Mail: laaksone@geeni.bio.vtt.fi I -- Snoopy laaksone@finfun I ****************************************************************************   Received: from vmb.brl.mil by VMB.BRL.MIL id ab19596; 11 May 90 7:20 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab19449; 11 May 90 7:10 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa19353; 11 May 90 6:46 EDT Received: from funet.fi by VGR.BRL.MIL id aa06956; 11 May 90 5:38 EDT Received: by funet.fi; id AA27079; Fri, 11 May 90 12:22:44 +0300 Received: from geeni.bio.vtt.fi by router.funet.fi; Fri, 11 May 90 10:56 EET Received: by geeni.bio.vtt.fi (5.52/1.1.geeni) id AA06476; Fri, 11 May 90 10:55:25 GMT Date: Fri, 11 May 90 10:55:25 GMT From: laaksone@geeni.bio.vtt.fi To: info-iris@BRL.MIL Message-Id: <9005111055.AA06476@geeni.bio.vtt.fi> We are running a SGI 4D/70GTB with a lot of problems. Most of the problems disappeared when we (finally) got IRIX 3.2.2 Now we have a new problem. The service personnel installed a new Fujitsu M2263E disk (ordered in August 89 arrived May 90, really fast service from SGI). There is even the option for that disk in fx. The installation went nice apart from one problem: all disk io to that disk is now about 20-30 times slower than that to the system disk. There are also no messages in the SYSLOG. Is there something special I should know about when I install a Fujitsu M2263E? All help is appreciated. This battle with the Fujitsu disk has gone on for a week now and I would be happy to do some real work again. -leif laaksonen -- **************************************************************************** Leif Laaksonen I "Learn from yesterday, Phone: 358-0-4565105 I live for today, Fax: 358-0-4552028 I look to tomorrow, I rest this afternoon." Mail: laaksone@geeni.bio.vtt.fi I -- Snoopy laaksone@finfun I ****************************************************************************   Received: from vmb.brl.mil by VMB.BRL.MIL id aa20531; 11 May 90 8:24 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa19110; 11 May 90 6:12 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa19085; 11 May 90 5:48 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa06882; 11 May 90 5:06 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA07858; Fri, 11 May 90 02:02:54 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 11 May 90 07:57:42 GMT From: Bruno Pape Organization: Silicon Graphics S.A., Zuerich, Switzerland Subject: IMSL Substitutes. Message-Id: <1990May11.075742.12243@sgzh.uucp> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Thank you! To everyone that replied with ideas for IMSL substitues. One idea that was passed through net-mail was some software distributed by: A.H. Morris Engineering & Information Systems Dept. Naval Surface Warfare Center Dahlgren, VA 22448-5000 USA from lang.fortran ------------------------------------------------------ I have just received the NSWC (Naval Surface Warfare Center) mathematics subroutine library. It is comparable in size and coverage to IMSL and NAG, and is distributed free of charge. It appears to have a slightly wider coverage in some areas than either NAG or IMSL, for instance, it includes linear programming. There are 878 routines in the library, which is distributed as Fortran source. The quality appears to be at least as high as that of NAG and IMSL. The manual - 3cm. thick, says: "Approved for public release; distribution unlimited." The first BAT file on floppy disk says: "echo You may copy this disk freely, without permission or payment." The code is available on 7-track and 9-track tapes, and compacted using LHARC on 5.25 inch, 1.2Mb floppy disk, self uncompacting. The reference for the manual is: NSWC TR 90-21. The source files on the disk were dated 8 March 1990. ------------------------------------------------------------------------- Thanks, Bruno   Received: from vmb.brl.mil by VMB.BRL.MIL id ab20531; 11 May 90 8:24 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa19817; 11 May 90 8:13 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa19784; 11 May 90 7:46 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa07374; 11 May 90 7:07 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA13746; Fri, 11 May 90 04:01:23 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 11 May 90 08:46:46 GMT From: Bruno Pape Organization: Silicon Graphics S.A., Zuerich, Switzerland Subject: Backing up remote machines? Message-Id: <1990May11.084646.12419@sgzh.uucp> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Objective: To back up remote machines from a central host. Constraints: Must be simple, operator is someone who changes tapes for a living. Failed attemps: Mount remote file systems and run bru localy: NFS hangs, times out, or becomes infinitly slow. 780MB disks, heavy traffic on local segment. rsh to remote machine and run bru and pipe to central host: Works good unless bru prompts for multiple tape volumes, prompt is not sent to window on central host if run from a script. The only thing that I have found that works is to rlogin to each remote machine and run bru across the net. This is deemed to complicated for the person whos primary job is to change tapes. Thank for any help, Bruno Pape That's bru no. Can I say "Fixed in the next release."?   Received: from vmb.brl.mil by VMB.BRL.MIL id ab23397; 11 May 90 9:43 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa22046; 11 May 90 9:14 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa21277; 11 May 90 8:47 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa07785; 11 May 90 8:05 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA16407; Fri, 11 May 90 05:01:07 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 11 May 90 08:42:38 GMT From: eru!luth!sunic!mcsun!cernvax!chx400!urz.unibas.ch!doelz@bloom-beacon.mit.edu Organization: University of Basel, Switzerland Subject: set autologout sysnatx wanted Message-Id: <1990May11.094238.723@urz.unibas.ch> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL I need to kick out some usere who forget to log out notoriously. I thought I'd put set autologout=10min in the /etc/cshrc but this doesn't work. Any ideas ? Pointers or suggestions appreciated. - Reinhard   Received: from vmb.brl.mil by VMB.BRL.MIL id aa24118; 11 May 90 9:54 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa23397; 11 May 90 9:43 EDT Date: Fri, 11 May 90 9:06:35 EDT From: Gary S. Moss (VLD/VMB) To: Ian Hoyle cc: info-iris@BRL.MIL Subject: Re: Toolchests under NeWS on SGI PIs. Message-ID: <9005110906.aa22435@VMB.BRL.MIL> [Mike Horton writes:] < The problem that we are trying to overcome is to provide a < two layer menu from a toolchest to control a system that we < have developed on SGI PIs. Mike, I wrote some code which reads your .rhosts file and based on comments in the file, creates a 2-level menu hierarchy under a toolchest. This was inspired by the remote toolchest in 4D gifts. Hopefully it will give you some clues. ----------------- rlogin.ps ------ % This file defines a procedure "makeHostsChest" which builds a tool chest % called "Hosts" containing menus, one for each category, as described in your % .rhosts file. The .rhosts file must contain comments (beginning with '#') % which partition off the host names into submenus. Whatever appears after % the '#' will be used as the title for the submenu. Leading spaces or tabs % (between the '#' and the label) will be ignored, but do not mix spaces and % tabs. Blank lines in the .rhosts file are ignored. Below is an example % .rhosts file that partitions menus by machine type: % % # Suns % junior.brl.mil % johnboy.brl.mil % % # Alliants % bigfx.brl.mil % % # Iris 4Ds % server1.brl.mil % grafix1.brl.mil % grafix2.brl.mil % This file can be loaded as follows: % (rlogin.ps) run % Or, if it lives in your ~/.4sight directory: % (rlogin.ps)(.4sight/) exch append run % The "Hosts" tool chest can then be built like this: % makeHostsChest % Or, ideally, run it from your "user.ps" file added to "RestartActions" % as follows: % /RestartActions % [ % ... other stuff here % { makeHostsChest } % ] store % Default "wsh" font. To override this, define /DefaultWSHFont somewhere % or comment the following out. systemdict /DefaultWSHFont known not { /DefaultWSHFont (Iris12.fm) def } if % % This is the argument template for the remote wsh to be forked % /RemoteWshTemplate { (wsh -t % -n % -f) DefaultWSHFont append ( -r2000 -s80,60 -m80,60 -C 0,7,4,1 -c rlogin %) append sprintf } def /printStack { (->) print pstack (\n) print } def % This procedure fills in RemoteWshTemplate and executes the rlogin command /rlogin % usage: hostname rlogin -> - { currentkey dup (host name: '%'\n) printf % get rid of all but first part of host name dup (.brl.mil) search { % true ->(host.brl.mil)()(.brl.mil)(host) exch pop % discard match ->(host.brl.mil)()(host) exch pop % discard post ->(host.brl.mil)(host) } if dup % ->(host.brl.mil)(host)(host) RemoteWshTemplate dup (command: '%'\n) printf forkunix % -> wshcommand forkunix } def /outputMenu % usage: [(item1)(item2)...] outputMenu -> - { currentdict /machine get % [...] (machine) currentdict /menus get currentdict /menucount get 1 sub 2 dict put % [...] (machine) dict[...] currentdict /menus get % [...] (machine) [dict[]...] currentdict /menucount get 1 sub get exch /title exch put % [...] currentdict /menus get currentdict /menucount get 1 sub get /menuitems % [...] dict[...] /menuitems 3 1 roll 3 1 roll [ { rlogin } ] % [...] dict[...] /menuitems (machine) [{rlogin}] /new DefaultMenu send put % - currentdict /menus get currentdict /menucount get 1 sub get /title get % (machine) /settitle % (machine) /settitle currentdict /menus get currentdict /menucount get 1 sub get /menuitems get send } def % This routine removes one or more instances of (pattern) if they occur % at the beginning of (string). /trimLeading % usage: (string) (pattern) -> (trimmed_string) { { search % find leading pattern { % -> post match pre dup length 0 eq % is it a leading character ? { % -> post match pre pop % -> post match } { % imbedded character, restore it 3 1 roll % pre post match exch % pre match post append append exit } ifelse } { % no match found exit } ifelse } loop } def % This routine removes leading spaces or tabs. NOTE: if they are mixed, % and spaces do not preceed all tabs, this won't do the job. /trimLeadingWhiteSpace % usage: (string) -> (string) { ( ) trimLeading % remove leading spaces (\t) trimLeading % remove leading tabs } def % This procedure opens ~/.rhosts for reading and pushs each line on the stack /readHostFile % usage: file readHostfile -> [(line1)(line2)(line3)...] { { (r) file } stopped { % (~/.rhosts)(r)file [ ] % if false do nothing } { % else -> file_object_opened_for_reading 5 dict begin % dictionary stack -> [] /menus 20 array def /remote exch def % /remote file def -> - % dictionary stack -> [remote:file] /count 0 def % dictionary stack -> % [remote:file,count:0] /menucount 0 def { remote 80 string % -> remote string[80] readline % next line from file read ->...(n-1)(n) not % test for EOF condition { % EOF encountered pop % discard empty string exit % done reading file } if % check for blank line dup length 0 eq { pop % discard empty string } { (#) search { % -> post (#) pre pop pop % -> post == trimLeadingWhiteSpace currentdict /count get dup 0 gt { 2 add 1 roll ] outputMenu } { pop } ifelse currentdict /menucount get 1 add /menucount exch def /count 0 def /machine exch def % dictionary stack -> % [remote:file,count:0,machine:name] [ } { % discard user name at end of line ( ) search { % post ( ) pre exch pop exch pop } if currentdict /count get 1 add /count exch def } ifelse } ifelse } loop ] outputMenu currentdict /HostsChest [ currentdict /menus get { dup null eq { pop exit } if dup [ exch % [dict] marker [dict] /title get % [dict] marker (machine) pullRightIcon ] % [dict] [(machine) pRI] exch /menuitems get % [(machine) pRI] menu } forall ] (Hosts) /new ToolChest send end % pop dictionary off current stack } ifelse } def /makeHostsChest { (HOME) getenv (/.rhosts) append % -> ~/.rhosts readHostFile } def % If you want the "Hosts" tool chest invoked from here rather than from % your user.ps, uncomment the next line. The ideal place is from % /RestartActions defined in your user.ps. % makeHostsChest   Received: from vmb.brl.mil by VMB.BRL.MIL id aa25294; 11 May 90 10:23 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ac24118; 11 May 90 10:12 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa23894; 11 May 90 9:47 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa08060; 11 May 90 9:08 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA19414; Fri, 11 May 90 06:02:24 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 11 May 90 11:33:10 GMT From: Bruno Pape Organization: Silicon Graphics S.A., Zuerich, Switzerland Subject: IBM 8 bit ascii char set? Message-Id: <1990May11.113310.12645@sgzh.uucp> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Does anyone have a definition of the IBM standard 8 bit ASCII character set for the font manager? If not how about some tools to define one with? Or any possibilities for a combination of form definition and upper and lower case characters? Thanks again, Bruno   Received: from vmb.brl.mil by VMB.BRL.MIL id aa26594; 11 May 90 11:51 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa26220; 11 May 90 11:40 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa26210; 11 May 90 11:22 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa08584; 11 May 90 10:36 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA24250; Fri, 11 May 90 07:24:52 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 11 May 90 10:12:24 GMT From: eagle.wesleyan.edu!gravishanker@cs.yale.edu Subject: SGI image library files Message-Id: <22374@eagle.wesleyan.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL I am a relatively new reader of this newsgroup. I have been using Silicon Graphics Iris machines for the past year. I have some general questions and would appreciate any help. 1. How do people get the good looking raster images into the system for view by ipaste? Are they directly fed in from some video recorder? What is exactly involved in getting these images into the machine as well as record images generated on the machine by some program onto a video tape? How much would such a setup cost? 2. I am interested in getting some of these image files that can be used in place of the user icons when the visual login procedure pandora starts up under version 3.2 of the OS. I have picked up some of the GIF files converted them to SGI imagelib images, but most of them are bigger than the maximum allowed (100 pixels by 100 pixels) allowed by pandora. 3. I struggled a lot to start the NeWS server succesfully when named is running under version 3.2 of the OS. Since the init.ps (in /usr/NeWS/lib/NeWS) has changed between version 3.1x and 3.2 several problems crop up. I finally solved the problem and if anyone is interested in hearing about it please let me know. Thanks for any help anyone can provide. G. Ravishanker Manager of Computer Graphics Wesleyan University Middletown, CT.   Received: from vmb.brl.mil by VMB.BRL.MIL id aa27061; 11 May 90 12:36 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa26909; 11 May 90 12:26 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa26787; 11 May 90 12:04 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa08735; 11 May 90 11:06 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA26389; Fri, 11 May 90 08:02:50 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 11 May 90 13:23:04 GMT From: mcsun!cernvax!chx400!urz.unibas.ch!doelz@uunet.uu.net Organization: University of Basel, Switzerland Subject: Re: Backing up remote machines? Next release ... ? Message-Id: <1990May11.142304.724@urz.unibas.ch> References: <1990May11.084646.12419@sgzh.uucp> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <1990May11.084646.12419@sgzh.uucp>, root@sgzh.uucp (Bruno Pape) writes: > rsh to remote machine and run bru and pipe to central host: > Works good unless bru prompts for multiple tape volumes, > prompt is not sent to window on central host if run from > a script. My convex (with a 'real' tape drive) doesn't know bru, and my SGI (with a streamer) doesn't know rdump, and my data are not backed up very well. Surely something to be improved ! Herewith I vote for rdump in the next release of IRIX. - Reinhard   Received: from vmb.brl.mil by VMB.brl.MIL id ab01751; 11 May 90 15:30 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa01246; 11 May 90 15:19 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa01005; 11 May 90 14:49 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa10220; 11 May 90 13:24 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA04556; Fri, 11 May 90 10:14:27 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 11 May 90 13:17:10 GMT From: Jeff Hanson Organization: NASA/Lewis Research Center, Cleveland Subject: More on how to use netlib Message-Id: <1990May11.131710.21170@eagle.lerc.nasa.gov> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL ===== How to use netlib ===== mail netlib@research.att.com send index The Internet address "netlib@research.att.com" refers to a gateway machine at AT&T Bell Labs in Murray Hill, New Jersey. This address should be understood on all the major networks. For systems having only uucp connections, use the address uunet!research!netlib. In this case, someone will be paying for long distance 1200 baud phone calls, so keep your requests to a reasonable size! For access from Europe, try the duplicate collection in Oslo: Internet: netlib@nac.no EARN/BITNET: netlib%nac.no@norunix.bitnet X.400: s=netlib; o=nac; c=no; EUNET/uucp: nac!netlib Background about netlib is in Jack J. Dongarra and Eric Grosse, Distribution of Mathematical Software Via Electronic Mail, Comm. ACM (1987) 30,403-407. -- *-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-* \ / \ / \ / \ / \ / \ / Jeff Hanson \ / \ / \ / \ / \ / \ / * ViSC: Better * tohanson@gonzo.lerc.nasa.gov * * * * * * / \ / \ Science / \ / \ NASA Lewis Research Center / \ / \ Through / \ / \ * * * * * * * Cleveland, Ohio 44135 * * * Pictures * * \ / \ / \ / \ Telephone - (216) 433-2284 Fax - (216) 433-2182 \ / \ / \ / *-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*   Received: from vmb.brl.mil by VMB.BRL.MIL id aa03482; 11 May 90 17:45 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa03344; 11 May 90 17:35 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa03323; 11 May 90 17:22 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa13863; 11 May 90 16:24 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA14950; Fri, 11 May 90 13:08:43 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 11 May 90 19:09:49 GMT From: Ted Wilcox Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Re: Toolchests under NeWS on SGI PIs. Message-Id: <7603@odin.corp.sgi.com> References: <1546@merlin.bhpmrl.oz.au> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <1546@merlin.bhpmrl.oz.au> ianh@bhpmrl.oz.au (Ian Hoyle) writes: >SUBJECT - Toolchests under NeWS on SGI PIs. > >The problem that we are trying to overcome is to provide a >two layer menu from a toolchest to control a system that we >have developed on SGI PIs. > >The toolchest is to be called 'Pits', after the system. . . . >I have managed to construct the first menu layer successfully by adding >the options to the toolchest definition script, but despite >the help of the remote toolchest in 4D gifts, I cannot get the second layer >of menus to work. In other words I can't seem to be able to read in >from a _file_ the menu choices that I want to 'pop up'. > >Having spent a futile two days in cranial wallbeating, i would appreciate >some assistance > >Mike Horton > Instead of starting with the 4Dgifts code, you can start with the code that actually starts up the "Demos" chest. If you look at "/usr/sbin/demochest", you will find a postscript executable that does just what you want. Hope this helps. ------------------------------------------------------------------------ |Ted. | "Reality is a horde of mice nibbling away in | |ted@sgi.com | the basement of your dreams." | | | -Bruce Sterling |   Received: from vmb.brl.mil by VMB.BRL.MIL id aa04214; 11 May 90 19:51 EDT Received: from vmb.brl.mil by VMB.brl.MIL id ab03832; 11 May 90 18:28 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab03807; 11 May 90 18:12 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa14194; 11 May 90 17:09 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA17530; Fri, 11 May 90 13:53:14 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 11 May 90 20:14:01 GMT From: Mark Bradley Organization: Solbourne Computer Systems Subject: Re: (none) Message-Id: <1990May11.201401.1159@Solbourne.COM> References: <9005111055.AA06476@geeni.bio.vtt.fi> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Leif Writes: >The service personnel installed a new Fujitsu >M2263E disk (ordered in August 89 arrived May 90, >really fast service from SGI). There is even the option for that >disk in fx. The installation went nice apart from one problem: >all disk io to that disk is now about 20-30 times slower than that to >the system disk. There are also no messages in the SYSLOG. > >Is there something special I should know about when I install a >Fujitsu M2263E? > >All help is appreciated. This battle with the Fujitsu disk has gone >on for a week now and I would be happy to do some real work again. When I was at SGI, some number of these drives were brought in to mfg. for limited shipment. The drive should not be 20-30 *times* slower, but more like 20-30 *percent* slower in some applications. I sincerely hope this is what you mean. If not, some other problems may exist on your system. fx entries are sometimes there for the convenience of those individuals that feel comfortable putting their own drives on the SGI box and to provide some flexibility as to what manufacturers disk drives can ship, not to make any statement as to which drives are the preferred OEM disk drive. As I said this was for only limited shipments, and to the best of my knowledge, will not be used as a standard OEM product for any long period of time. markb -- Mark Bradley Faster, faster, until the thrill I/O Subsystems of speed overcomes the fear of death. Solbourne Computer, Inc. --Hunter S. Thompson   Received: from vmb.brl.mil by VMB.BRL.MIL id aa04686; 11 May 90 21:46 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa04617; 11 May 90 21:36 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa04566; 11 May 90 21:22 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa15137; 11 May 90 20:22 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA29851; Fri, 11 May 90 17:15:43 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 11 May 90 19:12:41 GMT From: "James P. Loan" Organization: Computer Science Department, Stanford University Subject: Window Manager Bug? Message-Id: <1990May11.191241.8140@Neon.Stanford.EDU> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL I'm trying to use the DEPTHCHANGE event on the gl windows in my application. According to the manual, the value of the event is the gid of the window whose depth has been changed. All I ever get for a value is 0, making it impossible to track the depthchange of my windows. Is this a bug or am I doing something wrong? In case it matters, I'm running 3.2.1 on a 4D/25. Thanks for any help, pete loan loan@neon.stanford.edu   Received: from vmb.brl.mil by VMB.BRL.MIL id ab04686; 11 May 90 21:46 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab04617; 11 May 90 21:36 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab04566; 11 May 90 21:22 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa15140; 11 May 90 20:23 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA29369; Fri, 11 May 90 17:09:19 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 11 May 90 22:33:44 GMT From: dave "who can do? ratmandu!" ratcliffe Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Re: Invert greyscale of black and white image Message-Id: <7617@odin.corp.sgi.com> References: Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article elkins@topaz.rutgers.edu (George Elkins) writes: > >Is there a command or simple technique to invert the greyscale of a >black and white image? > >In the man page for tobw there is mention of an Appendix H, "Using the >Image Library" of the Users' Guide. Where is this? unfortunately the Appendix H "went away" a while back, but it didn't have anything that would've helped you here. you want to check out /usr/people/4Dgifts/iristools/imgtools/invert.c if you don't have /usr/people/4Dgifts loaded, scan yer latest release notes doc. for the section on in and load the "dev.sw.giftssrc" subsystem from yer latest "DEV" tape. in this way you'll have the source and can generate lots of image processing programs including invert which: inverts an image. pixel value 0 becomes 255 and pixel value 255 maps to 0. The result is to invert the tonal scale of an image. OR you can get yer latest eoe2 tape back out, run inst and get into manual mode with it and load "eoe2.sw.moregltools" (check to make sure inst knows about that subsystem on the tape), and it will put all the executable versions of these progs into /usr/sbin. -- daveus rattus yer friendly neighborhood ratman KOYAANISQATSI ko.yan.nis.qatsi (from the Hopi Language) n. 1. crazy life. 2. life in turmoil. 3. life out of balance. 4. life disintegrating. 5. a state of life that calls for another way of living.   Received: from vmb.brl.mil by VMB.BRL.MIL id aa10242; 18 May 90 9:02 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa08870; 18 May 90 8:13 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa08729; 18 May 90 7:55 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa03197; 18 May 90 7:45 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA22349; Fri, 18 May 90 04:30:46 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 12 May 90 01:20:23 GMT From: David Hinds Organization: AIR, Stanford University Subject: Catching floating point errors in FORTRAN ?!? Message-Id: <1990May12.012023.18420@portia.Stanford.EDU> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL I've had no luck trying to figure out how to detect a "not a number" error in MIPS FORTRAN. It must be in the manuals SOMEWHERE, but, of course, they are hopeless. I want to be able to test whether a real variable contains an "NaN" value. -David Hinds dhinds@popserver.stanford.edu   Received: from vmb.brl.mil by VMB.BRL.MIL id aa08776; 12 May 90 14:41 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa08494; 12 May 90 13:39 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa08467; 12 May 90 13:19 EDT Received: from cunyvm.cuny.edu by VGR.BRL.MIL id aa17899; 12 May 90 9:41 EDT Received: from MCCLB0.MED.NYU.EDU by CUNYVM.CUNY.EDU (IBM VM SMTP R1.2.2MX) with BSMTP id 5649; Sat, 12 May 90 09:38:50 EDT Received: from mcirps2.med.nyu.edu by MCCLB0.MED.NYU.EDU; Sat, 12 May 90 09:40 EDT Received: by mcirps2.med.nyu.edu (5.52/890607.SGI) (for @mcclb0.med.nyu.edu:info-iris@brl.arpa) id AA00638; Sat, 12 May 90 10:04:55 DSD Date: Sat, 12 May 90 10:04:55 DSD From: karron%mcirps2.med.nyu.edu@cunyvm.cuny.edu Subject: info-iris postings and news postings To: info-iris@BRL.MIL Message-id: <9005121704.AA00638@mcirps2.med.nyu.edu> X-Envelope-to: info-iris@brl.arpa Am I seeing things, or is mail to info-iris also getting posted on notes in comp.sys.iris (or somthing like that) ?. I noticed that some reply items were ONLY reply posted to notes. Since I do not have regular access to notes, I missed them. Is there a notes that runs on iris's ?, should I install it ?. What is the relationship between notes and info-iris mail. dan. . -- +-----------------------------------------------------------------------------+ | karron@nyu.edu Dan Karron | | . . . . . . . . . . . . . . New York University Medical Center | | 560 First Avenue \ \ Pager <1> (212) 397 9330 | | New York, New York 10016 \**\ <2> 10896 <3> | | (212) 340 5210 \**\__________________________________________ | +-----------------------------------------------------------------------------+   Received: from vmb.brl.mil by VMB.BRL.MIL id aa10297; 12 May 90 23:02 EDT Received: from vmb.brl.mil by VMB.brl.MIL id aa10188; 12 May 90 22:10 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa10184; 12 May 90 21:56 EDT Received: from cs.utah.edu by VGR.BRL.MIL id aa20047; 12 May 90 21:44 EDT Received: from adenosine.pharm.utah.edu by cs.utah.edu (5.61/utah-2.11-cs) id AA05698; Sat, 12 May 90 19:42:56 -0600 Date: Sat, 12 May 90 19:48:55 MDT From: "Darrell R. Davis" Posted-Date: Sat, 12 May 90 19:48:55 MDT Message-Id: <9005130148.AA03801@adenosine.pharm.utah.edu> Received: by adenosine.pharm.utah.edu (5.52/5.51) id AA03801; Sat, 12 May 90 19:48:55 MDT To: info-iris@BRL.MIL In-Reply-To: mcsun!cernvax!chx400!urz.unibas.ch!doelz@uunet.uu.net's message of 11 May 90 13:23:04 GMT <1990May11.142304.724@urz.unibas.ch> Subject: Backing up, possible solution. > rsh to remote machine and run bru and pipe to central host: mcsun> My convex (with a 'real' tape drive) doesn't know bru, and my SGI mcsun> (with a streamer) doesn't know rdump, and my data mcsun>are not backed up very well. Surely something to be improved ! I have been using gnu-tar (gtar) for a couple of months to do backups on my PI. I haven't been running it remotely, but the machine has about 1.1 gbytes of disk plus another 300 meg file system nfs mounted from a Sparcstation. I don't usually backup the whole works because a large portion are data files that are routinely archived by users. The standard *dump* is 2-3 tapes full. Some of what I think are major advantages to gtar over 'regular' tar and bru are: 1- ability to make multi-volume archives. Obviously first on the list. 2- compatibility with primitive tar, that is, gtar can read tar files. I'm pretty sure that in the default mode (no multi-volumes or other fancy stuff.) tar and gtar are completely compatible. 3- portability, you could put it on your convex, suns, iris, etc. 4- you can specify filesystems *not* to include in a full dump, this is particulary nice for server types with large storage. 5- gtar comes with a nice manual written in TeX. As with most (all?) fine gnu products, this is available by anonymous ftp from prep.ai.mit.edu. Although I don't recollect, you probably have to at least put "-I/usr/include/bsd" in your makefile, plus some other minor :-) system dependent flags. Good Luck, -------------------------------------------------------------- * Darrell R. Davis * * * "Faster, faster, until the Assistant Professor * * * thrill of speed overcomes Medicinal Chemistry *A**L**T**A* the fear of death." University of Utah * * * * * * --H.S. Thompson * --------------------------------------------------------------   Received: from vmb.brl.mil by VMB.BRL.MIL id aa13123; 13 May 90 15:30 EDT Received: from vmb.brl.mil by VMB.brl.MIL id aa12938; 13 May 90 14:28 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa12934; 13 May 90 14:15 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa22673; 13 May 90 13:54 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA28260; Sun, 13 May 90 10:44:20 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 13 May 90 17:44:47 GMT From: Jean Starkey Organization: Montana State University, Dept. of Microbiology, Bozeman MT 59717 Subject: Help on printers, disks, memory Message-Id: <1872@dali> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL We are looking at various ways of getting convenient color hardcopy from a 4D/25TG SGI, and would very much appreciate hearing from people who have used color (postscript?) printers with these systems. Any help with recommendations for third party disk and memory would also be very helpful. Please e-mail responses to: umbjs@terra.oscs.montana.edu or umbjs@mtsunix1.bitnet Thanks in advance Jean Starkey   Received: from vmb.brl.mil by VMB.BRL.MIL id aa14172; 13 May 90 19:06 EDT Received: from vmb.brl.mil by VMB.brl.MIL id aa13779; 13 May 90 18:14 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa13649; 13 May 90 17:58 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa23286; 13 May 90 17:24 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA09421; Sun, 13 May 90 14:13:19 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 13 May 90 21:04:32 GMT From: Michael Gleicher Organization: School of Computer Science, Carnegie Mellon University Subject: User Interface Toolkits Message-Id: Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL I was wondering what user interface tools people were using on Irises. (in C++) The ones that I know about: NASA panels Interviews (although I haven't been able to bring it up on the iris) The X toolkits (are there C++ hooks to any of them?) are there any others? As far as Interviews goes, has anyone brought it up under CFront (2.0 preferably, or 1.1)? If a program uses Interviews (or X in general), can it still have GL windows also? I appreciate any help. Mike -- Michael Lee Gleicher gleicher@cs.cmu.edu ski bum, graphics hacker, and SCS PhD student Carnegie Mellon University, Pittsburgh, PA 15213-3890 (412) CMU-STYX (school)   Received: from vmb.brl.mil by VMB.BRL.MIL id aa14512; 13 May 90 20:04 EDT Received: from vmb.brl.mil by VMB.brl.MIL id ab14172; 13 May 90 19:11 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa14163; 13 May 90 19:06 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa23837; 13 May 90 18:55 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA14144; Sun, 13 May 90 15:41:52 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 13 May 90 21:58:58 GMT From: "Scott R. Presnell" Subject: select()/accept() and sin_family Message-Id: Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Hey, I'm experiencing a problem with checking the data of an Internet host connection that I get from a select()/accept() call. System: IRIX 3.2.1 Machine: 4D/20G Code compiled with debugging options only (no optimization). The story goes like this. The program is a daemon that forks off and listens to requests on specific ports for requests to service. So in the initialization of the daemon, I make calls thus: socket(); bind(); listen(); without error. During the normal course of operation, the program select()s and accept()s incoming connections. If the connection comes in on the port defined for connections from other hosts, the program makes an attempt to verify the address and name of the caller in the function chkhost(). Normally, sin_family == AF_INET (as it should), and chkhost() can determine the host name from the address using gethostbyname(). But sometimes, the sin_family structure entry contains the value 0 (AF_UNSPEC) and the call to gethostbyname() fails (apparent address 0.0.0.0). Under normal operation the failure rate is about 1 in 100 - with lots of requests to the daemon the failure rate goes to maybe 1 in 20 (I'm guessing in a big way here). My question is: under what set of circumstances is it reasonable to see the value AF_UNSPEC for sin_family? Alternatively - is this a bug in select() or accept() or (more likely) is my code wrong? - Code fragments follow - struct sockaddr_in slo, fromlocal; struct sockaddr_in sin, frominet; fd_set defreadfds; int fromlen; int flocal, finet; /* socket file descriptors */ int nfds, s; /* While we're waiting for the alarm to go off... */ while (!alarmed) { FD_ZERO(&defreadfds); FD_SET(flocal, &defreadfds); FD_SET(finet, &defreadfds); if ((nfds = select(20, &defreadfds, 0, 0, 0)) <= 0) if (nfds < 0 && errno != EINTR) { warn(eslct); } else { continue; } if (FD_ISSET(flocal, &defreadfds)) { fromlen = sizeof(fromlocal); s = accept(flocal, &fromlocal, &fromlen); } else if (FD_ISSET(finet, &defreadfds)) { fromlen = sizeof(frominet); s = accept(finet, &frominet, &fromlen); } else { warn("null select? (0x%x)", nfds); continue; } if (s < 0) if (errno != EINTR) { warn(eacpt); } else { continue; } if ((pid = fork()) == 0) { (void) close(flocal); (void) close(finet); if (dup2(s, 1) < 0) warn("can't dup2() for ipc request"); (void) close(s); if (FD_ISSET(finet, &defreadfds)) chkhost(&frominet); } } chkhost(f) register struct sockaddr_in *f; { if (f->sin_family != AF_INET) warn("connection not AF_INET? (%d)", f->sin_family); if ((hp = gethostbyaddr(&f->sin_addr, sizeof(struct in_addr), f->sin_family)) == NULL) fatal("Host name for your address (%s) unknown", inet_ntoa(f->sin_addr)); } - end code - Any clues at all? Thanks for your help. - Scott -- Scott Presnell +1 (415) 476-9890 Pharm. Chem., S-926 Internet: srp@cgl.ucsf.edu University of California UUCP: ...ucbvax!ucsfcgl!srp San Francisco, CA. 94143-0446 Bitnet: srp@ucsfcgl.bitnet   Received: from vmb.brl.mil by VMB.BRL.MIL id aa22807; 14 May 90 9:41 EDT Received: from vmb.brl.mil by VMB.brl.MIL id aa20359; 14 May 90 8:48 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa20262; 14 May 90 8:31 EDT Received: from aero4.larc.nasa.gov by VGR.BRL.MIL id aa27293; 14 May 90 7:55 EDT Received: Mon, 14 May 90 07:55:13 EDT by aero4.larc.nasa.gov (5.52/5.6) Date: Mon, 14 May 90 07:55:13 EDT From: "Brent L. Bates AAD/TAB MS361 x42854" Message-Id: <9005141155.AA08313@aero4.larc.nasa.gov> To: umbjs@terra.oscs.montana.edu Subject: Re: Help on printers, disks, memory Cc: info-iris@BRL.MIL There are usually two basic routes to get hard copy. 1) Video interface. They hook up to your monitor video and what ever is on the screen is what you get. These systems are also usually limited to about 4k colors on the copy. 2) A parallel/serial interface. Files are created and sent to the printer like any other type of printer. If you have the right software you can capture portions of the screen, perform transformations on the image and then send it to the printer. You also don't have a 4k limit on colors. We bought a Tektronix 4693D unit and a parallel interface for our 3130. Most of the software we use has the ability to save a run length encoded image file and with software I wrote, we convert the images into the Tektronix format. We are very pleased with the results. Last year at SIGGRAPH '89, I was more impressed with the Kodak printer. If we were buying a unit now I would seriously consider the Kodak printer, the results looked better and were far more durable. Almost all of the color printers use thermal wax transfer and viewgraphs have a tendency to melt if the image is too dark. The Kodak unit didn't seem to have that problem. As far as memory is concered... Here is a list I have compiled from info-iris mail. All comments are from the original poster. I don't know how good or bad any of them are we haven't ordered anything yet. Sophisticated Circuits 19017 120th Ave N.E. Bothel, WA 98011 (206) 485-7172. Memory for 4D 20's, 1Mb simms, $125/Mb (January '90) 4MB chips also available. About 3day delivery. "...about the 4th time I have bought memory from them. Last year I got a batch of bad chips and they promptly replaced them with no hassle. The original memory I bought is still working fine, just about 2 years." Impediment, Inc. 333 Duxbury, MA 02332 (617) 837-8877, Alex Sunguroff (car phone): (617) 694-4488 Memory for 4D 20's, 1Mb simms, $90/Mb (early January '90) Memory for 4D 25, 4Mb simms 80ns, $525/4Mb (early March '90) Memory for 4D 200's $225/Mb (early January '90) (Have heard good things about this company, including 5 year replacement guarantee, not 90 days like some companies) ClearPoint 35 Parkwood Drive Hopkington,MA 07148 (800) 253-2778, (508) 435-2000 (A distributor: Dick Smith, Software Associtates: (916) 344-5558) Memory for 4D 20's, $76/Mb (GSA pricing) (late February '90) 1Mb sims, $100/Meg (80Ns) Memory for 4D/25, 4Mb sims 100ns, $730/4Meg (early March '90) Memory for SPARCstation 1+, 4Mb sims 80ns, $520/4Meg (mid April '90) (Life time warranty.) Parity Systems Inc. 504-B Vandell Way Campbell, CA 95008-9737, John Miller (408) 378-1000, FAX: (408) 378-1022 1MB 80ns SIMMS in Sun 3/x0's and SGI 4D/25's. $85US. (early Jan/90) No problems, fast ship, life-time warranty. Datarep Co., (408) 730-2121, David M. McWalters, Chip Meyers 100ns Piiceon, Helio: $575(/4Mb ?), 80ns: $625 (early March '90) South Coast Electronics, (213) 208-3260, Bob Rich 80ns Toshiba or Hitachi: $588(/4Mb ?) (early March '90) Systems Research, (415) 940-1890, Peter Kikos 80ns Hitachi: $595(/4Mb ?) Technology Works, Austin, TX, 800-622-2210 Memory for SPARCstation 1+, 4Mb sims 80ns, $520/4Meg (mid April '90) Note: Some one said they tried Hitachi SIMMS in a Personal and they didn't work. If anyone has any more names to add to the list or new prices please let me know. Also, please include an address or at least a phone number for the company. -- Brent L. Bates NASA-Langley Research Center M.S. 361 Hampton, Virginia 23665-5225 (804) 864-2854 E-mail: blbates@aero4.larc.nasa.gov or blbates@aero2.larc.nasa.gov   Received: from vmb.brl.mil by VMB.BRL.MIL id aa29043; 14 May 90 12:30 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa28661; 14 May 90 11:59 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa28532; 14 May 90 11:40 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa00787; 14 May 90 11:10 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA04514; Mon, 14 May 90 08:02:55 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 14 May 90 14:18:39 GMT From: mcsun!cernvax!chx400!urz.unibas.ch!doelz@uunet.uu.net Organization: University of Basel, Switzerland Subject: Re: Backing up, possible solution: gcc required. Message-Id: <1990May14.151840.725@urz.unibas.ch> References: <1990May11.142304.724@urz.unibas.ch>, <9005130148.AA03801@adenosine.pharm.utah.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <9005130148.AA03801@adenosine.pharm.utah.edu>, davis@ADENOSINE.PHARM.UTAH.EDU ("Darrell R. Davis") writes: > As with most (all?) fine gnu products, this is available by anonymous > ftp from prep.ai.mit.edu. Although I don't recollect, you probably > have to at least put "-I/usr/include/bsd" in your makefile, plus some > other minor :-) system dependent flags. Sounds promising ! I tried it, got it, and ... - well, it does even have a SGI makefile in it. Unfortunately, it also requires gnu cc and crashes with SGI cc. The gnu cc, thus, is 2.6 mega on the server, and besides the fact that it is not useful the net won't support ftp's of such large data files. I'd appreciate comments on whether is is worth to run gcc on the SGI and where to obtain it in a non-ftp procedure. Regards, Reinhard   Received: from vmb.brl.mil by VMB.BRL.MIL id aa00520; 14 May 90 14:06 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab29043; 14 May 90 12:42 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa28949; 14 May 90 12:21 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa01241; 14 May 90 11:55 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA07001; Mon, 14 May 90 08:50:30 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 14 May 90 14:54:43 GMT From: Jim Blue Organization: National Institute of Standards & Technology, Gaithersburg, MD Subject: xset -b Message-Id: <3513@fs1.cam.nist.gov> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL I am getting inconsistent results trying to turn off the beep on my PI. Sometimes "xset -b" or "xset b off" works and sometimes it doesn't. I see no consistency based on whether I give the command from a wsh or a xterm window, or anything else. Presumably there is a "hidden variable" that has eluded me. (xbiff -volume 0 doesn't turn off the beep, either, but it at least consistently never does it. I think.)   Received: from vmb.brl.mil by VMB.BRL.MIL id aa00921; 14 May 90 14:40 EDT Received: from vmb.brl.mil by VMB.brl.MIL id aa29544; 14 May 90 13:24 EDT Received: from adm.brl.mil by VMB.BRL.MIL id aa29487; 14 May 90 13:13 EDT Received: from ucbvax.Berkeley.EDU by ADM.BRL.MIL id aa21006; 14 May 90 12:40 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA09723; Mon, 14 May 90 09:35:59 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 14 May 90 16:01:22 GMT From: Jason Molenda Organization: Minnesota Supercomputer Institute Subject: MOVIE.BYU for the Iris Message-Id: <1990May14.160122.8466@s1.msi.umn.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL I'm sure the question has been asked a gazillion times and people don't want to hear it again, but please bear with me. Does anyone know of any companies distributing movie.byu for the Iris (PI or otherwise)? We'd like to get it up and going on our Personal but don't know who to talk to. Please reply via email and I'll post the results. Thanks! Jason Molenda, Minnesota Supercomputer Institute molenda@msi.umn.edu   Received: from vmb.brl.mil by VMB.BRL.MIL id aa03001; 14 May 90 15:51 EDT Received: from vmb.brl.mil by VMB.brl.MIL id aa02718; 14 May 90 15:40 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa02341; 14 May 90 15:16 EDT Received: from cunyvm.cuny.edu by VGR.BRL.MIL id aa00970; 14 May 90 14:54 EDT Received: from MCCLB0.MED.NYU.EDU by CUNYVM.CUNY.EDU (IBM VM SMTP R1.2.2MX) with BSMTP id 8081; Mon, 14 May 90 14:51:03 EDT Received: from mcirps2.med.nyu.edu by MCCLB0.MED.NYU.EDU; Mon, 14 May 90 13:52 EDT Received: by mcirps2.med.nyu.edu (5.52/890607.SGI) (for @mcclb0.med.nyu.edu:info-iris@brl.arpa) id AA04422; Mon, 14 May 90 14:18:02 DSD Date: Mon, 14 May 90 14:18:02 DSD From: root%mcirps2.med.nyu.edu@cunyvm.cuny.edu Subject: saving and restoring windows... To: info-iris@BRL.MIL Message-id: <9005142118.AA04422@mcirps2.med.nyu.edu> X-Envelope-to: info-iris@brl.arpa I want to save my window layout in a file, so I have an option in my graphics program to save window layout. This calls getsize() and getorigin(), and saves these values to a file. If this file is not present on the next invocation of the program, the user sizes and locates the graphics window in the usual way. However, if the file is present, then the program starts off with the window size and position that was last stored. I use the prefposition call prior to winopen(). The arg list to prefposition is lower x,y screen coords, and upper right x,y screen coords. I am finding that prefposition enforces a fixed window size, while letting the user move the window. The resize option on the window bar is grayed out and not available. The move option is available, and works. I want my users to pick the windowlayout that they want, and not have to resize the window on each program call. It should be saved. It should be resizable. prefposition() is acting like prefposition_forced_size() -- +-----------------------------------------------------------------------------+ | karron@nyu.edu Dan Karron | | . . . . . . . . . . . . . . New York University Medical Center | | 560 First Avenue \ \ Pager <1> (212) 397 9330 | | New York, New York 10016 \**\ <2> 10896 <3> | | (212) 340 5210 \**\__________________________________________ | +-----------------------------------------------------------------------------+   Received: from vmb.brl.mil by VMB.BRL.MIL id aa04100; 14 May 90 16:13 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab03001; 14 May 90 16:02 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa02913; 14 May 90 15:43 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa01189; 14 May 90 15:25 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA20113; Mon, 14 May 90 12:23:08 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 14 May 90 17:24:55 GMT From: Dan Watts Organization: Ki Research, Inc. Derry NH Subject: Print queue problems Message-Id: <756@ki.UUCP> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL I'm running on an SGI Personal Iris. My printer is connected to /dev/ttyd1. I'm running it at 9600 and it is supposed to use software handshakking (ie: XON/XOFF). The printer type is "Generic". My problem is that the XON/XOFF may not be working. On the first line of all print jobs, I see a line: interface/lpt1: ixon: not found What does this mean? I've also noticed with long jobs (more than five pages) or when printing multiple jobs, that the last page doesn't get completely printed. It looks like the last buffer of less than 512 bytes isn't getting flushed. I don't seem to have this problem if I "cat" directly to /dev/ttyd1 so I'm presuming the problem is in how I've configured the printer or the printer driver itself. Any help is greatly appreciated. -- dan watts -- ##################################################################### # CompuServe: >INTERNET:uunet.UU.NET!ki!dwatts Dan Watts # # UUCP : ...!uunet!ki!dwatts Ki Research, Inc. # ############### New Dimensions In Network Connectivity ##############   Received: from vmb.brl.mil by VMB.BRL.MIL id aa05616; 14 May 90 17:05 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa05542; 14 May 90 16:54 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa05260; 14 May 90 16:34 EDT Received: from aero4.larc.nasa.gov by VGR.BRL.MIL id aa01442; 14 May 90 16:12 EDT Received: Mon, 14 May 90 16:11:58 EDT by aero4.larc.nasa.gov (5.52/5.6) Date: Mon, 14 May 90 16:11:58 EDT From: "Brent L. Bates AAD/TAB MS361 x42854" Message-Id: <9005142011.AA01886@aero4.larc.nasa.gov> To: root%mcirps2.med.nyu.edu@cunyvm.cuny.edu Subject: Re: saving and restoring windows... Cc: info-iris@BRL.MIL After opening the window call wincontraints. If you don't call any of the contraint routines before wincontraints, the contraints are cleared. -- Brent L. Bates NASA-Langley Research Center M.S. 361 Hampton, Virginia 23665-5225 (804) 864-2854 E-mail: blbates@aero4.larc.nasa.gov or blbates@aero2.larc.nasa.gov   Received: from vmb.brl.mil by VMB.BRL.MIL id aa06420; 14 May 90 18:08 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa06221; 14 May 90 17:58 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa06212; 14 May 90 17:45 EDT Received: from cs.utah.edu by VGR.BRL.MIL id aa01765; 14 May 90 17:27 EDT Received: from adenosine.pharm.utah.edu by cs.utah.edu (5.61/utah-2.11-cs) id AA01312; Mon, 14 May 90 14:42:58 -0600 Date: Mon, 14 May 90 14:48:58 MDT From: "Darrell R. Davis" Posted-Date: Mon, 14 May 90 14:48:58 MDT Message-Id: <9005142048.AA05831@adenosine.pharm.utah.edu> Received: by adenosine.pharm.utah.edu (5.52/5.51) id AA05831; Mon, 14 May 90 14:48:58 MDT To: info-iris@BRL.MIL Subject: GNU tar, compilation thereof. Reinhard asked if gcc was required for compilation of GNU-tar, the answer I E-mailed to him was no, the correct answer is no, but you need to include alloca.c. I have verified this by rebuilding gtar on my 4D/20. Maybe with gcc you wouldn't need alloca.c, I don't know. If anyone wants to make life simple they can get an archive by anonymous ftp from adenosine.pharm.utah.edu (128.110.74.1) that has the altered Makefile (very minor change from dist.) and includes alloca.c. If you are really trusting, the tar and rmt executables are in the archive, give them a whirl (tar goes in /bin; rmt goes in /etc). Even if you don't have TeX (may the TeX God have mercy on your pitiful soul) the file tar.texinfo is ascii and you can wade through it. An iris 4D archive for the GNU version of tar is contained in gtar-iris.tar.Z The change that was required over the distribution was the inclusion of alloca.c and minor modification of the makefile to reflect this change. Good Luck, -------------------------------------------------------------- * Darrell R. Davis * * * "Faster, faster, until the Assistant Professor * * * thrill of speed overcomes Medicinal Chemistry *A**L**T**A* the fear of death." University of Utah * * * * * * --H.S. Thompson * --------------------------------------------------------------   Received: from vmb.brl.mil by VMB.BRL.MIL id aa06496; 14 May 90 18:28 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab06420; 14 May 90 18:17 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa06403; 14 May 90 18:06 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa01813; 14 May 90 17:42 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA28739; Mon, 14 May 90 14:29:31 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 14 May 90 18:40:44 GMT From: Michael Zeitlin Organization: Texaco Houston Res. Cntr Hou, Tx Subject: 1 vs. 4 mb dram Message-Id: <418@texhrc.UUCP> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Has anybody (or is it possible) upgraded their Personal Iris with 4 mb. drams in place of the 1mb memory .... I have a 4D25 with 16 mb currently and would like to go to know if I can go to 64 mb with just a swap out of the current chips... I know that I can go to 32 mb with 2mb dram cards..... thanks.... p.s. if it is possible....who supplies them that would be compatable... I have used Helios before....   Received: from vmb.brl.mil by VMB.BRL.MIL id ab06496; 14 May 90 18:28 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ac06420; 14 May 90 18:17 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa06412; 14 May 90 18:06 EDT Received: from cunyvm.cuny.edu by VGR.BRL.MIL id aa01825; 14 May 90 17:47 EDT Received: from MCCLB0.MED.NYU.EDU by CUNYVM.CUNY.EDU (IBM VM SMTP R1.2.2MX) with BSMTP id 1160; Mon, 14 May 90 17:44:16 EDT Received: from mcirps2.med.nyu.edu by MCCLB0.MED.NYU.EDU; Mon, 14 May 90 15:36 EDT Received: by mcirps2.med.nyu.edu (5.52/890607.SGI) (for @mcclb0.med.nyu.edu:info-iris@brl.arpa) id AA04914; Mon, 14 May 90 16:02:07 DSD Date: Mon, 14 May 90 16:02:07 DSD From: root%mcirps2.med.nyu.edu@cunyvm.cuny.edu Subject: prefposition() and winposition() To: info-iris@BRL.MIL Message-id: <9005142302.AA04914@mcirps2.med.nyu.edu> X-Envelope-to: info-iris@brl.arpa I am answering my previous posting about prefposition(). After reading the manual, there is a call winposition() which allows interactive resizing of the window size after winopen. Its name, winposition, implies that is SETS the window position. It also sets the size of the window, although there is a separate prefsize() call. Prefposition, implies that it SETS a PREFERRED window position, which can be changed. I think that this is bad nomenclature. Here is how you get an default window position and size that is resizable and repositionable. This took too much experimentation and careful reading of the manuals. The argument lists are not compatable between getsize, getorigin, and {pref|set}position. Call prefposition to set, ABSOLUTELY, the window position and size. You will not get resizing ticks at the window borders on winopen() if you stop there. Call winopen() to get your window. But don't stop there. Call winposition, with the same arguments as prefposition, to suggest a new, variable size and position window to 4sight. Call winconstraints() to make it happen. You will get resizing ticks at the border of the window. I don't know if the system does an entire redraw. It flashes by too fast to see the mechanism. Note that the arguments to {set|pref} position set both the origin and the size of the window, but they are specified as (x1,x2, y1,y2). You will have to take the origin and add it to the x and y sizes to get the absolute screen coords of the upper right corner. . . . prefposition(Window.xorg, Window.xorg+Window.wxsize, Window.yorg, Window.yorg+Window.wysize); winopen("YOUR WINDOW"); winposition(Window.xorg, Window.xorg+Window.wxsize, Window.yorg, Window.yorg+Window.wysize); winconstraints(); . . . after user resizes window, you will get a REDRAW token on the input queue. This is where you whould get the new window size, and calculate the aspect ratio (which you need for your new perspective) . . . case REDRAW: getsize(&Window.wxsize, &Window.wysize); getorigin(&Window.xorg,&Window.yorg); Window.aspect = (float) Window.wxsize / (float) Window.wysize; break; . . . I hope that this will save someone a few hours. It is a nice feature to save your window layout in a file once you are happy with the placement and size. My users did not want to place the window, they wanted it to cover most of the screen except for a textport at the bottom or top. dan. . -- +-----------------------------------------------------------------------------+ | karron@nyu.edu Dan Karron | | . . . . . . . . . . . . . . New York University Medical Center | | 560 First Avenue \ \ Pager <1> (212) 397 9330 | | New York, New York 10016 \**\ <2> 10896 <3> | | (212) 340 5210 \**\__________________________________________ | +-----------------------------------------------------------------------------+   Received: from vmb.brl.mil by VMB.BRL.MIL id ab06679; 14 May 90 19:09 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa06588; 14 May 90 18:59 EDT Date: Mon, 14 May 90 18:47:22 EDT From: Mike Muuss To: Info-Iris@BRL Subject: Re: Backing up remote machines? Next release ... ? Message-ID: <9005141847.aa06555@VMB.BRL.MIL> There is a University that has done a port of DUMP/RDUMP and RESTORE/RRESTORE from BSD to the Iris. It is good clean code, and seems to be stable. I would encourage all you SGI customers to pester your SGI sales droids (er, "marketing representatives") into supporting DUMP & RESTORE in some future "Irix" release. Suggest to SGI that they simply acquire this existing software from the University -- I'm sure the University would feel well compensated if SGI gave them something in return, perhaps some CPU boards. The reason I won't acknowledge which University has done this wonderful thing, is to prevent all of you from pounding on them for something that they can't give you (isn't software licensing wonderful?). But SGI knows who they are.... Best, -Mike   Received: from vmb.brl.mil by VMB.BRL.MIL id aa06903; 14 May 90 19:39 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa06808; 14 May 90 19:29 EDT Date: Mon, 14 May 90 19:14:19 EDT From: Mike Muuss To: Darrell R. Davis cc: info-iris@BRL.MIL Subject: Re: Backing up, possible solution. Message-ID: <9005141914.aa06751@VMB.BRL.MIL> We considered using GNU-Tar for our backups, and also PAX (ANSI CPIO), but rejected them because of these limitations: TAR 100 character maximum full path names CPIO 256 character maximum full path names A number of our users use software (such as InterLeaf) which encode things in file and directory names, making for very long full paths. Our TAR format "dumps" routinely encountered files with full path names longer than 100 characters. While none have hit the 256 limit, it is a scary prospect to have such a limitation built into the tape format. It is a shame than when ANSI standardized CPIO and TAR, they did not remove this 1.5-decade old restriction. Sigh. Lobby SGI for DUMP/RESTORE instead. Best, -Mike   Received: from vmb.brl.mil by VMB.BRL.MIL id aa07274; 14 May 90 20:39 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa07145; 14 May 90 20:28 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa07047; 14 May 90 20:16 EDT Received: from life.ai.mit.edu by VGR.BRL.MIL id aa02363; 14 May 90 19:55 EDT Received: from rice-chex (rice-chex.ai.mit.edu) by life.ai.mit.edu (4.1/AI-4.10) id AA06647; Mon, 14 May 90 19:54:07 EDT From: "William Y-P. Lim" Received: by rice-chex (4.1/AI-4.10) id AA14554; Mon, 14 May 90 19:52:36 EDT Date: Mon, 14 May 90 19:52:36 EDT Message-Id: <9005142352.AA14554@rice-chex> To: info-iris@BRL.MIL Subject: QIC 24 vs QIC 150 tapes Why can't QIC 24 tapes be used for tape dumps using the Personal Iris cartridge tape drive? Willie   Received: from vmb.brl.mil by VMB.BRL.MIL id ab07274; 14 May 90 20:39 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab07145; 14 May 90 20:28 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab07047; 14 May 90 20:17 EDT Received: from cunyvm.cuny.edu by VGR.BRL.MIL id aa02369; 14 May 90 19:59 EDT Received: from MCCLB0.MED.NYU.EDU by CUNYVM.CUNY.EDU (IBM VM SMTP R1.2.2MX) with BSMTP id 2470; Mon, 14 May 90 19:56:48 EDT Received: from mcirps2.med.nyu.edu by MCCLB0.MED.NYU.EDU; Mon, 14 May 90 19:56 EDT Received: by mcirps2.med.nyu.edu (5.52/890607.SGI) (for @mcclb0.med.nyu.edu:info-iris@brl.arpa) id AA06114; Mon, 14 May 90 20:22:26 DSD Date: Mon, 14 May 90 20:22:26 DSD From: root%mcirps2.med.nyu.edu@cunyvm.cuny.edu Subject: more arguments and parameters problems... To: info-iris@BRL.MIL Message-id: <9005150322.AA06114@mcirps2.med.nyu.edu> X-Envelope-to: info-iris@brl.arpa I noticed a previous item concerning passing/returning structs in c. This problem may be related. I am attempting to pass a large number of Coord that are elements of an array. First, I would like to pass a pointer to the base array element for some coords in the struct, but that is another story. I have been reduced to this horsey code, which STILL does not work. Checking what is received with dbx/edge shows that the first few values are wrong, while the rest are ok. I check the values with edge from the calling subroutine AlignConstellation and the receiving subroutine DoAlign. ============================================================== struct Database { char Skull[18]; struct { int Number; char FileName[40]; char MenuName[40]; char LongFileName[100]; char DataType; short LineStyle; char LongLabel[LONG_LABEL_SIZE]; short IsChanged; struct { char Pressed; float x, y, z; } Record[MAX_RECORDS]; int ActiveRecords; short ShowLabel; } Lines[MAX_LABELS]; int ActiveLabelLines; }; #ifdef ALLOCATE #else extern #endif struct Database SkullRecord; int AlignPoints[3]; void AlignConstellation() { DoAlign( SkullRecord.Lines[AlignPoints[0]].Record[0].x, SkullRecord.Lines[AlignPoints[0]].Record[0].y, SkullRecord.Lines[AlignPoints[0]].Record[0].z, SkullRecord.Lines[AlignPoints[0]].Record[1].x, SkullRecord.Lines[AlignPoints[0]].Record[1].y, SkullRecord.Lines[AlignPoints[0]].Record[1].z, SkullRecord.Lines[AlignPoints[1]].Record[0].x, SkullRecord.Lines[AlignPoints[1]].Record[0].y, SkullRecord.Lines[AlignPoints[1]].Record[0].z, SkullRecord.Lines[AlignPoints[1]].Record[1].x, SkullRecord.Lines[AlignPoints[1]].Record[1].y, SkullRecord.Lines[AlignPoints[1]].Record[1].z, SkullRecord.Lines[AlignPoints[2]].Record[0].x, SkullRecord.Lines[AlignPoints[2]].Record[0].y, SkullRecord.Lines[AlignPoints[2]].Record[0].z, SkullRecord.Lines[AlignPoints[2]].Record[1].x, SkullRecord.Lines[AlignPoints[2]].Record[1].y, SkullRecord.Lines[AlignPoints[2]].Record[1].z); return; } /*****************/ /****************************************************/ DoAlign(x11,y11,z11,x12,y12,z12,x21,y21,z21,x22,y22,z22,x31,y31,z31,x32,y32,z32) float x11,y11,z11,x12,y12,z12,x21,y21,z21,x22,y22,z22,x31,y31,z31,x32,y32,z32; { return; } /*******************************************************************/ -- +-----------------------------------------------------------------------------+ | karron@nyu.edu Dan Karron | | . . . . . . . . . . . . . . New York University Medical Center | | 560 First Avenue \ \ Pager <1> (212) 397 9330 | | New York, New York 10016 \**\ <2> 10896 <3> | | (212) 340 5210 \**\__________________________________________ | +-----------------------------------------------------------------------------+   Received: from vmb.brl.mil by VMB.BRL.MIL id aa07368; 14 May 90 21:01 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ac07274; 14 May 90 20:50 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa07217; 14 May 90 20:34 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa02367; 14 May 90 19:57 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA08650; Mon, 14 May 90 16:51:55 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 14 May 90 23:43:40 GMT From: "James D. Meiss" Organization: University of Colorado, Boulder Subject: bru backups with cron Message-Id: <21113@boulder.Colorado.EDU> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL I'm trying to get a script working for weekly backups. What I've got so far seems to "work" though it has two problems: 1) I get a message mailed to "root" which also goes to root at other machines on our local net. Actually I don't know how far it goes, but I've only received complaints from one local machine. I hope it doesn't wander out into the universie 2) The script seems to backup the entire disk! Here is what I have. I installed the following in "roots" crontabs file: # Backup # Rewind tape # Forward to end of recorded data # Backup files new since last backup # Touch creation/modification date of lastbackup 0 1 * * 0 mt rewind; mt feom ; bru -c -vf /dev/nrtape -s 150M -BR -n /etc/lastbackup >/etc/backup_logfile; touch /etc/lastbackup The file /ect/lastbacup has a creation date only one week old, so I expect only one weeks worth of stuff to be backed up. The message mailed to root is a warning that /etc/backup_logfile has grown during the backup. Please pardon my ignorance! Jim Meiss Program in Applied Mathematics jdm@boulder.colorado.edu   Received: from vmb.brl.mil by VMB.BRL.MIL id aa07689; 14 May 90 22:13 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa07609; 14 May 90 22:02 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa07598; 14 May 90 21:49 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa02687; 14 May 90 21:41 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA15174; Mon, 14 May 90 18:36:38 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 15 May 90 01:02:32 GMT From: Wiltse Carpenter Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Re: /dev/audio has a bug or a feature? Message-Id: <60195@sgi.sgi.com> References: <1990Apr26.162859.8089@sgzh.uucp>, <1990Apr27.075219.9522@sgzh.uucp> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL (Sorry this is kind of out of date) In article <1990Apr27.075219.9522@sgzh.uucp>, root@sgzh.uucp (Bruno Pape) writes: > In comp.sys.sgi I wrote: > > >>{ > >>/* char b[1000]; */ > >> char c, b[1000]; > >> int rtn, audio; > >> register i; > > > >>If you can tell me why, I will think you are real smart. > Steve Jay's response seems to be correct. I think he is real smart. > -------------------------------------------------------------------------- > This is just a guess, but with "char b[1000];", b will almost certainly > be at a word aligned address. With "char c,b[1000];", b will almost > certainly NOT be word aligned. I haven't looked at the documentation, > so I don't know if the audio device requires word alignment, or if > this is a bug. > -------------------------------------------------------------------------- > With "char c[4],b[1000];" or any other word alignment it works correctly. > Can you guys back home fix this so it works like any other file I/O. > The audio driver throws out the first or last 3 bytes of an unaligned transfer. While the hardware supports it, you would get ~40 usec glitches between the unaligned and the aligned portions of the transfer. So for best results, align your data (you can do this easily by declaring the buffer as an int or using the return from malloc), and using multiples of 4 as the length. If these seem hard to do in your program, try just using a longer sample buffer to minimize the error. The example given is the pathological case for the rounding that occurs since the length was so short. -Wiltse   Received: from vmb.brl.mil by VMB.BRL.MIL id aa09413; 15 May 90 3:05 EDT Received: from vmb.brl.mil by VMB.brl.MIL id aa08751; 15 May 90 1:31 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa08715; 15 May 90 1:16 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa03530; 15 May 90 0:46 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA20356; Mon, 14 May 90 19:59:46 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 15 May 90 02:42:20 GMT From: Shin Kurokawa Organization: University of Chicago Subject: Info wanted about /tmp/.ps_data Message-Id: <9273@tank.uchicago.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL I'd like to find about how /tmp/.ps_data is being used by the system (what's it for anyway??). Could anyone please post any useful info about this file? Cheers, Shin Kurokawa shin@tank.uchicago.edu Physical Sciences Numerical Calculation Laboratory, University of Chicago 5640 S.Ellis Ave., Chicago, IL 60637 USA   Received: from vmb.brl.mil by VMB.BRL.MIL id aa10080; 15 May 90 5:27 EDT Received: from vmb.brl.mil by VMB.brl.MIL id aa09800; 15 May 90 4:24 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa09768; 15 May 90 4:14 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa04029; 15 May 90 3:56 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA06356; Tue, 15 May 90 00:53:12 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 14 May 90 19:14:19 GMT From: Scott_Klosterman Organization: SDRC, Cincinnati Subject: Secure tftp for IRIS Message-Id: <1406@sdrc.UUCP> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Is there a secure tftpd for the IRIS? Please send info to: uunet!sdrc!crscott crscott@SDRC.UU.NET Thanks in advance Scott Klosterman   Received: from vmb.brl.mil by VMB.BRL.MIL id aa09240; 17 May 90 6:04 EDT Received: from vmb.brl.mil by VMB.brl.MIL id aa09127; 17 May 90 5:33 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa09110; 17 May 90 5:14 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa15604; 17 May 90 5:00 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA24147; Thu, 17 May 90 01:34:16 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 14 May 90 17:04:42 GMT From: Jim Rodger Organization: Queen's University, Kingston, Ontario, Canada Subject: video options for 4D50, 4D85 Message-Id: <707@qusunf.queensu.CA> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Our site is looking at alternatives for getting NTSC composite video from our 4D50G and 4D85GT machines, for recording on consumer VHS recorders. In light of recent negative comments on the CG2 board's performance, we would appreciate users' evaluations of it, the CG3 board, VideoCreator and any other products that people may be using. SGI sales types seem to be pretty uninformed about these products. Maybe SGI techies that read this group can respond. Jim Rodger rodger@qucis.queesnu.ca   Received: from vmb.brl.mil by VMB.BRL.MIL id aa11860; 15 May 90 8:02 EDT Received: from vmb.brl.mil by VMB.brl.MIL id ab11305; 15 May 90 7:52 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab11281; 15 May 90 7:41 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa04477; 15 May 90 7:11 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA15620; Tue, 15 May 90 04:01:34 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 15 May 90 09:21:09 GMT From: eru!luth!sunic!mcsun!cernvax!chx400!urz.unibas.ch!doelz@bloom-beacon.mit.edu Organization: University of Basel, Switzerland Subject: Compiler signal code ? (was:Re: GNU tar, compilation thereof.) Message-Id: <1990May15.102109.729@urz.unibas.ch> References: <9005142048.AA05831@adenosine.pharm.utah.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <9005142048.AA05831@adenosine.pharm.utah.edu>, davis@ADENOSINE.PHARM.UTAH.EDU ("Darrell R. Davis") writes: > need to include alloca.c. I have verified this by rebuilding gtar on > my 4D/20. Maybe with gcc you wouldn't need alloca.c, I don't know. If > anyone wants to make life simple they can get an archive by anonymous > ftp from adenosine.pharm.utah.edu (128.110.74.1) that has the altered > Makefile (very minor change from dist.) and includes alloca.c. If you I got it from there, no problem, and tried again; # make cc -g -DUSG -I/usr/include/bsd -DDEF_AR_FILE=\"/dev/tape\" -DDEFBLOCKING=20 -c tar.c Fatal error in: /usr/lib/ugen - core dumped Signal 139 *** Error code 1 Stop. The executable works, but , frankly speaking, I don't like to run ftp'ed binaries as root. Therefore, I suspect that the ugen is sick on 4D > 20. Mine is # ls -ls /usr/lib/ugen 609 -rwxr-xr-x 1 bin bin 311344 Dec 6 09:20 /usr/lib/ugen which is the primary installation date of 3.2 on our system. Currently we're running 3.2.2 but obviously, things didn't change. Explanations appreciated. Where can I find signal 139? My knowledge of unix, so far didn't need to deal with signals > 30. - Reinhard   Received: from vmb.brl.mil by VMB.BRL.MIL id aa12080; 15 May 90 8:23 EDT Received: from vmb.brl.mil by VMB.brl.MIL id aa11305; 15 May 90 7:51 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa11281; 15 May 90 7:41 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa04475; 15 May 90 7:11 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA15604; Tue, 15 May 90 04:01:23 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 15 May 90 08:33:41 GMT From: eru!luth!sunic!mcsun!cernvax!chx400!urz.unibas.ch!doelz@bloom-beacon.mit.edu Organization: University of Basel, Switzerland Subject: Re: record uptime? number of processes ? Message-Id: <1990May15.093341.728@urz.unibas.ch> References: <133@cutmcvax.OZ> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <133@cutmcvax.OZ>, architec@cutmcvax.OZ (Phil Dench ) writes: > > The irises been down for various reasons; > gomorrah - I did a movie on about 750 frames that grabbed heaps of memory > and then I couldn't kill it so gomorrah had to be reset. > lust - something I did froze the news_server so it had to be reset. Take another terminal, kill the session and /or grcond, and run /etc/gl/restartgl which will restart the graphics manager. > > I dont expect them all to stay up much longer though. Now > that the winter storms are here, blackouts will be pretty > common. Depending on how many users do how much, our SGI stays up only about a month or two. What counts, instead, is the number of processes run. We cycle through the max process number at least twice per week, i.e. we have 250000 processes at the minimum before the SGI box dies. The reason for volunteering the reset is that we encounter 'degradation' usually, i.e. we don't know why but the system gets slower. I try to keep a record on the phenomenon but it seems difficult to be precise. - Reinhard   Received: from vmb.brl.mil by VMB.BRL.MIL id aa13938; 15 May 90 9:17 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa12313; 15 May 90 8:44 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa12227; 15 May 90 8:28 EDT Received: from life.ai.mit.edu by VGR.BRL.MIL id aa04654; 15 May 90 8:09 EDT Received: from rice-chex (rice-chex.ai.mit.edu) by life.ai.mit.edu (4.1/AI-4.10) id AA14967; Tue, 15 May 90 08:07:09 EDT From: "William Y-P. Lim" Received: by rice-chex (4.1/AI-4.10) id AA16842; Tue, 15 May 90 08:05:40 EDT Date: Tue, 15 May 90 08:05:40 EDT Message-Id: <9005151205.AA16842@rice-chex> To: info-iris@BRL.MIL Subject: Messing around with sendmail.cf At my place of work, we have a Personal Iris connected to an internal network. There are Unix machines (BSD, Sparc stations, Ultrix) and Symbolics machines on that network. However we are having problems trying to get email to work on the Personal Iris (ftp and telnet seem to work though). We could send mail out from the Personal Iris to the other hosts. The problem is that we can't send mail to the Personal Iris. Any ideas on how sendmail.cf should be fixed to get mail to work properly? Thanks in advance. Willie   Received: from vmb.brl.mil by VMB.BRL.MIL id ac13938; 15 May 90 9:27 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa12664; 15 May 90 9:06 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa12630; 15 May 90 8:52 EDT Received: from aero4.larc.nasa.gov by VGR.BRL.MIL id aa04626; 15 May 90 8:00 EDT Received: Tue, 15 May 90 07:59:31 EDT by aero4.larc.nasa.gov (5.52/5.6) Date: Tue, 15 May 90 07:59:31 EDT From: "Brent L. Bates AAD/TAB MS361 x42854" Message-Id: <9005151159.AA04114@aero4.larc.nasa.gov> To: ki!dwatts@uunet.uu.net Subject: Re: Print queue problems Cc: info-iris@BRL.MIL Here is a line from an interface file I have: stty raw cs8 9600 -parenb -cstopb cread -clocal ixon ixany ixoff -ofill cr3 nl1 tab2 bs1 ff1 vt1 0<&1 I have software hand shaking enabled and delays timed instead of fill characters. I also specified port ttyf2 on my 3130, this allows hardware hand shaking too. I hope this is of some help. -- Brent L. Bates NASA-Langley Research Center M.S. 361 Hampton, Virginia 23665-5225 (804) 864-2854 E-mail: blbates@aero4.larc.nasa.gov or blbates@aero2.larc.nasa.gov   Received: from vmb.brl.mil by VMB.BRL.MIL id aa14586; 15 May 90 9:38 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab13938; 15 May 90 9:27 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa12665; 15 May 90 8:55 EDT Received: from DETRICK-EMH1.ARMY.MIL by VGR.BRL.MIL id aa04724; 15 May 90 8:29 EDT Date: 15 May 90 08:18:00 EDT From: "EAGLE::HACK" Subject: Command line history/editing To: info-iris Message-ID: <9005150829.aa04724@VGR.BRL.MIL> Another new forum member question: Does anyone know of a utility on the PI w/ 3.2 to allow for command line recall/ editing using the cursor keys? Dallas C. Hack U.S. Army Medical Research Institute of Infectious Diseases hack@detrick-emh1.army.mil   Received: from vmb.brl.mil by VMB.BRL.MIL id aa16246; 15 May 90 10:11 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa15379; 15 May 90 10:01 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa14756; 15 May 90 9:42 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa04884; 15 May 90 9:12 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA21805; Tue, 15 May 90 06:02:53 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 15 May 90 11:29:57 GMT From: mcsun!cernvax!chx400!urz.unibas.ch!doelz@uunet.uu.net Organization: University of Basel, Switzerland Subject: getrusage(2) hack wanted Message-Id: <1990May15.122957.730@urz.unibas.ch> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In order to port a BSD application, I need to know some data of resource usage. This is accomplished with the getrusage (2) call: NAME getrusage - get information about resource utilization SYNOPSIS #include #include #define RUSAGE_SELF 0 /* calling process */ #define RUSAGE_CHILDREN -1 /* terminated child processes */ getrusage(who, rusage) int who; struct rusage *rusage; DESCRIPTION Getrusage returns information describing the resources util- ized by the current process, or all its terminated child processes. The who parameter is one of RUSAGE_SELF and RUSAGE_CHILDREN. If rusage is nonzero, the buffer it points to will be filled in with the following structure: ... As you might have recognized, this man page cannot be from a SGI box because the system call isn't implemented. Does anyone in netland have a procedure mimicking this call ? Pointers, helps, etc. mostly appreciated. Reinhard   Received: from vmb.brl.mil by VMB.BRL.MIL id aa18735; 15 May 90 11:03 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa17848; 15 May 90 10:52 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa17357; 15 May 90 10:39 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa05339; 15 May 90 10:11 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA25675; Tue, 15 May 90 07:06:49 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 15 May 90 13:41:04 GMT From: George Elkins Organization: Rutgers Univ., New Brunswick, N.J. Subject: GNU Chess on Iris 4D/120 GTX Message-Id: Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Has anyone out there compiled Gnu Chess on a Silicon Graphics Iris workstation? George Elkins   Received: from vmb.brl.mil by VMB.BRL.MIL id ab18735; 15 May 90 11:04 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab17848; 15 May 90 10:53 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa17478; 15 May 90 10:40 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa05368; 15 May 90 10:13 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA25085; Tue, 15 May 90 06:57:27 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 15 May 90 09:55:56 GMT From: Rob Warnock Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Re: saving and restoring windows... Message-Id: <60230@sgi.sgi.com> References: <9005142118.AA04422@mcirps2.med.nyu.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <9005142118.AA04422@mcirps2.med.nyu.edu> root@MCIRPS2.MED.NYU.EDU writes: +--------------- | I want to save my window layout in a file, so I have an option in | my graphics program to save window layout. This calls getsize() and | getorigin(), and saves these values to a file... I am finding that | prefposition enforces a fixed window size... I want my users to pick | the windowlayout that they want, and not have to resize the window on | each program call. It should be saved. It should be resizable. | prefposition() is acting like prefposition_forced_size()... +--------------- After your "winopen()", add a call to "winconstraints()", like this: prefposition(oldx, oldx + oldxsize, oldy, oldy + oldysize); winopen("yournamegoeshere"); winconstraints(); This will clear the constraints, and you can resize the window. This may not be totally obvious the first time you read them "winopen()" and "winconstraints()" man pages, but it is there: WINOPEN(3G) winopen creates a graphics window as defined by the current values of the window constraints... After a call to winopen, the system resets the graphics state of the window (this includes window constraints) to its default value. ... WINCONSTRAINTS(3G) After binding these constraints to a window, winconstraints resets the window constraints to their default values, if any... If, after the "winopen()", you wanted *some* constraints but not others, you'd have to explicitly call the various constraint-setting routines (minsize(), maxsize(), keepaspect(), prefsize(), etc.) for *all* of the constraints you wanted, then call winconstraints(). All the constraints you *didn't* call again would be released to their defaults at that point. -Rob ----- Rob Warnock, MS-9U/510 rpw3@sgi.com rpw3@pei.com Silicon Graphics, Inc. (415)335-1673 Protocol Engines, Inc. 2011 N. Shoreline Blvd. Mountain View, CA 94039-7311   Received: from vmb.brl.mil by VMB.BRL.MIL id ab01581; 15 May 90 13:51 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa00090; 15 May 90 13:28 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa18959; 15 May 90 11:07 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa05661; 15 May 90 10:43 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA27461; Tue, 15 May 90 07:32:26 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 15 May 90 14:30:52 GMT From: Jean-Francois Lamy Subject: signals on multi-processor Message-Id: <90May15.103014edt.827@smoke.cs.toronto.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Assume that one has a process 1 that does setup HUP signal handler that sets a flag grab lock if HUP flag is set reload data files that were cached release lock and a process 2 that does grab lock mess up with data files process 1 has cached kill -HUP process 2 release lock The two processes are racing each other. In the picture I have of traditional Unix implementations, if process 2 seizes the lock first (I have in mind the simple-minded locks where the process creates a link to a file), the signal will be sent by setting a bit somewhere in 1's data structures, before process 2 releases the lock and the handler in process 1 will be triggered before 1 ever tests for the flag, by virtue of process 1 having to do system calls to acquire to lock before it tests for said flag (the act of locking requires a system call, at which point the bit set in the process data structures will get attention, assuming it did not before). So in traditional Unix this would be a workable if ugly way to warn a process that it's caches have been invalidated, or some such. I am fully aware that there are other ways of doing this that do not involve signals, but for the purpose of this discussion, could anyone tell me whether Irix delivers signals in a way that guarantees that process 2 will have processed the signal sent by process 1 before it gets to test for the flag? Or is signal delivery more asynchronous, meaning that the signal could float in space for a while and get delivered only after 1 has tested the flag and assumed the caches it held were valid? Finally, should this work in Irix as in other Unixes, is this mandated in some standard or just an accident? Jean-Francois Lamy lamy@cs.utoronto.ca, uunet!cs.utoronto.ca!lamy Department of Computer Science, University of Toronto, Canada M5S 1A4   Received: from vmb.brl.mil by VMB.BRL.MIL id af10362; 15 May 90 18:15 EDT Received: from vmb.brl.mil by VMB.brl.MIL id aa10163; 15 May 90 17:54 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa10086; 15 May 90 17:35 EDT Received: from [192.48.139.30] by VGR.BRL.MIL id aa00586; 15 May 90 13:51 EDT Received: from neumann (neumann.squibb.com) by dino.squibb.com; Tue, 15 May 90 13:41 EST Received: by neumann (5.52/5.7) id AA03711; Tue, 15 May 90 13:51:09 EDT Date: Tue, 15 May 90 13:51:09 EDT From: shaginaw@neumann.squibb.com Subject: TEK color emulation To: info-iris@BRL.MIL Message-id: <9005151751.AA03711@neumann> X-Envelope-to: info-iris@brl.mil Is anyone aware of any Tektronix color terminal emulators for the 4D? Specifically, I'd like to find a 4107 emulator. Nevertheless, I'd like any info anyone can provide on emulators for ANY of the color TEKs. Thanks in advance. -- Rich -- ------------------------------------------------------------------------------- Squibb Institute for Medical Research -- Bristol-Myers Squibb Company Richard J. Shaginaw Internet Address: shaginaw@squibb.com Principal Systems Engineer Telephone: 609-921-5184 Macromolecular Modeling Department FAX: 609-683-6607 ===============================================================================   Received: from vmb.brl.mil by VMB.BRL.MIL id ag10362; 15 May 90 18:15 EDT Received: from vmb.brl.mil by VMB.brl.MIL id ab10163; 15 May 90 17:54 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa10107; 15 May 90 17:35 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa00637; 15 May 90 13:57 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA10128; Tue, 15 May 90 10:48:25 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 15 May 90 17:11:46 GMT From: Aaron Schuman Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Re: Secure tftp for IRIS Message-Id: <7715@odin.corp.sgi.com> References: <1406@sdrc.UUCP> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In Irix 3.3, tftpd can be invoked with the -s ("secure") option. The daemon's run string includes a list of directories which can be accessed by the tftp user. If the -s option is omitted, the daemon adds "/" to the list, in effect showing the tftp user every world-readable file on the server. If the directory list is empty, tftpd's home directory, /usr/etc/boot, can still be read.   Received: from vmb.brl.mil by VMB.BRL.MIL id aa10605; 15 May 90 18:26 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa10362; 15 May 90 18:14 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa10175; 15 May 90 17:43 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa01044; 15 May 90 14:44 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA12906; Tue, 15 May 90 11:39:13 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 15 May 90 18:20:46 GMT From: Shin Kurokawa Organization: University of Chicago Subject: Wanted BSD ports of finger,w, and msgs Message-Id: <9288@tank.uchicago.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL What I'd like to see in the future release of the IRIX are better interface with "finger", and possibly the ports of "w" and "msgs". Basically, many non-IRIX finger works this way: someone types "finger" on that machine, and the listing such as the following would come up: Login Name TTY Idle When Where bozo Bozo the Clown p0 2:15 Tue 12:00 gumby.univbozo.edu foo Mr.Foo console 2 Wed 3:15 mongo Mongo p3 Wed 5:00 mongo.univbozo.edu Since the fingerd passes the incoming finger requests from outside to the finger program, the user@othermachine will see the similar output upon invoking finger @targethost. In IRIX, the user has to type some options after "finger" to get anything. Also, the info about the users' last logins would be very useful. The ucb program "w" is like "whodo", except one gets much nicer output, such as: 12:45pm up 3 days, 2:15, 2 users, load average: 2.68, 2.21, 2.02 User tty login@ idle JCPU PCPU what bozo ttyp0 12:00pm 2:15 21 10 rn comp.sys.sgi foo console 3:15am 40 23 rm -f *.* The "msgs" is a little utility that's usually used for system-wide announcements, like IRIX "news". The nice feature of the "msgs" is that one can send a mail message to msgs@host to get his/her message seen through the msgs utility. This can be helpful, when some folks from other sites want to send important messages like "gateway such-n-such will be shutdown tomorrow". Any succesful attempts at porting these to Iris, anyone? Cheers, Shin Kurokawa shin@tank.uchicago.edu Physical Sciences Numerical Calculation Laboratory, University of Chicago 5640 S.Ellis Ave., Chicago, IL 60637 USA   Received: from vmb.brl.mil by VMB.BRL.MIL id ab10605; 15 May 90 18:26 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab10362; 15 May 90 18:15 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab10224; 15 May 90 17:50 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa01436; 15 May 90 15:57 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA16751; Tue, 15 May 90 12:42:05 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 15 May 90 18:52:10 GMT From: "Scott R. Presnell" Subject: Re: Wanted BSD ports of finger,w, and msgs Message-Id: References: <9288@tank.uchicago.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL shin@tank.uchicago.edu (Shin Kurokawa) writes: >What I'd like to see in the future release of the IRIX are >better interface with "finger", and possibly the ports of "w" I've ported "w" and "top." If interested - drop me a line. - Scott -- Scott Presnell +1 (415) 476-9890 Pharm. Chem., S-926 Internet: srp@cgl.ucsf.edu University of California UUCP: ...ucbvax!ucsfcgl!srp San Francisco, CA. 94143-0446 Bitnet: srp@ucsfcgl.bitnet   Received: from vmb.brl.mil by VMB.BRL.MIL id ac10605; 15 May 90 18:26 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ac10362; 15 May 90 18:15 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ac10224; 15 May 90 17:50 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa01461; 15 May 90 15:58 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA17114; Tue, 15 May 90 12:49:02 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 15 May 90 19:03:16 GMT From: Paul Jackson Organization: Silicon Graphics, Research & Development Subject: Re: Info wanted about /tmp/.ps_data Message-Id: <7732@odin.corp.sgi.com> References: <9273@tank.uchicago.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <9273@tank.uchicago.edu>, shin@tank.uchicago.edu (Shin Kurokawa) writes: > I'd like to find about how /tmp/.ps_data is being used by the > system (what's it for anyway??). Could anyone please post any > useful info about this file? It is a temporary file created by and used by the ps(1) command. Try removing it, and observe that ps takes longer to run on the next invocation. Ps uses this file to store such things as the mapping of character major,minor special device numbers to /dev/tty* pathnames. The ps command automatically rebuilds /tmp/.ps_data if any of /unix, /dev or /etc/passwd have more recent modification times. Back in the dark ages (Version 7, System III, ...) before this performance hack to ps, it was a notoriously slow command, because of all the stat's it had to do in /dev at each invocation. Thanks, take care ... Paul Jackson (pj@asd.sgi.com), x1373   Received: from vmb.brl.mil by VMB.BRL.MIL id ad10605; 15 May 90 18:26 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ad10362; 15 May 90 18:15 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ad10224; 15 May 90 17:50 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa01478; 15 May 90 15:58 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA17133; Tue, 15 May 90 12:49:16 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 15 May 90 19:08:56 GMT From: Paul Jackson Organization: Silicon Graphics, Research & Development Subject: Re: Backing up remote machines? Next release ... ? Message-Id: <7733@odin.corp.sgi.com> References: <9005141847.aa06555@VMB.BRL.MIL> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <9005141847.aa06555@VMB.BRL.MIL>, mike@BRL.MIL (Mike Muuss) writes: > There is a University that has done a port of DUMP/RDUMP and RESTORE/RRESTORE > from BSD to the Iris. It is good clean code, and seems to be stable. > > I would encourage all you SGI customers to pester your SGI sales droids > (er, "marketing representatives") into supporting DUMP & RESTORE in some > future "Irix" release. > > The reason I won't acknowledge which University has done this wonderful > thing, is to prevent all of you from pounding on them for something that > they can't give you (isn't software licensing wonderful?). But SGI knows > who they are.... Yes - we know who they are (Thanks be to them). We intend to provide dump/restore in some future release. It won't make the release that comes out in the next few months, but we really are actively working on it, so someday ... Keep those cards and letters coming ... Thanks, take care ... Paul Jackson (pj@asd.sgi.com), x1373   Received: from vmb.brl.mil by VMB.BRL.MIL id ae10605; 15 May 90 18:26 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ae10362; 15 May 90 18:15 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab10229; 15 May 90 17:51 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa01651; 15 May 90 16:13 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA17760; Tue, 15 May 90 12:59:16 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 15 May 90 19:15:08 GMT From: "Jack P. Weldon" Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Re: Print queue problems Message-Id: <7735@odin.corp.sgi.com> References: <756@ki.UUCP> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <756@ki.UUCP> dwatts@ki.UUCP (Dan Watts) writes: > >I'm running on an SGI Personal Iris. My printer is connected to /dev/ttyd1. >I'm running it at 9600 and it is supposed to use software handshakking (ie: >XON/XOFF). The printer type is "Generic". My problem is that the XON/XOFF >may not be working. On the first line of all print jobs, I see a line: > > interface/lpt1: ixon: not found > From the 3.2.2 Release Notes (paraphrased): SCR 8824: If you used the visual admin tool to add a dumb (Generic) serial printer AND you chose "software handshaking," when that tool added the printer, it edited the template of the dumb script (in /usr/spool/lp/model), and installed the edited copy in /usr/spool/lp/interface/{printer-name}. The tool edits the file to setup the baud rate and the stty parameters. This was done incorrectly by leaving out the double quotes needed on the STTYPARAMS line. Fixed in 3.2.2 Example: # Change the following to set up defaults STTYPARAMS=-ixany ixon ixoff <--Change to: STTYPARAMS="-ixany ixon ixoff" BAUDRATE=9600 BANNER=1 TRAILER=0 RAW=0 Cheers, Jack P. Weldon (jweldon@sgi.com) SGI Product Support Engineering   Received: from vmb.brl.mil by VMB.BRL.MIL id aa10871; 15 May 90 18:50 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id af10605; 15 May 90 18:39 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa10594; 15 May 90 18:22 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa00251; 15 May 90 13:15 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA05050; Tue, 15 May 90 09:32:30 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 15 May 90 12:19:52 GMT From: Dan Watts Organization: Ki Research, Inc. Derry, NH Subject: Re: QIC 24 vs QIC 150 tapes Message-Id: <760@ki.UUCP> References: <9005142352.AA14554@rice-chex> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <9005142352.AA14554@rice-chex> wlim@AI.MIT.EDU ("William Y-P. Lim") writes: >Why can't QIC 24 tapes be used for tape dumps using the Personal Iris >cartridge tape drive? As I understand it, the tapes used for QIC-150 are certified for the higher density. Sort of like using Double Density vs Single Density disks. Now, you can use the non-certified tape (disk), but you run the risk of data errors. If you put a non-written on QIC-24 tape in the QIC-150 drive, it will write on it. It seems that the drive checks tapes to see if they are written on during its tape loading. If it sees the tape as QIC-24, it won't let you rewrite it at QIC-150 to prevent you from the possibility of data loss. >> dan << -- ##################################################################### # CompuServe: >INTERNET:uunet.UU.NET!ki!dwatts Dan Watts # # UUCP : ...!uunet!ki!dwatts Ki Research, Inc. # ############### New Dimensions In Network Connectivity ##############   Received: from vmb.brl.mil by VMB.BRL.MIL id aa11423; 15 May 90 20:18 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab11056; 15 May 90 19:45 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa11032; 15 May 90 19:30 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa02543; 15 May 90 18:29 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA26193; Tue, 15 May 90 15:19:02 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 15 May 90 22:15:41 GMT From: Tim Hall Organization: Boston University Subject: Color Map/RGB mode and lighting Message-Id: <57352@bu.edu.bu.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL This is on 4D 240's and 220's running 3.2 I have written some code for the SGI to do arbitrary slicing of volumetric data. What it does is to intersect a mesh with the volume (a mesh point at each voxel crossing) and then display the mesh using the hardware mesh routines. The value at each mesh point is given by an interpolation of the volumetric data. I then use this value in a lookup table to determine the color at the mesh point. I use the hardware lighting and Gouraud shading when displaying the mesh. While I wouldn't call my method "wrong", I don't think it's entirely correct. I'm interpolating in the data space and the hardware is interpolating in RGB space. What I would like to happen is to be able to specify to the hardware the color map I'm using, have it interpolate the data values (color indices) and then use that color in the lighting. In other words, I'd like the hardware to think it's in color map mode until it's time to do the lighting. Is this possible? Am I making any sense? -Tim Hall tjh@bu-pub.bu.edu   Received: from vmb.brl.mil by VMB.BRL.MIL id ab11423; 15 May 90 20:18 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa11237; 15 May 90 20:04 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa11042; 15 May 90 19:31 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa02611; 15 May 90 18:44 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA27246; Tue, 15 May 90 15:34:48 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 15 May 90 22:00:38 GMT From: Paul Haeberli Organization: Silicon Graphics, Inc., Mountain View, CA Subject: convert ALIAS image files to IRIS image files Message-Id: <60275@sgi.sgi.com> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL /* * fromalias - * Convert an Alias image to an Iris image. * * To compile: * cc fromalias.c -o fromalias -limage -I/usr/include/gl * * Paul Haeberli - 1986 */ #include "image.h" short rbuf[4096]; short gbuf[4096]; short bbuf[4096]; typedef struct hdr { short xsize; short ysize; short xorg; short yorg; short bitsdeep; } hdr; hdr ihdr; main(argc,argv) int argc; char **argv; { IMAGE *image; FILE *infile; int xsize, ysize; int y, z; if(argc<3) { fprintf(stderr,"usage: fromalias aliasimage outimage.rgb\n"); exit(1); } if ((infile = fopen(argv[1],"r")) == NULL ) { fprintf(stderr,"fromalias: can't open %s\n",argv[1]); exit(1); } fread(&ihdr,sizeof(ihdr),1,infile); xsize = ihdr.xsize; ysize = ihdr.ysize; image = iopen(argv[2],"w",RLE(1),3,xsize,ysize,3); for(y=0; y>= 8; g = val&0xff; val >>= 8; b = val&0xff; val >>= 8; count = val&0xff; n += count; while(count--) { *rbuf++ = r; *gbuf++ = g; *bbuf++ = b; } } }   Received: from vmb.brl.mil by VMB.BRL.MIL id ac11423; 15 May 90 20:18 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab11237; 15 May 90 20:04 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab11042; 15 May 90 19:31 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa02615; 15 May 90 18:45 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA27316; Tue, 15 May 90 15:35:20 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 15 May 90 22:01:49 GMT From: Paul Haeberli Organization: Silicon Graphics, Inc., Mountain View, CA Subject: convert IRIS image files into ALIAS image format Message-Id: <60276@sgi.sgi.com> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL /* * toalias - * Convert an Iris image to an Alias image. * * To compile: * cc toalias.c -o toalias -limage -I/usr/include/gl * * Paul Haeberli - 1986 */ #include "image.h" char rbuf[4096]; char gbuf[4096]; char bbuf[4096]; unsigned long lbuf[4096]; typedef struct hdr { short xsize; short ysize; short xorg; short yorg; short bitsdeep; } hdr; hdr ihdr; main(argc,argv) int argc; char **argv; { IMAGE *image; int xsize, ysize; int y, z; char dname[128]; FILE *outfile; if(argc<3) { fprintf(stderr,"usage: toalias inimage.rgb aliasimage\n"); exit(1); } if ((image = iopen(argv[1],"r")) == NULL ) { fprintf(stderr,"toalias: can't open %s\n",argv[1]); exit(1); } if ((outfile = fopen(argv[2],"w")) == NULL ) { fprintf(stderr,"toalias: can't open %s\n",argv[2]); exit(1); } xsize = image->xsize; ysize = image->ysize; ihdr.xsize = xsize; ihdr.ysize = ysize; ihdr.xorg = 0; ihdr.yorg = ysize-1; ihdr.bitsdeep = 24; fwrite(&ihdr,sizeof(ihdr),1,outfile); for(y=0; y Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Convert MacPaint files to IRIS image file format. Message-Id: <60270@sgi.sgi.com> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL /* * frommac - * Convert macpaint image files to IRIS image files. * * Paul Haeberli - 1989 */ #include "image.h" #define MAXXSIZE 576 #define MAXYSIZE 720 short sbuf[4096]; unsigned char cbuf[72+256]; main(argc,argv) int argc; char *argv[]; { if(argc<3) { fprintf(stderr,"usage: frommac image.mac image.bw\n"); exit(1); } if(readmac(argv[1],argv[2],512)) exit(0); else if(readmac(argv[1],argv[2],512+128)) exit(0); else { fprintf(stderr,"frommac: bad macpaint file %s\n",argv[1]); exit(1); } } readmac(iname,oname,offset) char *iname, *oname; int offset; { FILE *inf; IMAGE *oimage; int i, y; inf = fopen(iname,"r"); if(!inf) { fprintf(stderr,"frommac: can't open input file %s\n",iname); exit(1); } oimage = iopen(oname,"w",RLE(1),2,MAXXSIZE,MAXYSIZE,1); for (i=0; i=8) { sptr[0] = val; sptr[1] = val; sptr[2] = val; sptr[3] = val; sptr[4] = val; sptr[5] = val; sptr[6] = val; sptr[7] = val; sptr += 8; n -= 8; } while(n--) *sptr++ = val; } } bitstorow(bits,sbuf,n) unsigned char *bits; short *sbuf; int n; { int i, val, nbytes; nbytes = ((n-1)/8)+1; for(i = 0; i Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Convert IRIS image files to MacPaint format . . . Message-Id: <60269@sgi.sgi.com> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL /* * tomac - * Convert an IRIS image file to macpaint format. * * to compile: * cc tomac.c -o tomac -limage -I/usr/include/gl * * Paul Haeberli - 1989 * */ #include "image.h" #define HEADERSIZE 512 #define MAXXSIZE 576 #define MAXYSIZE 720 short sbuf[4096]; unsigned char ibits[MAXXSIZE/8+20]; unsigned char pbits[MAXXSIZE/8+20]; char header[HEADERSIZE]; main(argc,argv) int argc; char *argv[]; { if(argc<3) { fprintf(stderr,"usage: tomac image.bw image.mac\n"); exit(1); } writemac(argv[1],argv[2]); } writemac(iname,oname) char *iname, *oname; { FILE *outf; IMAGE *iimage; int i, y, n; int xmargin, ymargin; int xsize, ysize; iimage = iopen(iname,"r"); if(!iimage) { fprintf(stderr,"tomac: can't open input file %s\n",iname); exit(1); } outf = fopen(oname,"w"); if(!outf) { fprintf(stderr,"tomac: can't open output file %s\n",iname); exit(1); } xsize = iimage->xsize; ysize = iimage->ysize; xmargin = (MAXXSIZE-xsize)/2.0; if(xmargin<0) xmargin = 0; ymargin = (MAXYSIZE-ysize)/2.0; if(ymargin<0) ymargin = 0; for (i=0; iymargin && y<(ymargin+ysize)) getrow(iimage,sbuf+xmargin,ysize-1-(y-ymargin),0); else setrow(sbuf,255,MAXXSIZE); rowtobits(sbuf,ibits,MAXXSIZE); n = packbits(ibits,pbits,MAXXSIZE); fwrite(pbits,n,1,outf); } iclose(iimage); fclose(outf); return 1; } packbits(ibits,pbits,nbits) unsigned char *ibits, *pbits; int nbits; { int bytes; unsigned char *sptr; unsigned char *ibitsend; unsigned char *optr = pbits; int nbytes, todo, cc, count; nbytes = ((nbits-1)/8)+1; ibitsend = ibits+nbytes; while(ibits127 ? 127:count; count -= todo; *optr++ = todo-1; while(todo--) *optr++ = *sptr++; } if(ibits == ibitsend) break; sptr = ibits; cc = *ibits++; while( (ibits128 ? 128:count; count -= todo; *optr++ = 257-todo; *optr++ = cc; } } return optr - pbits; } /* * row - * support for operations on image rows. * */ zerorow(sptr,n) short *sptr; int n; { bzero(sptr,n*sizeof(short)); } setrow(sptr,val,n) short *sptr; int val, n; { if(val==0) zerorow(sptr,n); else { while(n>=8) { sptr[0] = val; sptr[1] = val; sptr[2] = val; sptr[3] = val; sptr[4] = val; sptr[5] = val; sptr[6] = val; sptr[7] = val; sptr += 8; n -= 8; } while(n--) *sptr++ = val; } } bitstorow(bits,sbuf,n) unsigned char *bits; short *sbuf; int n; { int i, val, nbytes; nbytes = ((n-1)/8)+1; for(i = 0; i Organization: UBC Computing Services, Vancouver, B.C., Canada Subject: Query: lpr/lpd on SGI? Message-Id: <7869@ubc-cs.UUCP> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL I know that SGI has promised Berkeley lpr/lpd printing support for the next release, but we were wondering if anyone has ported it to the current (3.2) release of IRIX.   Received: from vmb.brl.mil by VMB.BRL.MIL id aa12039; 15 May 90 21:27 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa11909; 15 May 90 21:17 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa11891; 15 May 90 21:05 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa03257; 15 May 90 20:42 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA05549; Tue, 15 May 90 17:40:18 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 16 May 90 00:22:20 GMT From: Wiltse Carpenter Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Re: 1 vs. 4 mb dram Message-Id: <60299@sgi.sgi.com> References: <418@texhrc.UUCP> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <418@texhrc.UUCP>, mjz@texhrc.UUCP (Michael Zeitlin) writes: > > > Has anybody (or is it possible) upgraded their Personal Iris with > 4 mb. drams in place of the 1mb memory .... > The PI will support 4 mb drams with a few restrictions. The first is that all simms must be the same size; you cannot mix 1 mb and 4 mb parts. The second is that we introduced a bug in the proms last summer (I don't know the rev #, but basically all machines made since the introduction of the 4D25) whereby we would read from a memory location before ever writing to it which causes a parity error. This only happens in the 4 mb case and only on some simms (apparently some brands of simms happen to power up with the correct parity). SGI doesn't officially support 4 mb simms yet, so don't call the hotline (or me) if they don't work, but if you're willing to experiment you may find that they work on your machine. The parity problem mentioned above can be worked around through some gross mucking around with the prom's ``fill'' command, but that will be left as an exercise to the reader :-). -Wiltse   Received: from vmb.brl.mil by VMB.BRL.MIL id aa13516; 16 May 90 0:41 EDT Received: from vmb.brl.mil by VMB.brl.MIL id aa13347; 16 May 90 0:30 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa13330; 16 May 90 0:18 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa04108; 15 May 90 23:58 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA17726; Tue, 15 May 90 20:53:46 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 16 May 90 03:40:05 GMT From: Andrew Cherenson Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Re: Wanted BSD ports of finger,w, and msgs Message-Id: <60320@sgi.sgi.com> References: <9288@tank.uchicago.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <9288@tank.uchicago.edu> shin@tank.uchicago.edu (Shin Kurokawa) writes: >What I'd like to see in the future release of the IRIX are >better interface with "finger", and possibly the ports of "w" >and "msgs". W, uptime, the BSD version of finger and a variant of top are in the IRIX 3.3 release, which will be coming "real soon now."   Received: from vmb.brl.mil by VMB.BRL.MIL id ab01469; 16 May 90 13:38 EDT Received: from vmb.brl.mil by VMB.brl.MIL id ab00783; 16 May 90 13:21 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab00374; 16 May 90 13:06 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa06216; 16 May 90 9:45 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA16261; Wed, 16 May 90 06:36:41 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 15 May 90 18:40:14 GMT From: zaphod.mps.ohio-state.edu!samsung!umich!srvr1!news@tut.cis.ohio-state.edu Organization: caen Subject: feedback buffer parser on 4D/GTX Message-Id: <1990May15.184014.663@caen.engin.umich.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Could someone tell me if there is the feedback buffer parser on 4D/GTX machines? I have anonymous ftp access, please let me know the ftp sites which contain the library and related header files. Thanks in advance. --Jim Zeng @ University of Michigan   Received: from vmb.brl.mil by VMB.BRL.MIL id ab05153; 16 May 90 16:21 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab04872; 16 May 90 16:11 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab04864; 16 May 90 16:01 EDT Received: from cunyvm.cuny.edu by VGR.BRL.MIL id aa11664; 16 May 90 15:45 EDT Received: from MCCLB0.MED.NYU.EDU by CUNYVM.CUNY.EDU (IBM VM SMTP R1.2.2MX) with BSMTP id 6485; Wed, 16 May 90 15:42:34 EDT Received: from mcirps2.med.nyu.edu by MCCLB0.MED.NYU.EDU; Tue, 15 May 90 18:53 EDT Received: by mcirps2.med.nyu.edu (5.52/890607.SGI) (for @mcclb0.med.nyu.edu:info-iris@brl.arpa) id AA08788; Tue, 15 May 90 18:54:37 DSD Date: Tue, 15 May 90 18:54:37 DSD From: root%mcirps2.med.nyu.edu@cunyvm.cuny.edu Subject: Makefile rules... To: info-iris@BRL.MIL Message-id: <9005160154.AA08788@mcirps2.med.nyu.edu> X-Envelope-to: info-iris@brl.arpa I have a collection of files that comprise a program. I want to make a makefile that understands in a rule what follows: prog : prog1.o prog2.o prog3.o cc link stage commands... prog1.o : prog1.c prog1.h Makefile globals.h cc compile stage commands prog2.o : prog2.c prog2.h Makefile globals.h cc compile stage commands and so on for each program. I have tried to use the .c.o : $(OBJECT_LIST), How do it get make to recognize that a change in the .h file is also a .o dependency and not just the .c . As the program grows, I want to only put a new basename in the OBJECT_LIST,and have it know that that should make a dependent .o out of the .c and the .h and the globals.h file. OBJECT_LIST=prog1 prog2 prog2 .myrule : $(OBJECT_LIST) cc compile stage commands... or .myrule : $(???).c $(???).h global.h Makefile cc compile time command How do I define myrule ? Is there a builtin macro that I can use ? My programming style is to have a .h file for each .c file, and a global.h file. All parts of the program depend on the makefile, so a change in the makefile or the globals.h file should trigger a complete remake of prog. I have an explicit makefile that spells out each of the dependencies, but there must be an easier way. Some really strange bugs in a program can be traced to a bad makefile in that I have fixed the source code .h, but the .o do not get updated and the prog does. Any ideas out there ??? -- +-----------------------------------------------------------------------------+ | karron@nyu.edu Dan Karron | | . . . . . . . . . . . . . . New York University Medical Center | | 560 First Avenue \ \ Pager <1> (212) 397 9330 | | New York, New York 10016 \**\ <2> 10896 <3> | | (212) 340 5210 \**\__________________________________________ | +-----------------------------------------------------------------------------+   Received: from vmb.brl.mil by VMB.BRL.MIL id aa14283; 16 May 90 2:53 EDT Received: from vmb.brl.mil by VMB.brl.MIL id aa14196; 16 May 90 2:42 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab14122; 16 May 90 2:26 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa04669; 16 May 90 2:14 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA25649; Tue, 15 May 90 23:09:48 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 16 May 90 04:48:13 GMT From: Dave Olson Subject: Re: QIC 24 vs QIC 150 tapes Message-Id: <7776@odin.corp.sgi.com> References: <9005142352.AA14554@rice-chex>, <760@ki.UUCP> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL dwatts@ki.UUCP (Dan Watts) writes: | In article <9005142352.AA14554@rice-chex> wlim@AI.MIT.EDU ("William Y-P. Lim") writes: | >Why can't QIC 24 tapes be used for tape dumps using the Personal Iris | >cartridge tape drive? | As I understand it, the tapes used for QIC-150 are certified for the higher | density. Sort of like using Double Density vs Single Density disks. Now, | you can use the non-certified tape (disk), but you run the risk of data errors. | If you put a non-written on QIC-24 tape in the QIC-150 drive, it will write | on it. It seems that the drive checks tapes to see if they are written on | during its tape loading. If it sees the tape as QIC-24, it won't let you | rewrite it at QIC-150 to prevent you from the possibility of data loss. Sorry, this is incorrect. Few (if any) QIC-150 drives are capable of writing on QIC-24 cartridges, whether they are blank or not. This certainly includes the drives used in the PI. The PI drives CAN write in QIC-120 format on 600A style cartridges, yielding 120 Mb on a 600 foot tape. The QIC-24 tapes can be read. The QIC-150 cartridges are made to tighter mechanical tolerances than either the QIC 24 or QIC120 cartridges. -- Dave Olson Life would be so much easier if we could just look at the source code.   Received: from vmb.brl.mil by VMB.BRL.MIL id aa00095; 16 May 90 12:04 EDT Received: from vmb.brl.mil by VMB.brl.MIL id aa17396; 16 May 90 8:58 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa17330; 16 May 90 8:45 EDT Received: from prandtl.nas.nasa.gov by VGR.BRL.MIL id aa05615; 16 May 90 8:35 EDT Received: Wed, 16 May 90 05:33:55 -0700 from csduts1.lerc.nasa.gov by prandtl.nas.nasa.gov (5.61/1.2) Received: Wed, 16 May 90 08:35:21 EDT by csduts1.lerc.nasa.gov (5.51/LeRC(1.0)) Received: Wed, 16 May 90 09:01:05 EDT by avelon.lerc.nasa.gov (5.52/LeRC(1.0)) Date: Wed, 16 May 90 09:01:05 EDT From: Tony Facca Message-Id: <9005161301.AA06124@avelon.lerc.nasa.gov> To: shaginaw@neumann.squibb.com Subject: Re: TEK color emulation Cc: info-iris%brl.mil@prandtl.nas.nasa.gov > > Is anyone aware of any Tektronix color terminal emulators for the 4D? >Specifically, I'd like to find a 4107 emulator. Nevertheless, I'd like >any info anyone can provide on emulators for ANY of the color TEKs. > I have used TGRAF/SGI from Grafpoint and it does a good job. It will emulate 4107 and 4115/4125 Tektronix terminals. It runs on all the 4D's under the window manager. All the Tek hardware features are emulated. Call: Mason Killebrew 408-446-0666 While your making phone calls, get a hold of your SGI sales rep and ask for a copy of the most recent Geometry Partners Directory. It lists most all of the third party software available for the Iris's. -- ----------------------------------------------------------------------------- Tony Facca | fsfacca@avelon.lerc.nasa.gov | phone: 216-433-8318 ----------------------------------------------------------------------------- You are at Witt's end. Passages lead off in *all* directions.   Received: from vmb.brl.mil by VMB.BRL.MIL id aa01968; 16 May 90 13:48 EDT Received: from vmb.brl.mil by VMB.brl.MIL id ad00783; 16 May 90 13:22 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id al00374; 16 May 90 13:07 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa09919; 16 May 90 12:16 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA25180; Wed, 16 May 90 09:01:42 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 16 May 90 15:53:14 GMT From: George Hartzell Organization: MCD Biology, University of Colorado, Boulder Subject: need math board for 2400T. Message-Id: <21200@boulder.Colorado.EDU> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Does anyone out there have a math coprocessor board for an IRIS 2400Turbo that they would be willing to sell? How much? g. George Hartzell (303) 492-4535 MCD Biology, University of Colorado-Boulder, Boulder, CO 80309 hartzell@Boulder.Colorado.EDU ..!ncar!boulder!hartzell   Received: from vmb.brl.mil by VMB.BRL.MIL id ab01968; 16 May 90 13:48 EDT Received: from vmb.brl.mil by VMB.brl.MIL id ae00783; 16 May 90 13:22 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id am00374; 16 May 90 13:07 EDT Received: from NYU.EDU by VGR.BRL.MIL id aa10017; 16 May 90 12:27 EDT Received: from XP.PSYCH.NYU.EDU by cmcl2.NYU.EDU (5.61/1.34) id AA14198; Wed, 16 May 90 12:26:20 -0400 Message-Id: <9005161626.AA14198@cmcl2.NYU.EDU> Received: by xp.psych.nyu.edu; Wed, 16 May 90 12:23:27 EDT Date: Wed, 16 May 90 12:23:27 EDT From: "Steve Azueta (/w Aries" MMDF-Warning: Parse error in original version of preceding line at BRL.MIL To: info-iris@BRL.MIL Subject: hardware inventory table I am developing a graphics application that will be doing some hardware dependent stuff, and I would like to use getinvent() to check the graphics options at run-time. If you run the enclosed program and then do a hinv, you can begin to make assumptions about the classes, but the types escape me. Does anybody know where one can find definitions of the classes, types, and state of hardware? (Any help would be appreciated, Thanks in advance, ...) /* * hinvent.c ==> prints the information found in the hware inv table. */ #include #include main() { inventory_t *inv; do { inv = getinvent(); printf("class:\t\t%d\n", inv->class); printf("type:\t\t%d\n", inv->type); printf("controller:\t%d\n", inv->controller); printf("unit:\t\t%d\n", inv->unit); printf("state:\t\t%u\n\n", inv->state); } while (inv->next != NULL); endinvent(); } -- Steve Azueta azueta@xp.psych.nyu.edu   Received: from vmb.brl.mil by VMB.BRL.MIL id ac02614; 16 May 90 14:21 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ac01968; 16 May 90 14:01 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa01698; 16 May 90 13:40 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa10569; 16 May 90 13:30 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA29708; Wed, 16 May 90 10:18:25 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 16 May 90 17:07:51 GMT From: "Loren (Buck" MMDF-Warning: Parse error in original version of preceding line at BRL.MIL Organization: Computer Sciences Corporation @ NASA Goddard Space Flight Center Subject: 9-track tape drives Message-Id: <2060@dftsrv.gsfc.nasa.gov> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Well we are getting close to signing on the dotted line on getting a 9-track tape drive for our PI. Does any one out there have any experience with the Pertec FS-2000? It is a quad density auto-loader and is rack mountable. The lowest quote we have received is under $7000. B Cing U Buck Loren Buchanan | buck@drax.gsfc.nasa.gov | #include CSC, 1100 West St. | ...!ames!dftsrv!drax!buck | typedef int by Laurel, MD 20707 | (301) 497-2531 | void where_prohibited(by law){} CD International lists over 40,000 pop music CDs, collect the whole set.   Received: from vmb.brl.mil by VMB.BRL.MIL id ab02991; 16 May 90 14:39 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa02614; 16 May 90 14:21 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa02121; 16 May 90 13:56 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa10650; 16 May 90 13:44 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA00382; Wed, 16 May 90 10:27:32 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 16 May 90 17:00:06 GMT From: "Philip L. Karlton" Organization: Silicon Graphics, System Software Division Subject: Re: PEX Plans? Message-Id: <7783@odin.corp.sgi.com> References: <20674@shamash.cdc.com> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <20674@shamash.cdc.com>, dwb@hare.udev.cdc.com (dw block x-4621) writes: |> Does SGI have any plans on implementing PEX (PHIGS Extensions to X Windows)? SGI will provide a competitive and conforming implementation of PEX. The exact scheduling of the release is still undetermined. The GL will continue to provide the best performance for 3D interactive graphics. The nature of PHIGS (or any data base driven graphics standard) makes it extremely difficult for it to be optimized across many domains. PK -- Phil Karlton karlton@wpd.sgi.com   Received: from vmb.brl.mil by VMB.BRL.MIL id ac02991; 16 May 90 14:39 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab02614; 16 May 90 14:21 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab02121; 16 May 90 13:56 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa10656; 16 May 90 13:45 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA00339; Wed, 16 May 90 10:26:51 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 16 May 90 15:35:43 GMT From: Thant Tessman Organization: Silicon Graphics Inc. Subject: Re: Color Map/RGB mode and lighting Message-Id: <7779@odin.corp.sgi.com> References: <57352@bu.edu.bu.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <57352@bu.edu.bu.edu>, tjh@bu-pub.bu.edu (Tim Hall) writes: > This is on 4D 240's and 220's running 3.2 > [...] > > While I wouldn't call my method "wrong", I don't think it's entirely > correct. I'm interpolating in the data space and the hardware is > interpolating in RGB space. What I would like to happen is to > be able to specify to the hardware the color map I'm using, have it > interpolate the data values (color indices) and then use that > color in the lighting. In other words, I'd like the hardware to > think it's in color map mode until it's time to do the lighting. > > Is this possible? > > Am I making any sense? Yes, and yes. On the VGX you could just use a one dimensional texture and it would do the right thing for you, including interpolating between texture map entries when you zoom in, and averageing the texture map when you zoom out. :) However... The GTX graphics hardware will light in colormap mode. Knowing that, there is a trick you can do to to draw a surface that is both lit, and interpolated in colormap space. It involves using writemask in colormap mode and making two passes on the object. The color map is 12 bits big. Use the bottom 6 for lighting and the top 6 bits for color. This means you need to put 64 differently colored lighting ramps that are each 64 colormap entries long. Set up your lighting and material to use the colormap (see lmdef) as if your lighting ramp were from colormap entries 0 to 63. (I set it up so that there is a highlight at the top quarter of the ramp. It looks great.) Do writemask(0x00003f) and draw the object lit. Next, turn off lighting and do writemask(0x000fc0). When you draw the object the second time, specify the color at each vertex, but multiply the color by 64. Swap buffers and repeat. > > -Tim Hall > tjh@bu-pub.bu.edu thant Sorry I don't actually have some code to give you.   Received: from vmb.brl.mil by VMB.BRL.MIL id aa03571; 16 May 90 14:50 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa02991; 16 May 90 14:38 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa02629; 16 May 90 14:13 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa10717; 16 May 90 13:59 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA01953; Wed, 16 May 90 10:54:28 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 16 May 90 17:48:55 GMT From: David Pratt Organization: Naval Postgraduate School, Monterey CA Subject: Spaceball source code request Message-Id: <1004@cs.nps.navy.mil> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Does anybody out in netland have a public domain application that uses a spaceball? If you do would you mind sharing? Drop me a line on e-mail on how I can get it from you. Thanks, Dave pratt@cs.nps.navy.mil 408 646-2865 These aren't the Navy's opinions. I not even sure that they are mine.   Received: from vmb.brl.mil by VMB.BRL.MIL id aa03728; 16 May 90 15:00 EDT Received: from vmb.brl.mil by VMB.brl.MIL id ac00783; 16 May 90 13:21 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ai00374; 16 May 90 13:06 EDT Received: from enh.nist.gov by VGR.BRL.MIL id aa09211; 16 May 90 11:40 EDT Return-path: rbriber@poly1.nist.gov Received: from poly1.nist.gov by ENH.NIST.GOV; Wed, 16 May 90 11:36 EST Received: by poly1.nist.gov (5.52/890607.SGI) (for @enh.nist.gov:Info-Iris@BRL.MIL) id AA00977; Wed, 16 May 90 11:43:03 EDT Date: Wed, 16 May 90 11:43:03 EDT From: rbriber@poly1.nist.gov Subject: more named problems To: Info-Iris@BRL.MIL Message-Id: <9005161543.AA00977@poly1.nist.gov> We have just tried bringing up 'named' on our 4D/80GT and ran into a conflict with the windowserver. This is our first UNIX machine so forgive any ignorance on my part. I noticed after looking through the usergroup archives that there has been some discussion on problems between NeWS and 'named' in that unless the address 127.0.0.1 is resolved correctly then windows won't open. Ours problem is sort of backwards. Upon login windows open correctly but the various chests (windowchest, toolchest, systemchest and demochest) do not appear. Looking in SYSLOG the error is: "getsocktpeername: Can't find name for 129.6.154.1" and then a bunch of stuff from init.ps. This is the address of our computer! It should find its own name of poly1.nist.gov but is failing somewhere. Where does getsocketpeername look to resolve addresses? Does it ask 'named'? Does it look in /etc/hosts? We are using a set of files for named that were passed on to us from SGI by fax, but after sending them they haven't been much help and I haven't made any progress. The toolchests don't appear on initial login but will appear if invoked from the command line after login. Our goal is to resolve our own name and "localhost" on our machine and then go to one of our site nameservers (129.6.16.1, enh.nist.gov) for all other names. Copies of the relevant files are appended. Any help would be GREATLY appreciated. -------------/etc/hosts----------------------------- 127.1 localhost 129.6.154.1 poly1.nist.gov poly1 129.6.16.1 enh.nist.gov enh 129.6.154.254 pgw224.nist.gov pgw224 ---------------------------------------------------- -----------------/usr/etc/resolv.conf--------------- # Domain name resolver configuration file # domain nist.gov nameserver 0 ---------------------------------------------------- --------/usr/etc/named.d/named.boot----------------- ; ; @(#)named.boot.slave 1.13 (Berkeley) 87/07/21 ; ; boot file for secondary name server ; Note that there should be one primary entry for each SOA record. ; ; sortlist 10.0.0.0 directory /usr/etc/named.d ; type domain source host/file backup file cache . root.cache primary 0.0.127.IN-ADDR.ARPA localhost.rev primary 1.154.6.129.IN-ADDR.ARPA nist.rev ; ---------------------------------------------------- -------/usr/etc/named.d/root.cache------------------ ; ; @(#)root.cache 1.13 (Berkeley) 88/04/15 ; ; Initial cache data for root domain servers. ; . 99999999 IN NS enh.nist.gov. ; ; Prep the cache (hotwire the addresses). Order does not matter ; enh.nist.gov. 99999999 IN A 129.6.16.1 ---------------------------------------------------- ----------/usr/etc/named.d/localhost.rev------------ ; ; localhost.rev -- PTR record for 127.1 ; @ IN SOA poly1.nist.GOV. root.poly1.nist.GOV. ( 1.1 ; Serial 3600 ; Refresh 300 ; Retry 3600000 ; Expire 14400 ) ; Minimum IN NS poly1.nist.gov. 1 IN PTR localhost. --------------------------------------------------- ----------/usr/etc/named.d/nist.rev---------------- ; ; @(#)named.rev 1.1 (Berkeley) 86/02/05 ; @ IN SOA poly1.nist.gov. root.poly1.nist.gov. ( 1.1 ; Serial 10800 ; Refresh 3 hours 3600 ; Retry 1 hour 3600000 ; Expire 1000 hours 86400 ) ; Minimum 24 hours IN NS poly1.nist.gov. poly1.nist.gov. IN A 129.6.154.1 1.154.6.129 IN PTR poly1.nist.gov. --------------------------------------------------- -- ---------------------------------------------------------------------------- | Adios Amoebas, | | | | "I've tried and I've tried and I'm still mystified, | | Robert Briber | I can't do it anymore and I'm not satisfied." | | 224/B210 | --Elvis | | NIST |------------------------------------------------------| | Gaithersburg, MD | rbriber@enh.nist.gov (Internet) | | 20899 USA | rbriber@poly1.nist.gov (Internet) | |(301) 975-6775(voice)| rbriber@nbsenh (Bitnet) | |(301) 975-2128 (fax) | | ----------------------------------------------------------------------------   Received: from vmb.brl.mil by VMB.BRL.MIL id aa05153; 16 May 90 16:21 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa04872; 16 May 90 16:10 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa04864; 16 May 90 16:00 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa11622; 16 May 90 15:44 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA08343; Wed, 16 May 90 12:30:56 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 16 May 90 18:26:35 GMT From: Thilaka Sumanaweera Organization: /user/sumane/.organization Subject: 4Meg SIMMS on IRIS 4D/25 Message-Id: Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Ok. As I promised, here are the results of the 4Meg SIMMS installation test on our SGI IRIS 4D/25. We purchased 32Megs worth 4Meg-SIMMS for $475 a piece from Impediment Inc. We tried to use these 4Meg SIMMS with the 1Meg SIMMS that came with the machine. When the machine came up it reported having 0Megs of memory. So this means you cannot mix 1Megs and 4Megs. Our machine seems to be happy with the 4Meg SIMMS alone. This new memory has doubled the image display refreshing. (We display 4 512x512 images on the screen at a time.) The prices of 4Meg SIMMS are supposed to hit $300 range by the end of summer. (i.e. about the same price as 1Megs). Does anybody know if 16Meg SIMMS are going to be available soon? Thilaka   Received: from vmb.brl.mil by VMB.BRL.MIL id aa05497; 16 May 90 16:50 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa05290; 16 May 90 16:40 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa05258; 16 May 90 16:28 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa12438; 16 May 90 16:16 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA10910; Wed, 16 May 90 13:09:55 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 16 May 90 19:42:11 GMT From: "Jeff P. M. Hultquist" Organization: NAS - Applied Research Office, NASA Ames Subject: Re: PEX Plans? Message-Id: References: <20674@shamash.cdc.com>, <7783@odin.corp.sgi.com> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <7783@odin.corp.sgi.com> karlton@sgi.com (Philip L. Karlton) writes: > > From: karlton@sgi.com (Philip L. Karlton) > Organization: Silicon Graphics, System Software Division > > |> Does SGI have any plans on implementing PEX (PHIGS Extensions to X)? > > SGI will provide a competitive and conforming implementation of PEX. The > exact scheduling of the release is still undetermined. ... but we _can_ assume that this will happen sometime after they provide a "competitive and conforming implementation" of X-windows! -- jeff -- -- Jeff Hultquist hultquis@nas.nasa.gov NASA - Ames Research Center (415) 604-4970 Disclaimer: "I am not a rocket scientist."   Received: from vmb.brl.mil by VMB.BRL.MIL id aa06667; 16 May 90 19:21 EDT Received: from vmb.brl.mil by VMB.brl.MIL id aa06454; 16 May 90 18:47 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa06404; 16 May 90 18:28 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa13581; 16 May 90 18:17 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA18367; Wed, 16 May 90 15:09:20 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 16 May 90 21:38:03 GMT From: John D Mccalpin Organization: College of Marine Studies, Univ. of Delaware Subject: Interactive Volume Modelling? Message-Id: <6483@vax1.acs.udel.EDU> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL I just read an article in "Oceanography" magazine on volume visualization. The product used was called "Interactive Volume Modelling", from an organization called "Dynamic Graphics, Inc". Anybody have any pointers on how I can find these folks? Of course, the software was run on a Silicon Graphics 4D machine.... -- John D. McCalpin mccalpin@vax1.udel.edu Assistant Professor mccalpin@delocn.udel.edu College of Marine Studies, U. Del. mccalpin@scri1.scri.fsu.edu   Received: from vmb.brl.mil by VMB.BRL.MIL id aa06779; 16 May 90 19:57 EDT Received: from vmb.brl.mil by VMB.brl.MIL id aa06556; 16 May 90 19:11 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa06535; 16 May 90 18:45 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa13636; 16 May 90 18:32 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA18969; Wed, 16 May 90 15:18:53 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 16 May 90 21:17:42 GMT From: Jerre Bowen Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Re: signals on multiprocessor Message-Id: <7801@odin.corp.sgi.com> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL >From: lamy@cs.utoronto.ca (Jean-Francois Lamy) >Newsgroups: comp.sys.sgi >Subject: signals on multi-processor > >Assume that one has a process 1 that does > > setup HUP signal handler that sets a flag > grab lock > if HUP flag is set reload data files that were cached > release lock > >and a process 2 that does > > grab lock > mess up with data files process 1 has cached > kill -HUP process 2 > release lock > >The two processes are racing each other. In the picture I have of traditional >Unix implementations, if process 2 seizes the lock first (I have in mind the >simple-minded locks where the process creates a link to a file), the signal >will be sent by setting a bit somewhere in 1's data structures, before process >2 releases the lock and the handler in process 1 will be triggered before 1 >ever tests for the flag, by virtue of process 1 having to do system calls to >acquire to lock before it tests for said flag (the act of locking requires a >system call, at which point the bit set in the process data structures will >get attention, assuming it did not before). > >So in traditional Unix this would be a workable if ugly way to warn a process >that it's caches have been invalidated, or some such. I am fully aware that >there are other ways of doing this that do not involve signals, but for the >purpose of this discussion, could anyone tell me whether Irix delivers signals >in a way that guarantees that process 2 will have processed the signal sent by >process 1 before it gets to test for the flag? Or is signal delivery more >asynchronous, meaning that the signal could float in space for a while and >get delivered only after 1 has tested the flag and assumed the caches it >held were valid? Finally, should this work in Irix as in other Unixes, is >this mandated in some standard or just an accident? > >Jean-Francois Lamy lamy@cs.utoronto.ca, uunet!cs.utoronto.ca!lamy >Department of Computer Science, University of Toronto, Canada M5S 1A4 The above algorithm and your explanation in the subsequent paragraphs are too abreviated to make clear to me exactly what you're trying to do (e.g. "process 2 will have processed the signal sent by process 1": the algo doesn't say anything about process 1 sending signals), and it doesn't sound like it was meant to be a concise algorithm. Rearranging the order of the lines of the algorithm will make the race situation better, especially on single processor machines, but the fact is that no guarantees of event-order can be made about signal delivery, *especially* on mp machines, in any version of BSD or AT&T UNIX. Other methods of synchronization are available as you say (indeed, the above algorithm may be simplified), and coding order can reduce or eliminate races, but signals are by definition extremely asynchronous. If an implementation of the above algorithm (as I understand it) worked, it was on a single-processor, and it was an accident. If this doesn't answer your question, you may want to email me (bowen@sgi.com). Jerre Bowen   Received: from vmb.brl.mil by VMB.BRL.MIL id ab06779; 16 May 90 19:57 EDT Received: from vmb.brl.mil by VMB.brl.MIL id ab06667; 16 May 90 19:36 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa06619; 16 May 90 19:13 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa13617; 16 May 90 18:30 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA18693; Wed, 16 May 90 15:14:25 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 16 May 90 21:39:58 GMT From: John D Mccalpin Organization: College of Marine Studies, Univ. of Delaware Subject: RGB->NTSC converters Message-Id: <6484@vax1.acs.udel.EDU> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Does anyone out there have a good feel for the current market for RGB->NTSC converters? Used units are especially interesting. I do a fair bit of live animation on my 4D-25TG and would like my own video setup.... -- John D. McCalpin mccalpin@vax1.udel.edu Assistant Professor mccalpin@delocn.udel.edu College of Marine Studies, U. Del. mccalpin@scri1.scri.fsu.edu   Received: from vmb.brl.mil by VMB.BRL.MIL id aa07150; 16 May 90 20:59 EDT Received: from vmb.brl.mil by VMB.brl.MIL id aa07046; 16 May 90 20:49 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa07036; 16 May 90 20:37 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa14076; 16 May 90 20:30 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA26537; Wed, 16 May 90 17:18:37 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 16 May 90 23:11:05 GMT From: "Lee Ward CS.DEPT" Organization: University of New Mexico, Albuquerque Subject: Help with "Ethernet init failed" from dog Message-Id: <1990May16.231105.17956@unmvax.cs.unm.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL We have two old SGI Iris machines. One is a 2400 and the other is a 2400T. We just got the 2400T so thought it would be fun to try out /usr/people/demos/dog. Unfortunately, both machines deliver the error message: Ethernet init failed I called SGI and all they say is "/etc/services" is messed up. I have both in the local /etc/services and in the YP masters: sgi-dogfight 5130/udp dogfight 3011/udp I wrote a program that looks up these services and it succeeded, no problem. The people I have talked with at SGI say they have no clue as to why the error is reported or what it means. By the way, a strings of the dog program on the newest SGI machines reveals that they contain the same error message. So, if you've seen it on a newer box the fix is probably applicable for me too. The turbo machine is running operating system release number GL2-W3.5r1 and dog itself is version 2.3 So the question is: Can somebody tell me what "Ethernet init failed" means or what else I might do to get past it's little problem? --Lee   Received: from vmb.brl.mil by VMB.BRL.MIL id aa07335; 16 May 90 21:37 EDT Received: from vmb.brl.mil by VMB.brl.MIL id aa07261; 16 May 90 21:17 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa07186; 16 May 90 21:03 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa14175; 16 May 90 21:00 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA27518; Wed, 16 May 90 17:34:35 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 17 May 90 00:09:54 GMT From: Dan Watts Organization: Ki Research, Inc. Derry, NH Subject: Re: QIC 24 vs QIC 150 tapes Message-Id: <762@ki.UUCP> References: <9005142352.AA14554@rice-chex>, <760@ki.UUCP>, <7776@odin.corp.sgi.com> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <7776@odin.corp.sgi.com> olson@anchor.esd.sgi.com (Dave Olson) writes: >dwatts@ki.UUCP (Dan Watts) writes: >| << text deleted >> >| If you put a non-written on QIC-24 tape in the QIC-150 drive, it will write >| on it. >| << text deleted >> >Sorry, this is incorrect. Few (if any) QIC-150 drives are capable >of writing on QIC-24 cartridges, whether they are blank or not. Sorry about the incorrect answer. I was just working from actual experience. I've put QIC-24 tapes in my QIC-150 drive and used them just fine (though once I found the part number for QIC-150 tapes, I quickly changed). At least I thought they were QIC-24. I was using DC600 tapes that I had been using on a Sun and Tektronix system. I've since changed to the DC6150's. -- ##################################################################### # CompuServe: >INTERNET:uunet.UU.NET!ki!dwatts Dan Watts # # UUCP : ...!uunet!ki!dwatts Ki Research, Inc. # ############### New Dimensions In Network Connectivity ##############   Received: from vmb.brl.mil by VMB.BRL.MIL id ab09240; 17 May 90 6:04 EDT Received: from vmb.brl.mil by VMB.brl.MIL id ab09127; 17 May 90 5:33 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab09110; 17 May 90 5:14 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa15606; 17 May 90 5:00 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA23317; Thu, 17 May 90 01:18:10 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 17 May 90 13:11:29 GMT From: Phil Dench Organization: Curtin University of Technology, Maths & Comp Sc Subject: object routines Message-Id: <136@cutmcvax.OZ> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Thanks to all those who replied to my query about the object routines. Brian McClendon from SG told me that all usefull routines (which ones aren't?) will be supported in DLs from 3.3 on and that there will be no performance loss if you use the shared library. Phil   Received: from vmb.brl.mil by VMB.BRL.MIL id aa10502; 17 May 90 8:08 EDT Received: from vmb.brl.mil by VMB.brl.MIL id aa09854; 17 May 90 7:47 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa09846; 17 May 90 7:36 EDT Received: from life.ai.mit.edu by VGR.BRL.MIL id aa16027; 17 May 90 7:25 EDT Received: from rice-chex (rice-chex.ai.mit.edu) by life.ai.mit.edu (4.1/AI-4.10) id AA03363; Thu, 17 May 90 07:23:22 EDT From: "William Y-P. Lim" Received: by rice-chex (4.1/AI-4.10) id AA04931; Thu, 17 May 90 07:21:47 EDT Date: Thu, 17 May 90 07:21:47 EDT Message-Id: <9005171121.AA04931@rice-chex> To: info-iris@BRL.MIL Subject: sendmail.cf replies I got many replies to my query on "messing with sendmail.cf" to make our Personal Iris send and receive mail from our local network. Before posting the replies to info-iris, I would like to know if those people who have kindly replied my message mind if I post their replies on info-iris. There are too many of these kind hearted souls for me to ask individually. If there are no objections, I'll post the replies tonight. Thanks to all who have replied. Willie   Received: from vmb.brl.mil by VMB.BRL.MIL id aa11291; 17 May 90 9:01 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab10502; 17 May 90 8:19 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa10441; 17 May 90 8:04 EDT Received: from aero4.larc.nasa.gov by VGR.BRL.MIL id aa16144; 17 May 90 7:51 EDT Received: Thu, 17 May 90 07:50:35 EDT by aero4.larc.nasa.gov (5.52/5.6) Date: Thu, 17 May 90 07:50:35 EDT From: "Brent L. Bates AAD/TAB MS361 x42854" Message-Id: <9005171150.AA11621@aero4.larc.nasa.gov> To: unmvax!carina.unm.edu!lee@ucbvax.berkeley.edu Subject: Re: Help with "Ethernet init failed" from dog Cc: info-iris@BRL.MIL Since you said your /etc/services is correct, I know of only one other problem that gives that error. We had the same problem when we first got our 3130. Dog was compiled using the XNS libraries instead of TCP. If you just want to run it with out trying to access the network, use the "-i infile" and/or "-o outfile" options. -i lets you read in an "airshow" file, -o lets you create one. If you run dog with just -o your flight trajectory and missile fire events are recorded in the outfile. You can then read that file in later. When the end of the input file is reached, the file is rewond and started over. If you cycle through this process several times you can have multiple planes flying. -- Brent L. Bates NASA-Langley Research Center M.S. 361 Hampton, Virginia 23665-5225 (804) 864-2854 E-mail: blbates@aero4.larc.nasa.gov or blbates@aero2.larc.nasa.gov   Received: from vmb.brl.mil by VMB.BRL.MIL id aa11742; 17 May 90 9:43 EDT Received: from vmb.brl.mil by VMB.brl.MIL id ab11291; 17 May 90 9:10 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa11146; 17 May 90 8:51 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa16517; 17 May 90 8:45 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA05996; Thu, 17 May 90 05:31:56 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 17 May 90 08:33:43 GMT From: "Tim Everett,," Organization: Philips Research Laboratories, Eindhoven the Netherlands Subject: Using the audio device without getting audible "ticks" Message-Id: <1658@prles2.prl.philips.nl> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL I hope someone can help, I want to be able to get continuous sound out of the audio device of a PI (running 3.2). The trouble is I need to be able to change the sample about every 10th of a second, and up to now I can't get rid of the tick as the change takes place. (even when the individual blocks come from a continous sampled source) I guess it has something to do with context switches. I have tried upping the priority to NDPHIMAX combined with plock(PROCLOCK) which stops the playback being interrupted by other system activity, but doesn't get rid of the ticks. The solution I'm looking for should not take up too much CPU time, ie. something along the lines of ... given a collection of audio samples (all 1/10th of a second long) while (TRUE) { write(audio_device,selected_sample,length); // another process selects a new sample. } Any suggestions? Tim Everett Philips Research Labs. Eindhoven The Netherlands email: tim@vangogh.prl.philips.nl   Received: from vmb.brl.mil by VMB.BRL.MIL id ab11742; 17 May 90 9:43 EDT Received: from vmb.brl.mil by VMB.brl.MIL id aa11464; 17 May 90 9:32 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa11354; 17 May 90 9:03 EDT Received: from prandtl.nas.nasa.gov by VGR.BRL.MIL id aa16525; 17 May 90 8:47 EDT Received: Thu, 17 May 90 05:45:58 -0700 from csduts1.lerc.nasa.gov by prandtl.nas.nasa.gov (5.61/1.2) Received: Thu, 17 May 90 08:47:26 EDT by csduts1.lerc.nasa.gov (5.51/LeRC(1.0)) Received: Thu, 17 May 90 09:00:55 EDT by avelon.lerc.nasa.gov (5.52/LeRC(1.0)) Date: Thu, 17 May 90 09:00:55 EDT From: Tony Facca Message-Id: <9005171300.AA12704@avelon.lerc.nasa.gov> To: rbriber@poly1.nist.gov Subject: Re: more named problems Cc: info-iris%brl.mil@prandtl.nas.nasa.gov > >We have just tried bringing up 'named' on our 4D/80GT and ran into a conflict >with the windowserver. This is our first UNIX machine so forgive any > [ stuff deleted ] >"getsocktpeername: Can't find name for 129.6.154.1" >and then a bunch of stuff from init.ps. >This is the address of our computer! It should find its own name of >poly1.nist.gov but is failing somewhere. Where does getsocketpeername look to [ stuff deleted ] We have had the same problem. The short answer is that the /etc/hosts file is NEVER consulted for information about addresses once the /usr/etc/resolv.conf is created. What most likely happened in your case was that the 127.0.0.1 was found in the nameserver but your NEW host (129.6.154.1) has not yet been added to it. This is supposed to be changed in the next release of the software. You will be able to use the host table first and then the nameserver (or vice versa). What we have done is to write a short script which allows the sysadmin to toggle the use of the nameserver. This way, if you are using a remote machine as the nameserver and it becomes unavailable, you can switch to using the host table. I suppose a copy of this is in the archives since I've posted it a couple of times already. If not, let me know and I'll get you a copy. Hope this helps.. -- ----------------------------------------------------------------------------- Tony Facca | fsfacca@avelon.lerc.nasa.gov | phone: 216-433-8318 ----------------------------------------------------------------------------- You are at Witt's end. Passages lead off in *all* directions.   Received: from vmb.brl.mil by VMB.BRL.MIL id ac11742; 17 May 90 9:43 EDT Received: from vmb.brl.mil by VMB.brl.MIL id ac11464; 17 May 90 9:32 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa11447; 17 May 90 9:18 EDT Received: from Princeton.EDU by VGR.BRL.MIL id aa16614; 17 May 90 9:06 EDT Received: from acm.Princeton.EDU by Princeton.EDU (5.61++/2.37/relay) id AA22946; Thu, 17 May 90 09:04:01 -0400 Received: from taylor.Princeton.EDU by acm.Princeton.EDU (5.61/1.102) id AA19372; Thu, 17 May 90 09:04:11 -0400 Received: by taylor.Princeton.EDU (5.61/1.95) id AA07945; Thu, 17 May 90 09:04:08 -0400 Date: Thu, 17 May 90 09:04:08 -0400 From: ams@acm.princeton.edu Message-Id: <9005171304.AA07945@taylor.Princeton.EDU> To: info-iris@BRL.MIL Subject: Re: video options for 4D50, 4D85 There are really a lot of options to consider when you wish to make movies from your Iris. The big contenders as I see it are: VideoCreator/Genlock boards Real Time Scan Converters Abekas style boxes Direct Film Recorders The Applied Math Program chose a Real Time Scan converter from RGB Technologies complimented by an SVHS Lyon Lamb minivas animation controller and a Panasonic SVHS video recorder. This system gives us excellent results for both VHS and SVHS recording, without being locked into doing movies only from SGI screens. We looked at three vendors when we were choosing the scan converter (we saw just about everyone at SIGGRAPH '89). These were Lyon Lamb, RGB Technologies and Folsom. We liked the LL, but at $25,000 it was too much money. Folsom produced unacceptable results. RGB was a good trade off--excellent quality, reduced cost $11K, but less sophisticated than the LL. Were we to do this again, I think we would still choose RGB. I am told they have improved our model so their current offering should be even better. Here is some vendor information: Microdisk: RGB dealer, good prices on professional tape decks and monitors. 609/486-0330 RGB Technologies: 415/848-0180 Lyon Lamb: 818/843-4831 Folsom Research: 916/983-1500 WEB Associates: Folsom dealer, as well as Focus Film Recorder products. 215/630-8999 --ams _________________________________________________________________________ Andrew Simms, System Manager Program in Applied and Computational Mathematics Princeton University 218 Fine Hall, Washington Road ams@acm.princeton.edu Princeton, NJ 08544 609/258-5324 609/258-1054 (fax)   Received: from vmb.brl.mil by VMB.BRL.MIL id aa12036; 17 May 90 10:04 EDT Received: from vmb.brl.mil by VMB.brl.MIL id ab11464; 17 May 90 9:32 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab11354; 17 May 90 9:03 EDT Received: from lvax.brl.mil by VGR.BRL.MIL id aa16527; 17 May 90 8:48 EDT Date: 17 May 90 08:47:00 EDT From: "Earl N. Ferry" Subject: RE: more named problems To: info-iris Message-ID: <9005170848.aa16527@VGR.BRL.MIL> In Message-Id: <9005161543.AA00977@poly1.nist.gov> Robert Briber writes: >>We have just tried bringing up 'named' on our 4D/80GT and ran into a conflict >>with the windowserver. This is our first UNIX machine so forgive any >>ignorance on my part. I noticed after looking through the usergroup archives >>that there has been some discussion on problems between NeWS and 'named' in >>that unless the address 127.0.0.1 is resolved correctly then windows won't >>open. Ours problem is sort of backwards. Upon login windows open correctly >>but the various chests (windowchest, toolchest, systemchest and demochest) >>do not appear. Looking in SYSLOG the error is: >>"getsocktpeername: Can't find name for 129.6.154.1" >>and then a bunch of stuff from init.ps. >>This is the address of our computer! It should find its own name of >>poly1.nist.gov but is failing somewhere. Where does getsocketpeername look to >>resolve addresses? Does it ask 'named'? Does it look in /etc/hosts? >>We are using a set of files for named that were passed on to us from SGI by >>fax, but after sending them they haven't been much help and I haven't made >>any progress. The toolchests don't appear on initial login but will appear >>if invoked from the command line after login. Our goal is to resolve our >>own name and "localhost" on our machine and then go to one of our site >>nameservers (129.6.16.1, enh.nist.gov) for all other names. >>Copies of the relevant files are appended. Any help would be GREATLY >>appreciated. [Stuff Deleted...] We had a power failure Tuesday (15-May-90) and when our systems (2 4D/20G's) came back up they had the exact same problems you described. After spending some time trying to figure out what was wrong and checking all the files you mentioned I could find nothing wrong. So I started looking for a common link between the two systems since it is VERY uncommon for a power failure to affect two identical systems in exactly the same way as far as possible file system corruption goes, and the only thing I could come up with was the network. I had our network people reset our gateway and the problems cleared up immediately! I'm still running today with NO problems. This might not be the same situation you are in, but sometimes any help or suggestions can help. Earl Ferry, Jr. US Army Ballistic Research Laboratory Aberdeen Proving Grounds, MD 21005 EMAIL: eferry@brl.mil Phone (301)278-2916   Received: from vmb.brl.mil by VMB.BRL.MIL id aa14999; 17 May 90 13:14 EDT Received: from vmb.brl.mil by VMB.brl.MIL id aa14429; 17 May 90 12:43 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa14400; 17 May 90 12:23 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa17695; 17 May 90 12:16 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA19195; Thu, 17 May 90 09:09:37 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 17 May 90 12:22:35 GMT From: Jeff Hanson Organization: NASA/Lewis Research Center, Cleveland Subject: Re: Interactive Volume Modelling? Message-Id: <1990May17.122235.13440@eagle.lerc.nasa.gov> References: <6483@vax1.acs.udel.EDU> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Dynamic Graphics 2855 Telegraph Ave, Suite 405 Berkely CA 94705 415-845-8180 They make several related products, see pages 35-37 of Fall 1989 Geometry Partners directory. I have no experience with these products. -- *-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-* \ / \ / \ / \ / \ / \ / Jeff Hanson \ / \ / \ / \ / \ / \ / * ViSC: Better * tohanson@gonzo.lerc.nasa.gov * * * * * * / \ / \ Science / \ / \ NASA Lewis Research Center / \ / \ Through / \ / \ * * * * * * * Cleveland, Ohio 44135 * * * Pictures * * \ / \ / \ / \ Telephone - (216) 433-2284 Fax - (216) 433-2182 \ / \ / \ / *-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*   Received: from vmb.brl.mil by VMB.BRL.MIL id aa02470; 17 May 90 18:06 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa01540; 17 May 90 17:15 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa01500; 17 May 90 17:06 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa00227; 17 May 90 15:47 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA01254; Thu, 17 May 90 12:18:13 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 17 May 90 18:09:07 GMT From: Alex Woo RAC Organization: NASA Ames Research Center, Mtn Vw CA 94035 Subject: Imagetools Manual Wanted Message-Id: <6180@amelia.nas.nasa.gov> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL A number of SGI's imagetools have been recently posted to the network. Is there a nice manual somewhere that explains all of them? There seems to be a few programs in /usr/sbin with a few scattered man pages. A pointer to a collection of man pages would be nice. Thanks. ====================================================================== Alex Woo, MS 227-2 | woo@pioneer.arc.nasa.gov NASA Ames Research Center | woo@ames-nas.arpa Moffett Field, CA 94035 | {seismo,topaz,lll-crg,ucbvax}! Phone: (415) 694-6010 | ames!pioneer!woo ====================================================================== {hplabs,hao,att,decwrl,allegra,tektronix,menlo70}!ames!pioneer!woo ======================================================================   Received: from vmb.brl.mil by VMB.BRL.MIL id af02470; 17 May 90 18:15 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa02127; 17 May 90 17:55 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa02039; 17 May 90 17:42 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa00690; 17 May 90 17:32 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA08439; Thu, 17 May 90 14:10:55 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 17 May 90 21:03:42 GMT From: Gautam Mehrotra Organization: National Center for Supercomputing Applications Subject: Help !! Message-Id: <1990May17.210342.17949@ux1.cso.uiuc.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL WHile running a program, I got this message in my console window : NOTICE: File table overflow The program failed to find some files ( which were there ). The disks were not full and I could see no other abnormality ..... I reran the program immediately afterwards and it ran just fine. Any clues to whats happening ? Why ? cheers, gautam PS : The machine was a 4D/240GTX ( 128 Meg memory ). Gautam Mehrotra, National Center for Supercomputing Applications, Champaign.   Received: from vmb.brl.mil by VMB.BRL.MIL id ag02470; 17 May 90 18:15 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab02127; 17 May 90 17:56 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab02039; 17 May 90 17:42 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa00692; 17 May 90 17:32 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA08496; Thu, 17 May 90 14:11:22 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 17 May 90 19:13:29 GMT From: Al Broderick Organization: Rutgers Univ., New Brunswick, N.J. Subject: SGI include file problem? Message-Id: Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL I have a SGI 4D/70G running IRIX 3.2.1. I have a few questions. I get this error message when I compile (a program that compiled cleanly on 3.1): cpp: error /usr/include/bsd/sys/ioctl.h:9: Can't find include file net/soioctl.h cpp: error /usr/include/bsd/sys/ioctl.h:10: Can't find include file sys/ttychars.h The offending code in /usr/include/bsd/sys/ioctl.h says this: #include "../../sys/ioctl.h" #include #include Should it really look like this? #include "../../sys/ioctl.h" #include #include I have not tried this on my machine yet because I am hoping that I am doing something wrong. Could someone from SGI please advise? Thank you. Alfred J. Broderick broderic@topaz.rutgers.edu   Received: from vmb.brl.mil by VMB.BRL.MIL id aa02669; 17 May 90 18:26 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ac02470; 17 May 90 18:15 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab02262; 17 May 90 17:53 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa00737; 17 May 90 17:45 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA09981; Thu, 17 May 90 14:34:04 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 17 May 90 20:14:09 GMT From: Dave Olson Subject: Re: QIC 24 vs QIC 150 tapes Message-Id: <7836@odin.corp.sgi.com> References: <760@ki.UUCP>, <7776@odin.corp.sgi.com>, <762@ki.UUCP> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL dwatts@ki.UUCP (Dan Watts) writes: | In article <7776@odin.corp.sgi.com> olson@anchor.esd.sgi.com (Dave Olson) writes: | >dwatts@ki.UUCP (Dan Watts) writes: | >| << text deleted >> | >| If you put a non-written on QIC-24 tape in the QIC-150 drive, it will write | >| on it. | >| << text deleted >> | >Sorry, this is incorrect. Few (if any) QIC-150 drives are capable | >of writing on QIC-24 cartridges, whether they are blank or not. | Sorry about the incorrect answer. I was just working from actual experience. | I've put QIC-24 tapes in my QIC-150 drive and used them just fine (though | once I found the part number for QIC-150 tapes, I quickly changed). At | least I thought they were QIC-24. I was using DC600 tapes that I had been | using on a Sun and Tektronix system. I've since changed to the DC6150's. DC600 tapes are usually actively QIC-120 tapes, not QIC-24. -- Dave Olson Life would be so much easier if we could just look at the source code.   Received: from vmb.brl.mil by VMB.BRL.MIL id ab02669; 17 May 90 18:26 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ad02470; 17 May 90 18:15 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ac02262; 17 May 90 17:53 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa00740; 17 May 90 17:45 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA09948; Thu, 17 May 90 14:33:48 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 17 May 90 17:40:31 GMT From: Jesse Rendleman Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Re: more named problems Message-Id: <7823@odin.corp.sgi.com> References: <9005161543.AA00977@poly1.nist.gov> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <9005161543.AA00977@poly1.nist.gov> rbriber@POLY1.NIST.GOV writes: > >We have just tried bringing up 'named' on our 4D/80GT and ran into a conflict >with the windowserver. This is our first UNIX machine so forgive any ... >--------/usr/etc/named.d/named.boot----------------- >sortlist 10.0.0.0 >directory /usr/etc/named.d >; type domain source host/file backup file >cache . root.cache >primary 0.0.127.IN-ADDR.ARPA localhost.rev >primary 1.154.6.129.IN-ADDR.ARPA nist.rev >---------------------------------------------------- ... >----------/usr/etc/named.d/nist.rev---------------- >@ IN SOA poly1.nist.gov. root.poly1.nist.gov. ( > 1.1 ; Serial > 10800 ; Refresh 3 hours > 3600 ; Retry 1 hour > 3600000 ; Expire 1000 hours > 86400 ) ; Minimum 24 hours > IN NS poly1.nist.gov. >poly1.nist.gov. IN A 129.6.154.1 >1.154.6.129 IN PTR poly1.nist.gov. >--------------------------------------------------- I believe your problem is in having "too much" info in the nist.rev file. The PTR record in this file already has "1.154.6.129" from the "domain" "1.154.6.129.IN-ADDR.ARPA" which nist is the database for. Try using... >----------/usr/etc/named.d/nist.rev---------------- >@ IN SOA poly1.nist.gov. root.poly1.nist.gov. ( > 1.1 ; Serial > 10800 ; Refresh 3 hours > 3600 ; Retry 1 hour > 3600000 ; Expire 1000 hours > 86400 ) ; Minimum 24 hours > IN NS poly1.nist.gov. >poly1.nist.gov. IN A 129.6.154.1 > IN PTR poly1.nist.gov. >--------------------------------------------------- good luck,   Received: from vmb.brl.mil by VMB.BRL.MIL id ac02669; 17 May 90 18:26 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ah02470; 17 May 90 18:15 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa02441; 17 May 90 18:05 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa00788; 17 May 90 18:01 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA11079; Thu, 17 May 90 14:51:33 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 17 May 90 17:40:44 GMT From: Mike Walker Organization: NASA Langley Research Center, Hampton VA, 23665 Subject: fontmanager aliases and the font path Message-Id: Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL I recently have had to convert and install a large number of fonts for a Sun based X application on a personal iris and this has brought up a couple of questions which I can't find in the manuals. If anyone can point me at existing documentation or provide some information I would appreciate it. The gray area is exactly what should go in the fmaliases file in /usr/lib/fmfonts. After converting a family of fonts from sun vfont format to fontmanager format with dumpfont and running bldfamily I get an fmaliases file which contains (for the times roman fonts): "*" pstimr 10 "*" pstimr 11 "*" pstimr 12 "*" pstimr 13 "*" pstimr 14 "*" pstimr 16 "*" pstimr 18 "*" pstimr 24 "*" pstimr 9 When this is appended to the ``real'' fmaliases file xlsfonts doesn't see any new fonts (after an xset fp rehash). If I further add a font alias such as: "pstimr-9" pstimr 9 an xlsfonts now shows the fonts pstimr-9 and pstimr.9 (where the latter is the name of the original sunview font). So I guess my question is can I define all the various fonts in a font family without adding a specific alias for each in the fmaliases file? Next I would like to install these new fonts in a separate directory and modify the default font path on a system wide basis. Can this be done? Thanks for any help, Mike -- Mike Walker AS&M Inc/NASA LaRC Voice: +1 804 864 2305 Fax: +1 804 864 6134   Received: from vmb.brl.mil by VMB.BRL.MIL id aa02788; 17 May 90 18:37 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab02470; 17 May 90 18:15 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa02262; 17 May 90 17:53 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa00727; 17 May 90 17:44 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA09924; Thu, 17 May 90 14:33:22 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 17 May 90 17:01:57 GMT From: Archer Sully Subject: Re: Help with "Ethernet init failed" from dog Message-Id: <7821@odin.corp.sgi.com> References: <1990May16.231105.17956@unmvax.cs.unm.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <1990May16.231105.17956@unmvax.cs.unm.edu> lee@carina.unm.edu (Lee Ward CS.DEPT) writes: >The turbo machine is running operating system release number GL2-W3.5r1 >and dog itself is version 2.3 >So the question is: >Can somebody tell me what "Ethernet init failed" means or what else >I might do to get past it's little problem? It all depends on what version of dog you're running. If it is the one that came with the machine you need to be running the XNS kernel. If you are running XNS, or have the UDP version of dog, then I haven't a clue. -- Archer Sully | I'm 27 years old. That's 54 in Nerd Years (archer@esd.sgi.com) | -- Keith Rienzi   Received: from vmb.brl.mil by VMB.BRL.MIL id aa03160; 17 May 90 19:18 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa02966; 17 May 90 19:07 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa02942; 17 May 90 18:54 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa00951; 17 May 90 18:42 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA11810; Thu, 17 May 90 15:03:12 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 17 May 90 17:23:13 GMT From: Mike Walker Organization: NASA Langley Research Center, Hampton VA, 23665 Subject: fontmanager aliases and the font path Message-Id: Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL I recently have had to convert and install a large number of fonts for a Sun based X application on a personal iris and this has brought up a couple of questions which I can't find in the manuals. If anyone can point me at existing documentation or provide some information I would appreciate it. The gray area is exactly what should go in the fmaliases file in /usr/lib/fmfonts. After converting a family of fonts from sun vfont format to fontmanager format with dumpfont and running bldfamily I get an fmaliases file which contains (for the times roman fonts): "*" pstimr 10 "*" pstimr 11 "*" pstimr 12 "*" pstimr 13 "*" pstimr 14 "*" pstimr 16 "*" pstimr 18 "*" pstimr 24 "*" pstimr 9 When this is appended to the ``real'' fmaliases file xlsfonts doesn't see any new fonts (after an xset fp rehash). If I further add a font alias such as: "pstimr-9" pstimr 9 an xlsfonts now shows the fonts pstimr-9 and pstimr.9 (where the latter is the name of the original sunview font). So I guess my question is can I define all the various fonts in a font family without adding a specific alias for each in the fmaliases file? Next I would like to install these new fonts in a separate directory and modify the default font path on a system wide basis. Can this be done? I tried using fmsetpath() but that doesn't change the path on a sysmte wide basis. Thanks for any help, Mike -- Mike Walker AS&M Inc/NASA LaRC Voice: +1 804 864 2305 Fax: +1 804 864 6134   Received: from vmb.brl.mil by VMB.BRL.MIL id aa03915; 17 May 90 21:02 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa03676; 17 May 90 20:30 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa03651; 17 May 90 20:18 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa01343; 17 May 90 20:14 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA18838; Thu, 17 May 90 17:00:56 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 17 May 90 23:59:47 GMT From: yan xiao Organization: Indust. Eng., U. of Toronto. Subject: Redirect error message to a file. Message-Id: <1990May17.235947.10153@gpu.utcs.utoronto.ca> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Can I re-direct system error message and warnings to a file (e.g. /usr/adm/SYSLOG)? As when I use the console, some of the messages (like NFS timeout) are displayed on the screen and interrupt my display. And it is also important to keep a record of error messages. (We are using a 4D and a 3000 serial). Thanks. xiao (xiao@vered.rose.utoronto.ca) . Keywords:   Received: from vmb.brl.mil by VMB.BRL.MIL id ab04162; 17 May 90 21:32 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ac03915; 17 May 90 21:11 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab03856; 17 May 90 20:57 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa01414; 17 May 90 20:30 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA19823; Thu, 17 May 90 17:19:48 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 18 May 90 00:19:13 GMT From: Robert Lansdale Organization: CSRI, University of Toronto Subject: Problems with tgetent() Message-Id: <1990May17.201913.5931@jarvis.csri.toronto.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL I'm having a small problem with the tgetent() termlib emulation routine in the curses package (I belive that's where it's from). When I start up my 3d renderer with the line: nu_iris -r script_file < /dev/null > errors & I get the message "[1] + Stopped (tty output)". I traced the tty output to the tgetent() library call. Why does this routine need to do any terminal I/O? The problem does not occur on the SUN3 version, but this is probably because it uses termlib instead of Curses. Is there something I can do other than check to see whether the stdout has been redirected to a file and not call tgetent() in such a case? -->Rob Lansdale -------------------------------------------------------------------------- Robert Lansdale - (416) 978-6619 Dynamic Graphics Project Internet: lansd@dgp.toronto.edu Computer Systems Research Institute UUCP: ..!uunet!dgp.toronto.edu!lansd University of Toronto Bitnet: lansd@dgp.utoronto Toronto, Ontario M5S 1A4, CANADA   Received: from vmb.brl.mil by VMB.BRL.MIL id aa04597; 17 May 90 22:13 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa04346; 17 May 90 22:02 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa04309; 17 May 90 21:48 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa01585; 17 May 90 21:32 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA23379; Thu, 17 May 90 18:20:30 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 18 May 90 00:03:33 GMT From: "Lee Ward CS.DEPT" Organization: University of New Mexico, Albuquerque Subject: Re: Ethernet init failed Message-Id: <1990May18.000333.14885@unmvax.cs.unm.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Thanks for all the help. Both posts and email. The problem is that dog is for XNS and we are running IP. Is there an IP version I could ftp from somewhere? --Lee   Received: from vmb.brl.mil by VMB.BRL.MIL id aa05158; 17 May 90 23:44 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa04967; 17 May 90 23:23 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa04961; 17 May 90 23:14 EDT Received: from enh.nist.gov by VGR.BRL.MIL id aa01781; 17 May 90 23:08 EDT Return-path: rbriber@poly1.nist.gov Received: from poly1.nist.gov by ENH.NIST.GOV; Thu, 17 May 90 23:06 EST Received: by poly1.nist.gov (5.52/890607.SGI) (for @enh.nist.gov:info-iris@brl.army.mil) id AA00594; Thu, 17 May 90 23:12:11 EDT Date: Thu, 17 May 90 23:12:11 EDT From: rbriber@poly1.nist.gov Subject: named and NeWS To: info-iris@BRL.MIL Message-Id: <9005180312.AA00594@poly1.nist.gov> I posted a question about a problem with named and NeWs recently and I've had a number of suggestions for changing the nist.rev file I appended on the end of my message. Unfortunately none of the suggestions have helped. I had a couple of other suggestions about the rest of the named configuration files which I tried out but so far no luck. I still get the error: "getsocketpeername: cant find name for 129.6.154.1 ..." in SYSLOG from init.ps. I've also made a few other observations. For example, after I login I can get the chests to come up by starting them from the command line in the CONSOLE WINDOW ONLY. If I try to start them from a wsh window I get nothing or "Connection reset by peer". I should also note any other PostScript processes (such as/usr/demos/bin/eye.ps or anything in /usr/NeWS/demo) also fail and can only be started in the console window. This is currently not a satisfactory solution. I use nslookup I can find the address for any hosts I query about but the default server announced when the program starts is: Default Server: localhost.nist.gov Address: 127.0.0.1 and not my machine's name (poly1.nist.gov, 129.6.154.1). If I then ask for the address associatied with my name (poly1) I get the non-autoritative answer: 129.16.154.1, which is OK. BUT if I do 'nslookup -q=ptr 129.16.154.1' to ask for the name associated with my address I get the error: ** No domain name pointer information is available for 1.154.6.129.in-addr.arpa Which must be related to the fact that getsocketpeername is failing in init.ps. I've had a number of suggestions which involved setting NetSecurityWanted in init.ps to false. Is this dangerous? Oh, I also forgot to mention in my posting that I don't currently get the interest group mailings (thought I would have been subscribed by now but these things take time) so pleas e-mail any further ideas. I suspect if I can resolve the nslookup error init.ps will run. -- ---------------------------------------------------------------------------- | Adios Amoebas, | | | | "I've tried and I've tried and I'm still mystified, | | Robert Briber | I can't do it anymore and I'm not satisfied." | | 224/B210 | --Elvis | | NIST |------------------------------------------------------| | Gaithersburg, MD | rbriber@enh.nist.gov (Internet) | | 20899 USA | rbriber@poly1.nist.gov (Internet) | |(301) 975-6775(voice)| rbriber@nbsenh (Bitnet) | |(301) 975-2128 (fax) | | ----------------------------------------------------------------------------   Received: from vmb.brl.mil by VMB.BRL.MIL id ab17671; 18 May 90 16:17 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab17349; 18 May 90 16:07 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ac17189; 18 May 90 15:44 EDT Received: from [129.112.1.12] by VGR.BRL.MIL id aa05778; 18 May 90 15:31 EDT Received: from baby.swmed.utexas.edu by utsw.swmed.utexas.edu with SMTP; Thu, 17 May 1990 11:44:31 CDT Received: by baby.swmed.utexas.edu (5.52/890619.SGI) (for @utsw.swmed.utexas.edu:info-iris@brl.mil) id AA14106; Thu, 17 May 90 11:47:54 CDT Date: Thu, 17 May 90 11:47:54 CDT From: Rose Oguz Message-Id: <9005171647.AA14106@baby.swmed.utexas.edu> To: info-iris@BRL.MIL Subject: Rotations about an arbitrary axis Hi all! I'm trying to do rotations about an arbitrary axis and I'm running into some difficulties. This algorithm is described in Newman and Sproull's PRINCIPLES OF INTERACTIVE COMPUTER GRAPHICS and Foley and van Dam's FUNDAMENTALS OF INTERACTIVE COMPUTER GRAPHICS. I have followed the algorithm, but I still have a problem. My problem seems to stem from the calculation of the axis of rotation. Basically, I have two vectors coming from the same point and one of the vectors needs to be rotated into the other. So, I calculate the axis of rotation as the cross product of the two vectors. The resulting vector can be in one of two directions from the point and depends on the "order" that the vectors are crossed. I spoke with someone who did this a few years ago and he added checks to the algorithm. Based on the signs of certain values, other values may be negated. However, it didn't work every time. So, by interactively doing this, he was able to determine if the rotation was correct and if it wasn't, some value was negated and the algorithm was executed again. In any case, I tried to do it similarly, but I couldn't get it to work properly. Also, I don't want to do this interactively. So, if anyone has any ideas on this, please let me know. Thanks a million! Rose Oguz rose@baby.swmed.utexas.edu   Received: from vmb.brl.mil by VMB.BRL.MIL id aa06847; 18 May 90 4:33 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa06485; 18 May 90 3:20 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa06471; 18 May 90 3:10 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa02466; 18 May 90 3:00 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA11824; Thu, 17 May 90 23:45:21 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 18 May 90 04:53:17 GMT From: Rob Warnock Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Re: Problems with tgetent() Message-Id: <60553@sgi.sgi.com> References: <1990May17.201913.5931@jarvis.csri.toronto.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <1990May17.201913.5931@jarvis.csri.toronto.edu> lansd@dgp.toronto.edu (Robert Lansdale) writes: +--------------- | I'm having a small problem with the tgetent() termlib emulation | routine in the curses package (I belive that's where it's from). When | I start up my 3d renderer with the line: | nu_iris -r script_file < /dev/null > errors & | I get the message "[1] + Stopped (tty output)". I traced the tty output | to the tgetent() library call. Why does this routine need to do any | terminal I/O? The problem does not occur on the SUN3 version, but this | is probably because it uses termlib instead of Curses. Is there something | I can do other than check to see whether the stdout has been redirected | to a file and not call tgetent() in such a case? +--------------- Ah... but not *all* your terminal output has been redirected to a file. Your "stderr" output is still going to the window you ran this in. Try either, for "sh": nu_iris -r script_file < /dev/null > errors 2>&1 & or, for "csh": nu_iris -r script_file < /dev/null >& errors & and see if that doesn't help. -Rob ----- Rob Warnock, MS-9U/510 rpw3@sgi.com rpw3@pei.com Silicon Graphics, Inc. (415)335-1673 Protocol Engines, Inc. 2011 N. Shoreline Blvd. Mountain View, CA 94039-7311   Received: from vmb.brl.mil by VMB.BRL.MIL id aa09862; 18 May 90 8:40 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa09439; 18 May 90 8:30 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa09164; 18 May 90 8:20 EDT Received: from aero4.larc.nasa.gov by VGR.BRL.MIL id aa03280; 18 May 90 8:12 EDT Received: Fri, 18 May 90 08:12:41 EDT by aero4.larc.nasa.gov (5.52/5.6) Date: Fri, 18 May 90 08:12:41 EDT From: "Brent L. Bates AAD/TAB MS361 x42854" Message-Id: <9005181212.AA16274@aero4.larc.nasa.gov> To: finn.drablos@sintef.no Subject: Re: Help on printers, disks, memory Cc: info-iris@BRL.MIL I put the source to the program I wrote to convert PLOT3D/GAS/SURF ".ras" files (run length encoded) to Tektronix 4693D, black & white and grayscale PostScript formats in the info-iris pub directory. It also has an option to convert the files to Celco film recorder format, but it isn't all there. I am in the process of converting my original FORTRAN source to C and haven't finished yet. -- Brent L. Bates NASA-Langley Research Center M.S. 361 Hampton, Virginia 23665-5225 (804) 864-2854 E-mail: blbates@aero4.larc.nasa.gov or blbates@aero2.larc.nasa.gov   Received: from vmb.brl.mil by VMB.BRL.MIL id ab09862; 18 May 90 8:41 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab09439; 18 May 90 8:30 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab09164; 18 May 90 8:20 EDT Received: from aero4.larc.nasa.gov by VGR.BRL.MIL id aa03286; 18 May 90 8:13 EDT Received: Fri, 18 May 90 08:13:35 EDT by aero4.larc.nasa.gov (5.52/5.6) Date: Fri, 18 May 90 08:13:35 EDT From: "Brent L. Bates AAD/TAB MS361 x42854" Message-Id: <9005181213.AA16281@aero4.larc.nasa.gov> To: finn.drablos@sintef.no Subject: Re: Help on printers, disks, memory Cc: info-iris@BRL.MIL P.S. The file name is "change" -- Brent L. Bates NASA-Langley Research Center M.S. 361 Hampton, Virginia 23665-5225 (804) 864-2854 E-mail: blbates@aero4.larc.nasa.gov or blbates@aero2.larc.nasa.gov   Received: from vmb.brl.mil by VMB.BRL.MIL id aa10780; 18 May 90 9:34 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa10465; 18 May 90 9:23 EDT Date: Fri, 18 May 90 9:07:23 EDT From: Gary S. Moss (VLD/VMB) To: Rob Warnock cc: info-iris@BRL.MIL Subject: Re: Problems with tgetent() Message-ID: <9005180907.aa10358@VMB.BRL.MIL> errors & <| I get the message "[1] + Stopped (tty output)". I traced the tty output <| to the tgetent() library call. Why does this routine need to do any <| terminal I/O? The problem does not occur on the SUN3 version, but this <| is probably because it uses termlib instead of Curses. Is there something <| I can do other than check to see whether the stdout has been redirected <| to a file and not call tgetent() in such a case? <+--------------- [ stuff about redirecting stderr as well as stdout ] Well, chances are the root of the problem is calling tgetent. You really don't want to call tgetent if stdout is not connected to a terminal. Use isatty(3) to determine if a file descriptor is attached to a terminal before attempting screen-oriented output. I usually check stdin, since if a user is providing input from a file or pipe, he probably doesn't need termlib style output: if( (tty = isatty( 0 )) ) InitTermCap(); /* initially */ if( tty ) DoScreenStuff(); /* later */   Received: from vmb.brl.mil by VMB.BRL.MIL id ab10780; 18 May 90 9:34 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab10465; 18 May 90 9:23 EDT Date: Fri, 18 May 90 9:10:46 EDT From: Gary S. Moss (VLD/VMB) To: Al Broderick cc: info-iris@BRL.MIL Subject: Re: SGI include file problem? Message-ID: <9005180910.aa10444@VMB.BRL.MIL> < cpp: error /usr/include/bsd/sys/ioctl.h:9: Can't find include file net/soioctl.h < cpp: error /usr/include/bsd/sys/ioctl.h:10: Can't find include file sys/ttychars.h < The offending code in /usr/include/bsd/sys/ioctl.h says this: < #include "../../sys/ioctl.h" < #include < #include < Should it really look like this? < #include "../../sys/ioctl.h" < #include < #include It looks like you have explicitly #included . Rather than do this, use the -I/usr/include/bsd option on the compile line.   Received: from vmb.brl.mil by VMB.BRL.MIL id aa15058; 18 May 90 12:56 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa14700; 18 May 90 12:45 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa14599; 18 May 90 12:25 EDT Received: from wugate.wustl.edu by VGR.BRL.MIL id aa04518; 18 May 90 12:09 EDT Received: by wugate.wustl.edu (5.61++/WUSTL-0.3) with SMTP id AA28277; Fri, 18 May 90 11:08:57 -0500 Received: by castor.wustl.edu (5.52/890607.SGI) (for @wugate.wustl.edu:info-iris@brl.mil) id AA00871; Fri, 18 May 90 11:08:21 CDT Date: Fri, 18 May 90 11:08:21 CDT From: "Martin S. Weinhous" Message-Id: <9005181608.AA00871@castor.wustl.edu> To: info-iris@BRL.MIL Subject: DWB, LaserWriter support and the screen Cc: weinhous@castor.wustl.edu We just purchased DWB and LaserWriter support software from SGI. How does one preview a postscript document on the screen? Thanks in advance. -- Marty Weinhous weinhous@castor.wustl.edu Radiation Oncology, Washington U. !uunet!wucs1!dinorah!weinhous St. Louis, MO 63105, 314/362-2600   Received: from vmb.brl.mil by VMB.BRL.MIL id aa17671; 18 May 90 16:17 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa17349; 18 May 90 16:07 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab17189; 18 May 90 15:43 EDT Received: from enh.nist.gov by VGR.BRL.MIL id aa05747; 18 May 90 15:30 EDT Return-path: root@poly1.nist.gov Received: from poly1.nist.gov by ENH.NIST.GOV; Fri, 18 May 90 15:28 EST Received: by poly1.nist.gov (5.52/890607.SGI) (for @enh.nist.gov:Info-Iris@brl.army.mil) id AA00610; Fri, 18 May 90 15:33:05 EDT Date: Fri, 18 May 90 15:33:05 EDT From: root@poly1.nist.gov Subject: named&NeWS now working To: Info-Iris@BRL.MIL Message-Id: <9005181933.AA00610@poly1.nist.gov> I have found a *Non-authoratative* answer to getting 'named' and NeWS to work together on our machine without having to turn NetSecurityWanted off. Currently we are able to use our own machine to resolve our own name and localhost and then go to one of our nameservers for anything else. The windows all open and the chests all appear on login. A generic set of the files we are using is given below for the arbitrary machine *iris.mydomain.gov* (address 500.50.5.1). The computer *knowitall.mydomain.gov* (address 500.50.6.1)is used by sendmail for mail forwarding and by 'named' as the first nameserver when iris.mydomain.gov doesn't know the answer. -------------/etc/hosts--------------------------------- 127.1 localhost | 500.50.5.1 iris | <= Our machine. 500.50.6.1 knowitall | <= The machine sendmail uses for forwarding | and also will be our 1st nameserver. 500.50.5.254 gtwy224 | <= Our local subnet gateway. -------------------------------------------------------- -----------/usr/etc/resolv.conf------------------------- domain mydomain.gov nameserver 127.0.0.1 -------------------------------------------------------- -----------/usr/etc/named.d/named.boot------------------ directory /usr/etc/named.d ; type domain source host/file cache . root.cache primary 0.0.127.IN-ADDR.ARPA localhost.rev primary 1.6.50.500.IN-ADDR.ARPA domain.rev -------------------------------------------------------- -----------/usr/etc/named.d./root.cache----------------- ; Initial cache data for root domain servers. ; . 99999999 IN NS knowitall.mydomain.gov. ; ; Prep the cache (hotwire the addresses). Order does not matter ; knowitall.mydomain.gov. 99999999 IN A 500.50.6.1 -------------------------------------------------------- -----------/usr/etc/named.d/localhost.rev--------------- @ IN SOA iris.mydomain.gov. root.iris.mydomain.gov. ( 1.1 ; Serial 3600 ; Refresh 300 ; Retry 3600000 ; Expire 14400 ) ; Minimum 1 IN PTR localhost. IN NS iris.mydomain.gov. ------------------------------------------------------- -----------/usr/etc/named.d/mydomain.rev--------------- @ IN SOA iris.mydomain.gov. root.iris.mydomain.gov. ( 1.1 ; Serial 10800 ; Refresh 3 hours 3600 ; Retry 1 hour 3600000 ; Expire 1000 hours 86400 ) ; Minimum 24 hours IN PTR iris.mydomain.gov. IN NS iris.mydomain.gov. ------------------------------------------------------ As strange as it may seem (at least it seems strange to me) the order of the last 2 lines in mydomain.rev is *very* important. If the IN NS line is before the IN PTR line neither NeWS (through getsocketpeername) or nslookup can find a name for the localhost's own numeric address. If I have them as shown above then everything resolves OK. Based on that I also switched the last two lines in localhost.rev to put the IN PTR line first, though it seems to work either way. Thanks to everyone the net and the people @sgi.com (joyce, arc, strang, jesse) who suggested things. This resulted from a synthesis of everything and the divine inspiration that comes from lack of sleep. -- ---------------------------------------------------------------------------- | Adios Amoebas, | "I've tried and I've tried and I'm still mystified, | | Robert Briber | I can't do it anymore and I'm not satisfied." | | 224/B210 | --Elvis | | NIST |------------------------------------------------------| | Gaithersburg, MD | rbriber@poly1.nist.gov (Internet) | | 20899 USA | rbriber@enh.nist.gov (Internet) | |(301) 975-6775(voice)| rbriber@nbsenh (Bitnet) | |(301) 975-2128 (fax) | | ----------------------------------------------------------------------------   Received: from vmb.brl.mil by VMB.BRL.MIL id aa18295; 18 May 90 17:09 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa18126; 18 May 90 16:58 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa18052; 18 May 90 16:40 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa06093; 18 May 90 16:31 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA23576; Fri, 18 May 90 12:46:20 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 18 May 90 19:42:26 GMT From: Fred Velijanian Organization: University of California, Irvine Subject: IRIX support of BSD printers Message-Id: <26545022.559@orion.oac.uci.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL We are a new SGI site, and apologies in advance if this topic has been discussed before in this newsgroup. Our two 4D20G's are printer-less and I'd like to find a way to use our existing printers which are all on BSD machines. Is there a way to accomplish this or do we have to wait until SGI includes the BSD lpr/lpd services in the next release of IRIX? Thank you very much for any pointers you may provide. Fred Velijanian Univ of Calif, Irvine fredv@vmsd.oac.uci.edu   Received: from vmb.brl.mil by VMB.BRL.MIL id ab18483; 18 May 90 17:55 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa18437; 18 May 90 17:38 EDT Received: from adm.brl.mil by VMB.BRL.MIL id aa18431; 18 May 90 17:27 EDT Received: from ucbvax.Berkeley.EDU by ADM.BRL.MIL id aa23145; 18 May 90 16:48 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA25705; Fri, 18 May 90 13:21:50 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 18 May 90 16:40:13 GMT From: Archer Sully Subject: Re: SGI include file problem? Message-Id: <7875@odin.corp.sgi.com> References: Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article broderic@topaz.rutgers.edu (Al Broderick) writes: >I have a SGI 4D/70G running IRIX 3.2.1. I have a few questions. >I get this error message when I compile (a program that compiled >cleanly on 3.1): >cpp: error /usr/include/bsd/sys/ioctl.h:9: Can't find include file net/soioctl.h >cpp: error /usr/include/bsd/sys/ioctl.h:10: Can't find include file sys/ttychars.h >The offending code in /usr/include/bsd/sys/ioctl.h says this: >#include "../../sys/ioctl.h" >#include >#include >Should it really look like this? >#include "../../sys/ioctl.h" >#include >#include >I have not tried this on my machine yet because I am hoping that >I am doing something wrong. >Could someone from SGI please advise? Yes. You are probably doing something wrong :-) You need to add the following incantation in your CFLAGS (however you do it with your makefiles): -I/usr/include/bsd That should clear everything up. -- Archer Sully | I'm 27 years old. That's 54 in Nerd Years (archer@esd.sgi.com) | -- Keith Rienzi   Received: from vmb.brl.mil by VMB.BRL.MIL id ab18672; 18 May 90 18:15 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa18483; 18 May 90 17:55 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa18447; 18 May 90 17:37 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa06195; 18 May 90 16:56 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA25864; Fri, 18 May 90 13:24:42 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 18 May 90 17:18:14 GMT From: dave "who can do? ratmandu!" ratcliffe Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Re: Imagetools Manual Wanted Message-Id: <7881@odin.corp.sgi.com> References: <6180@amelia.nas.nasa.gov> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <6180@amelia.nas.nasa.gov> woo@pioneer.arc.nasa.gov (Alex Woo RAC) writes: > >A number of SGI's imagetools have been recently posted to >the network. Is there a nice manual somewhere that explains >all of them? There seems to be a few programs in >/usr/sbin with a few scattered man pages. > >A pointer to a collection of man pages would be nice. unfortunately alex, that which you seek is still in a state of "evolvement"... however, i can help fill in the gaps to a degree: as an admittedly poor substitute for: -- a complete man page cross-referencing set, and -- a detailed and comprehensive "IMGTOOLS USER'S GUIDE", you want to study that which exists under /usr/people/4Dgifts/iristools/imgtools there's a fairly long history to this subdirectory--and the entire subtree of ~4Dgifts/iristools. at this point, let me give a breakdown of what is currently bundled together. **CURRENTLY: there are 3 subsystems you want to check out/be familiar with: dev.sw.giftssrc eoe2.sw.gltools eoe2.sw.moregltools on the DEV tape there is: ----------------------------- "dev.sw.giftssrc": this is the subsystem containing the entire /usr/people/4Dgifts subtree. this is a sample account (like guest) with lots of hooks in it's own user.ps, and the files that it loads, for customizing yer own NeWS environment when you login on the graphics console. it also contains various source code examples (consult the ~4Dgifts/README file for a breakdown of this subtree's contents), including the whole of the ~4Dgifts/iristools subtree. the majority of image tools yer interested in live under ~4Dgifts/iristools/imgtools (there are a couple you might also find useful (like colorbars, gamcal, mag) in ~4Dgifts/iristools/tools. all of this stuff is the source code used that builds the executables you are finding in /usr/sbin. ---> for now the documentation consists of the incomplete man pages you are finding, PLUS the ~4Dgifts/iristools/imgtools/README, PLUS the source code files themselves (yes, this is not optimal, but it *is* what/all that is currently available). on the EOE2 tape there is: ------------------------------ "eoe2.sw.gltools" is the subsystem which contains the following executables: blanktime gamcal *imgexp *loadmap *scrsave cedit gamma interp mag showmap clock ical *ipaste mousewarp *snapshot dialwarp *icut *istat *savemap textcolors "eoe2.sw.moregltools" is the subsystem containing the following executables: *abs *dotgen *hist *movie *rle *tobin *add *fieldmerge *histeq *mult *saturate *tobw *addframe *fromalias *iflip *noblack scanner *tops *addnoise *frombin imged *nullimg *scope *toscitex *assemble *fromcmap *imgwrap *oneband *scrsave *totarga *blend *fromdi *invert *over *setlum *verbatim *blur *fromrla *iset palette *showimg vis *cglue *fromtarga *izoom *postcard *subimg *writeimg colorbars *gammawarp *mapimg *quant *thresh *convolve *greyscale *max *randimg *tile cycol grid mousemon *readimg *toalias each of the above executable programs, when loaded off the EOE2 tape, live in /usr/sbin. "eoe2.sw.gltools" gets loaded by default on every system. both "eoe2.sw.moregltools", as well as "dev.sw.giftssrc", DO NOT GET LOADED BY DEFAULT and, on a "vanilla" 4D system, these 2 subsystems must initially be loaded manually via "Manual" mode inside inst. once these subsystems are initially loaded explicitly from inst inside Manual mode, the install tool "knows" about them from that time on, and they do not need to be explicitly tagged again via Manual mode for being updated when a new release of the system software becomes available or with any maintenance release software. all of the executables listed above which are preceeded by an asterisk have equivalent source files that live in ~4Dgifts/iristools/imgtools. all you hardocre comp.sys.sgi user's will note that the following are recent additions, courtesy of Paul "iristools OVERLORD" Haeberli, which are not currently a part of "dev.sw.giftssrc" or "eoe2.sw.moregltools": frommac.c (convert macpaint image files to IRIS image files) tomac.c (convert an IRIS image file to macpaint format) fromgif.c (convert a GIF file into an IRIS image file) fromyuv (uuencoded prog. which converts Abekas A60 digital video format to IRIS image file format) toyuv (uuencoded prog. which converts an IRIS image file to Abekas A60 digital video format) i am currently engaged in writing man pages for all those executables in eoe2.sw.moregltools that do not, at present exist. these are being initially generated from the ~4Dgifts/iristools/imgtools/README file with additional information/cross-referencing being added as i go. i do not yet know when these will be available. stay tuned. -- daveus rattus yer friendly neighborhood ratman KOYAANISQATSI ko.yan.nis.qatsi (from the Hopi Language) n. 1. crazy life. 2. life in turmoil. 3. life out of balance. 4. life disintegrating. 5. a state of life that calls for another way of living.   Received: from vmb.brl.mil by VMB.BRL.MIL id aa18827; 18 May 90 18:26 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa18672; 18 May 90 18:15 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa18503; 18 May 90 17:53 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa06213; 18 May 90 17:03 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA25875; Fri, 18 May 90 13:24:58 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 18 May 90 18:10:45 GMT From: Vernon Schryver Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Re: SGI include file problem? Message-Id: <60579@sgi.sgi.com> References: Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article , broderic@topaz.rutgers.edu (Al Broderick) writes: >... > Should it really look like this? > #include "../../sys/ioctl.h" > #include > #include No, the contents of /usr/include are correct as shipped (at least in this regard). There is (or should be) a note on every man page for BSD stuff like: ]NOTE ] To compile and link a program that calls this routine, follow the ] procedures for section (3B) routines as described in intro(3). That text says to use -I/usr/include/bsd and -lbsd. This is necessary to get BSD compatibility stuff. The next release substantially relaxes this requirement, because we have moved lots of BSD stuff to libc, and used other mechanisms to select among conflicting standards (e.g. readdir and signals). Vernon Schryver vjs@sgi.com   Received: from vmb.brl.mil by VMB.BRL.MIL id ab18827; 18 May 90 18:26 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ac18672; 18 May 90 18:16 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa18536; 18 May 90 17:55 EDT Received: from aero4.larc.nasa.gov by VGR.BRL.MIL id aa06280; 18 May 90 17:25 EDT Received: Fri, 18 May 90 17:26:44 EDT by aero4.larc.nasa.gov (5.52/5.6) Date: Fri, 18 May 90 17:26:44 EDT From: "Brent L. Bates AAD/TAB MS361 x42854" Message-Id: <9005182126.AA18797@aero4.larc.nasa.gov> To: weinhous@castor.wustl.edu Subject: Re: DWB, LaserWriter support and the screen Cc: info-iris@BRL.MIL You don't. (Preview PostScript files on the screen) -- Brent L. Bates NASA-Langley Research Center M.S. 361 Hampton, Virginia 23665-5225 (804) 864-2854 E-mail: blbates@aero4.larc.nasa.gov or blbates@aero2.larc.nasa.gov   Received: from vmb.brl.mil by VMB.BRL.MIL id aa19669; 18 May 90 22:06 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa19590; 18 May 90 21:56 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa19571; 18 May 90 21:39 EDT Received: from [132.206.4.10] by VGR.BRL.MIL id aa07002; 18 May 90 21:21 EDT Received: by frodo.Physics.McGill.CA id AA20028; Fri, 18 May 90 21:20:22 EDT (5.59++/IDA-1.1S) Date: Fri, 18 May 90 21:20:22 EDT From: Loki Jorgenson Rm421 Message-Id: <9005190120.AA20028@frodo.Physics.McGill.CA> To: info-iris@vgr.brl.mil Subject: TEK4010/4014 Is there any such beast as a TEK4010/4014 terminal emulator for an IRIS Personal console? This may sound a bit perverse but it would make life here much easier. Thanks, Loki Jorgenson node: loki@physics.mcgill.ca Physics, McGill University fax: (514) 398-3733 Montreal Quebec CANADA phone: (514) 398-6531 <.... Free the mind and the police will surely follow ....>   Received: from vmb.brl.mil by VMB.BRL.MIL id aa24161; 19 May 90 16:03 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa23916; 19 May 90 15:11 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ad23880; 19 May 90 15:01 EDT Received: from cunyvm.cuny.edu by VGR.BRL.MIL id aa09859; 19 May 90 14:52 EDT Received: from MCCLB0.MED.NYU.EDU by CUNYVM.CUNY.EDU (IBM VM SMTP R1.2.2MX) with BSMTP id 6389; Sat, 19 May 90 14:33:26 EDT Received: from mcirps2.med.nyu.edu by MCCLB0.MED.NYU.EDU; Sat, 19 May 90 14:33 EDT Received: by mcirps2.med.nyu.edu (5.52/890607.SGI) (for @mcclb0.med.nyu.edu:info-iris@brl.arpa) id AA17407; Sat, 19 May 90 14:37:25 DSD Date: Sat, 19 May 90 14:37:25 DSD From: karron%mcirps2.med.nyu.edu@cunyvm.cuny.edu Subject: Horizontal Lines only in Pandora To: info-iris@BRL.MIL Message-id: <9005192137.AA17407@mcirps2.med.nyu.edu> X-Envelope-to: info-iris@brl.arpa Here is an odd problem: We get dirty horizontal square 'dots' that wander all over the screen. It looks like bad bitplanes, but the problem only manifests its self in the pandora login program. The prom monitor and wsh are fine. What does pandora do that is not done elseware in the machine that could have somthing to do with this problem. We are ignoring it for now. Could this have somthing to do with the slow graphics performance we are seeing ( on our 70g running 3.2) dan. -- +-----------------------------------------------------------------------------+ | karron@nyu.edu Dan Karron | | . . . . . . . . . . . . . . New York University Medical Center | | 560 First Avenue \ \ Pager <1> (212) 397 9330 | | New York, New York 10016 \**\ <2> 10896 <3> | | (212) 340 5210 \**\__________________________________________ | +-----------------------------------------------------------------------------+   Received: from vmb.brl.mil by VMB.BRL.MIL id aa24881; 19 May 90 19:28 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa24843; 19 May 90 19:17 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa24830; 19 May 90 19:09 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa10537; 19 May 90 19:03 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA16909; Sat, 19 May 90 15:48:22 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 19 May 90 22:32:21 GMT From: "Daniel L. Severance" Organization: Purdue University Subject: Why does a 3.5" floppy cost $1000?? Message-Id: <4830@mace.cc.purdue.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Hi!, First, let me say that I have found SGI to have fast machines at a great price. Ok, now for the bad news: Could someone from SGI or elsewhere tell me why a 3 1/2" floppy drive for a personal IRIS lists for $1000, while the one in our SPARCstation I lists (from SUN) for $250 ??? Even $300-$400 would be ok, but $1000?? There is NO WAY that a $150 PC floppy with a SCSI adapter has to cost $1000! My sales rep. in New Haven (we're moving to Yale in June) said he talked to the product manager about the unreasonable price, and was told that "well, we don't sell that many of them, anyway". Frankly, I can't say I'm surprised. There are 3 research groups here at Purdue Chem, and 1 central facility which would all be interested in the SoftPC/AT emulator AND floppy drives. The software isn't much use if you don't have easy access to a floppy, and no one is willing to plunk down $1000 (- discount) for the things. The floppies are also quite nice for archiving small amounts of data. Don't get me wrong, the P.I.'s put the SUNs to shame in CPU performance, A 4D-25x is over twice the speed on all of our programs (Molecular modelling) and closer to 3x on some. We have no complaints there. SGI's prices on hard disks and memory are also quite reasonable. We buy SGI peripherals rather than 3rd party so the service contracts are simpler, but only since they are competitive (not cheaper, but close enough to be reasonable). In lieu of spending $1000, does anyone know how to make use of the floppy in a SPARCstation I over the network with the SoftPC emulator on the IRIS? Dan Severance Internet: dan@rani.chem.purdue.edu   Received: from vmb.brl.mil by VMB.BRL.MIL id aa25103; 19 May 90 20:25 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa25028; 19 May 90 20:14 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa25008; 19 May 90 20:02 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa10626; 19 May 90 19:48 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA19605; Sat, 19 May 90 16:40:01 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 19 May 90 23:39:18 GMT From: Edward Vielmetti Organization: University of Michigan Math Dept., Ann Arbor MI. Subject: one IRIS, the rest Suns. how easy? Message-Id: Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Hi. I may be faced this summer with the task of putting one Iris onto a network full of Suns. I'd like to figure out how easy this is going to be given that I won't have a lot of time to port systems software to the IRIS -- so for bsd'ish things like syslog, sendmail, lpr, dump, X windows et al I need to know either that they work well or it might as well be not at all. From a review of recent postings on comp.sys.sgi it looks like I need to worry about: - usability of 4MB simms on Personal IRIS - support for "dump" and "restore" - support for the GNU C compiler - conversion of IRIS bitmap files to pbm/ppm (portable bitmap/pixmap) formats - support for Berkeley printing (lpr and lpd) - availability, usability and goodness of X Windows server - support for Berkeley "who", "w", and "uptime" programs - support for process monitoring tools like "top" - availability of TeX previewer under some window system. Some of these appear to be addressed in IRIX 3.3, a target date for that release would be nice to know. The goal is to be able to drop this on the desk of someone who's currently using a Sun 3/60 with SunView and get them up to speed quickly; I'm willing to forego porting every last program in the world from the Suns to the IRIS since after all the Suns will still be around. Thanks for any info -- replies back to the net are fine. --Ed Edward Vielmetti, U of Michigan math dept. emv@math.lsa.umich.edu   Received: from vmb.brl.mil by VMB.BRL.MIL id aa25557; 19 May 90 22:56 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa25371; 19 May 90 21:53 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa25364; 19 May 90 21:47 EDT Received: from cunyvm.cuny.edu by VGR.BRL.MIL id aa10915; 19 May 90 21:40 EDT Received: from MCCLB0.MED.NYU.EDU by CUNYVM.CUNY.EDU (IBM VM SMTP R1.2.2MX) with BSMTP id 7966; Sat, 19 May 90 21:39:08 EDT Received: from mcirps2.med.nyu.edu by MCCLB0.MED.NYU.EDU; Sat, 19 May 90 15:06 EDT Received: by mcirps2.med.nyu.edu (5.52/890607.SGI) (for @mcclb0.med.nyu.edu:info-iris@brl.arpa) id AA18900; Sat, 19 May 90 15:10:20 DSD Date: Sat, 19 May 90 15:10:20 DSD From: karron%mcirps2.med.nyu.edu@cunyvm.cuny.edu Subject: bsd include files... To: broderic@topaz.rutgers.edu, info-iris@BRL.MIL Message-id: <9005192210.AA18900@mcirps2.med.nyu.edu> X-Envelope-to: info-iris@brl.arpa Al Broderick writes: >I have a SGI 4D/70G running IRIX 3.2.1. I have a few questions. >I get this error message when I compile (a program that compiled >cleanly on 3.1): >cpp: error /usr/include/bsd/sys/ioctl.h:9: Can't find include file net/soioctl. h > >cpp: error /usr/include/bsd/sys/ioctl.h:10: Can't find include file sys/ttychar s.h > >The offending code in /usr/include/bsd/sys/ioctl.h says this: >#include "../../sys/ioctl.h" >#include >#include > >Should it really look like this? >#include "../../sys/ioctl.h" >#include >#include > >I have not tried this on my machine yet because I am hoping that >I am doing something wrong. It is a bsd hack. I have had this problem, and I was told that to do with bsd defines that conflict with sysV defines. Try -I/usr/include/bsd to include the bsd directory in your search path. This will give you bsd flavor instead of sysV flavor defines. Since these includes are not on 3.2, it should not hurt. Also try -Ibsd. This should just append bsd/ to your base include search path which defaults to "./:/usr/include", i.e, "./:/usr/include/:/usr/include/bsd", but I am not certain. You may also have problems with the dir parts of the include names. Multiple -I should solve it. I will test it at some point. If there is not name conflict, why does sgi not just make symbolic links for these includes ? A side note concerning these types of include problems: /usr/lib/cpp has a -M option that will generate a makefile code fragment expanding all of the include file dependencies. I am playing with getting a proper makefile for my own stuff which has to run on number of different dialect machines. I have not tested this code below yet, but running this prior to porting code to a new version will flag problems in your #include statments or Makefile. MAKEFILE FRAGMENT BELOW depend : cp Makefile Makefile.bak sed -e "/^### File Dependencies/,$$$$d" < Makefile > Makefile.dep echo "### File Dependencies" >> Makefile.dep for file in $(CSOURCES) ; do \ /lib/cpp -M $(INCLUDES) $$file >> Makefile.dep ; \ done mv Makefile.dep Makefile ### File Dependencies circle.o: circle.c circle.o: /usr/include/device.h circle.o: /usr/include/gl.h circle.o: /usr/include/malloc.h circle.o: /usr/include/math.h circle.o: /usr/include/sgimath.h circle.o: /usr/include/values.h circle.o: /usr/include/stdio.h circle.o: ./common.h circle.o: ./draw.h circle.o: ./circle.h This was contributed by Scott Townsend at NASA Lewis Research Center. I think that it is very neat. dan. -- +-----------------------------------------------------------------------------+ | karron@nyu.edu Dan Karron | | . . . . . . . . . . . . . . New York University Medical Center | | 560 First Avenue \ \ Pager <1> (212) 397 9330 | | New York, New York 10016 \**\ <2> 10896 <3> | | (212) 340 5210 \**\__________________________________________ | +-----------------------------------------------------------------------------+   Received: from vmb.brl.mil by VMB.BRL.MIL id aa25689; 19 May 90 23:32 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa25631; 19 May 90 23:22 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa25612; 19 May 90 23:08 EDT Received: from umich.edu by VGR.BRL.MIL id aa11144; 19 May 90 23:01 EDT Received: from ummts.cc.umich.edu by umich.edu (5.61/1123-1.0) id AA01896; Sat, 19 May 90 23:00:41 -0400 Date: Sat, 19 May 90 22:59:34 EDT From: Tim_Buxton@um.cc.umich.edu To: info-iris@BRL.MIL Message-Id: <6226910@um.cc.umich.edu> Subject: We need help getting IRISPLOT, Please We would like to get IRISPLOT via anonymous ftp. I have contacted system@math.arizona.edu and gotten the address of the archive machine, connemara.math.arizona.edu. Repeated attempts to connect to that machine via InterNet have failed, apparently because the machine is down. Could someone else who has IRISPLOT make it available over the net? Thanks a lot. Tim Buxton OptiMetrics, Inc. Tim_Buxton@um.cc.umich.edu   Received: from vmb.brl.mil by VMB.BRL.MIL id aa25989; 20 May 90 0:12 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa25760; 20 May 90 0:01 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa25743; 19 May 90 23:50 EDT Received: from umich.edu by VGR.BRL.MIL id aa11220; 19 May 90 23:45 EDT Received: from ummts.cc.umich.edu by umich.edu (5.61/1123-1.0) id AA01840; Sat, 19 May 90 22:56:46 -0400 Date: Sat, 19 May 90 22:55:40 EDT From: Tim_Buxton@um.cc.umich.edu To: info-iris@BRL.MIL Message-Id: <6226902@um.cc.umich.edu> Subject: What to do after you run tomac? Thanks to Paul Haeberli for the tomac.c source. We would like to use it to move images into PageMaker on the Mac, but are unable to PLACE the graphic in a document. When we try, an icon with a curved arrow shaped like a sideways "s" appears, then disappears, and the graphic has not been placed. The number of bytes is the same on the Mac as it was on the SGI, so I believe the transfer was binary and correct. The file icon indicated a text type file. If someone has succeeded in this operation, we would appreciate your help. Thank you. -Tim Buxton OptiMetrics, Inc Tim_Buxton@um.cc.umich.edu   Received: from vmb.brl.mil by VMB.BRL.MIL id aa27762; 20 May 90 7:56 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa27713; 20 May 90 7:46 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa27691; 20 May 90 7:33 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa12389; 20 May 90 7:19 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA20395; Sun, 20 May 90 04:12:39 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 20 May 90 07:18:55 GMT From: John H Merritt Organization: Goddard Space Flight Center Climate and Radiation Branch Subject: Re: Why does a 3.5" floppy cost $1000?? Message-Id: <2100@dftsrv.gsfc.nasa.gov> References: <4830@mace.cc.purdue.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <4830@mace.cc.purdue.edu> irg@mace.cc.purdue.edu (Daniel L. Severance) writes: > SGI's prices on hard disks and memory are also quite reasonable. ^^^^^unreasonable. > We buy SGI peripherals rather than 3rd party so the service contracts > are simpler, but only since they are competitive (not cheaper, > but close enough to be reasonable). :-/ I bought 8 MBytes of RAM and two 760 MByte disks for the price of ONE SGI 780 MByte disk! Hardly close enough for me. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ John H. Merritt # Yesterday I knew nothing, Applied Research Corporation # Today I know that. merritt@iris613.gsfc.nasa.gov #   Received: from vmb.brl.mil by VMB.BRL.MIL id aa27926; 20 May 90 8:37 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa27855; 20 May 90 8:26 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa27826; 20 May 90 8:08 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa12414; 20 May 90 7:33 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA20834; Sun, 20 May 90 04:18:45 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 20 May 90 08:26:14 GMT From: John H Merritt Organization: Goddard Space Flight Center Climate and Radiation Branch Subject: Getting window id the mouse points to. Message-Id: <2101@dftsrv.gsfc.nasa.gov> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In a multiwindow program, I can't seem to find the routine which returns the window id for the window pointed to by the mouse. This example program opens up 2 windows and attempts to have two different popup menus for each window, however no matter which window the mouse is in the last window id is returned by 'winget'. Without going through a list of currently opened windows, is there a routine which returns the window id of the pointed to window? I do not want to have to do a winset and check the origin and dimension of each opened window to determine if the mouse is with in it. ---------- cut -------------------- /* * Multiwindow multiple popup menu example * * Can we have different popup menus in different windows? * * Answer is: */ #include #include #include int win1, win2; int menu1, menu2; main() { win1 = open_win("window 1"); printf("window 1: %d\n",win1); menu1 = make_menu1(); win2 = open_win("window 2"); printf("window 2: %d\n",win2); menu2 = make_menu2(); queue_devices(); wait_for_quit(); } int make_menu1() { int menu; menu = defpup("MENU 1 %t|item1|item2|item3"); return menu; } int make_menu2() { int menu; menu = defpup("MENU 2 %t|num 1|num 2|num 3"); return menu; } queue_devices() { qdevice(MENUBUTTON); qdevice(WINQUIT); } int open_win(name) char *name; { int winid; winid = winopen(name); return winid; } wait_for_quit() { int event; short val; int menu; int cur_win_id; do { switch(event = qread(&val)) case MENUBUTTON: if (val) { /* * THE MOUSE IS IN WHICH WINDOW ? * --- winget is not enough */ cur_win_id = winget(); printf("current window: %d\n",cur_win_id); if (cur_win_id == win1) { menu = dopup(menu1); printf("MENU1 item: %d\n",menu); } else if (cur_win_id == win2) { menu = dopup(menu2); printf("MENU2 item: %d\n",menu); } else printf("Not in valid window for menu.\n"); } } while (event != WINQUIT); } ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ John H. Merritt # Yesterday I knew nothing, Applied Research Corporation # Today I know that. merritt@iris613.gsfc.nasa.gov #   Received: from vmb.brl.mil by VMB.BRL.MIL id ab27926; 20 May 90 8:37 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab27855; 20 May 90 8:27 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab27826; 20 May 90 8:08 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa12446; 20 May 90 7:49 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA21636; Sun, 20 May 90 04:33:28 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 20 May 90 11:25:03 GMT From: Steve Lamont Organization: Foo Bar Brewers Cooperative Subject: Re: Why does a 3.5" floppy cost $1000?? Message-Id: <2180@speedy.mcnc.org> References: <4830@mace.cc.purdue.edu>, <2100@dftsrv.gsfc.nasa.gov> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <2100@dftsrv.gsfc.nasa.gov> merritt@iris613.UUCP (John H Merritt) writes: >In article <4830@mace.cc.purdue.edu> irg@mace.cc.purdue.edu (Daniel L. Severance) writes: >> SGI's prices on hard disks and memory are also quite reasonable. > ^^^^^unreasonable. >... I bought 8 MBytes of RAM and two 760 MByte disks for the price > of ONE SGI 780 MByte disk! Hardly close enough for me. Agreed. We bought our 280GTX with only 16 MB and went third party for the balance of the 128 MB. We wouldn't have been able to buy the machine if we had to pay SGI's prices! The salesperson mumbled something about warrantee service or lack thereof as the reason for the difference in price but, frankly, I can go out and buy a whole load of spares and still have bucks left over. In defense of SGI, who probably need no defense, they *have* lowered their memory prices to where they're only twice what the commodity market charges. With discounts and a little creative salespersonship we can get the prices down to something that only gives one high blood pressure, rather than cause to call out the paramedics! :-) Now, if I could just talk them into trading out the GTX boards for VGX and the 25 MHz boards for 33 MHz boards for free :-) :-) :-) spl (the p stands for preparing to hold my breath for a long time) -- Steve Lamont, sciViGuy (919) 248-1120 EMail: spl@ncsc.org NCSC (The other one), Box 12889, Research Triangle Park, NC 27709 "It must be swell to be that snooty." - Raymond Chandler, 1950, "Red Wind"   Received: from vmb.brl.mil by VMB.BRL.MIL id aa29018; 20 May 90 14:58 EDT Received: from vmb.brl.mil by VMB.brl.MIL id aa28970; 20 May 90 14:47 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa28931; 20 May 90 14:34 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa13171; 20 May 90 14:18 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA07926; Sun, 20 May 90 11:06:46 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 20 May 90 18:05:20 GMT From: Tim Hall Organization: Boston University Computer Graphics Lab Subject: lmcolor and normals Message-Id: <57555@bu.edu.bu.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In displaying a multi-colored planar mesh I use the routine lmcolor( LMC_DIFFUSE ) inorder to use the routine "cpack" to specify the vertex colors. The man page for lmcolor led me to believe that after the call to cpack the material color in the hardware should now be the color specified in cpack. However, experience shows that you have to specify the vertex normal after the cpack call before the color takes effect. Is this the correct behavior? Why isn't the already specified normal used? (Since my mesh is planar I had dreams of only specifing the normal once.) -Tim Hall tjh@bu-pub.bu.edu   Received: from vmb.brl.mil by VMB.BRL.MIL id aa29242; 20 May 90 16:01 EDT Received: from vmb.brl.mil by VMB.brl.MIL id ab28970; 20 May 90 14:47 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab28931; 20 May 90 14:34 EDT Received: from cs.utah.edu by VGR.BRL.MIL id aa13187; 20 May 90 14:22 EDT Received: from adenosine.pharm.utah.edu by cs.utah.edu (5.61/utah-2.11-cs) id AA23746; Sun, 20 May 90 12:22:22 -0600 Date: Sun, 20 May 90 12:28:26 MDT From: "Darrell R. Davis" Posted-Date: Sun, 20 May 90 12:28:26 MDT Message-Id: <9005201828.AA12733@adenosine.pharm.utah.edu> Received: by adenosine.pharm.utah.edu (5.52/5.51) id AA12733; Sun, 20 May 90 12:28:26 MDT To: info-iris@BRL.MIL In-Reply-To: Tim_Buxton@um.cc.umich.edu's message of Sat, 19 May 90 22:59:34 EDT <6226910@um.cc.umich.edu> Subject: We need help getting IRISPLOT, Please on Sat, 19 May 90 22:59:34 EDT, Tim_Buxton@um.cc.umich.edu said: Tim> We would like to get IRISPLOT via anonymous ftp. I have Tim> contacted system@math.arizona.edu and gotten the Tim> someone else who has IRISPLOT make it available over the I have Irisplot up and running, though haven't used it much so far. You can get it via anonymous ftp from adenosine.pharm.utah.edu (128.110.74.1). Be advised that it may disappear without notice. The file is Irisplot.2.0.tar.Z -------------------------------------------------------------- * Darrell R. Davis * * * "Faster, faster, until the Assistant Professor * * * thrill of speed overcomes Medicinal Chemistry *A**L**T**A* the fear of death." University of Utah * * * * * * --H.S. Thompson * --------------------------------------------------------------   Received: from vmb.brl.mil by VMB.BRL.MIL id aa01044; 20 May 90 22:02 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa01004; 20 May 90 21:52 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa00992; 20 May 90 21:39 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa14228; 20 May 90 21:19 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA28889; Sun, 20 May 90 18:10:19 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 20 May 90 22:45:21 GMT From: "Louis M. McDonald" Organization: The Aerospace Corporation, El Segundo, CA Subject: Message-Id: <73603@aerospace.AERO.ORG> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL I have been trying to compiling some of the X software (XView), and have run into a problem with the file. It makes a reference to a typedef of ".globl", however, I cannot find where this is defined (another .h file or part of cpp?). This currently is putting a halt on the building of XView (or at least an attempt at it). Anyone have any clues into this ".globl" reference? -- Louis McDonald Internet: louis@aerospace.aero.org The Aerospace Corporation 213-336-8914   Received: from vmb.brl.mil by VMB.BRL.MIL id aa01243; 20 May 90 23:02 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab01044; 20 May 90 22:10 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa01024; 20 May 90 21:55 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa14290; 20 May 90 21:49 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA00564; Sun, 20 May 90 18:43:17 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 21 May 90 01:42:37 GMT From: Tim Hall Organization: Boston University Computer Graphics Lab Subject: While on the subject of meshes.... Message-Id: <57560@bu.edu.bu.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL I saw in the latest "Iris World" that the VGX supports quad meshes. When rendering a quad mesh does the VGX break it up into triangles? For example, I use the quad mesh to render a square with a different color at each vertex. Is the pixel in the center of the square an average of all the colors at the four verticies or is it an average of two diagonal verticies? -Tim Hall tjh@bu-pub.bu.edu   Received: from vmb.brl.mil by VMB.BRL.MIL id aa02151; 21 May 90 1:20 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa01785; 21 May 90 0:42 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa01772; 21 May 90 0:31 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa14811; 21 May 90 0:20 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA08145; Sun, 20 May 90 21:12:45 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 20 May 90 20:23:47 GMT From: Fred Fish Organization: Amiga Library Services Subject: Re: [comp.sys.sgi] Backing up, possible solution. Message-Id: <107@fishpond.UUCP> References: <12027@stag.math.lsa.umich.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <12027@stag.math.lsa.umich.edu> davis@ADENOSINE.PHARM.UTAH.EDU ("Darrell R. Davis") writes: > Some of what I think are major >advantages to gtar over 'regular' tar and bru are: >1- ability to make multi-volume archives. Obviously first on the list. BRU does this just fine thank you. In fact, at the lowest I/O level it implements a virtual archive of infinite length in which (in theory) any block from any volume can be accessed randomly. The actual archiving code that calls upon these virtual archive routines does not in fact make use of much of these features at this time. >2- compatibility with primitive tar, that is, gtar can read tar files. > I'm pretty sure that in the default mode (no multi-volumes or other > fancy stuff.) tar and gtar are completely compatible. Definitely a plus for gtar. The first version of BRU also had this feature but the archive format was changed to make it more reliable (checksums on EVERY block, block sequence numbers, file sequence numbers, etc). >3- portability, you could put it on your convex, suns, iris, etc. BRU is probably more portable that gtar. We have yet to find any Unix like system where a port takes more than a few hours. On most systems, the autoconfiguration feature figures out everything necessary and installation is as simple as typing "make". >4- you can specify filesystems *not* to include in a full dump, this > is particulary nice for server types with large storage. This is arguably something that should be handled by a frontend process, be it a shell script or some archive management process. Once implemented, "normal" tar, gtar, or BRU should work just fine with it. >5- gtar comes with a nice manual written in TeX. As does BRU (the manual is written in AmigaTeX). Although the manual source is not normally included as part of the source distribution, we would have no qualms about supplying it to source customers. -Fred (just keeping the record straight) -- # Fred Fish, 1835 E. Belmont Drive, Tempe, AZ 85284, USA # 1-602-491-0048 asuvax!mcdphx!fishpond!fnf   Received: from vmb.brl.mil by VMB.BRL.MIL id aa03001; 21 May 90 3:01 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab02751; 21 May 90 2:19 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab02625; 21 May 90 2:04 EDT Received: from uxa.cso.uiuc.edu by VGR.BRL.MIL id aa15024; 21 May 90 1:50 EDT Received: by uxa.cso.uiuc.edu (5.61+/IDA-1.2.8) id AA17662; Mon, 21 May 90 00:50:10 -0500 Date: Mon, 21 May 90 00:50:10 -0500 From: Marc/TALK TO ME Message-Id: <9005210550.AA17662@uxa.cso.uiuc.edu> To: info-iris@BRL.MIL Subject: more on meshes Speaking of meshes... Between the two ways of rendering a planar surface in 3D - that is, specifying each 4-vertex polygon at a time vs. using a T-mesh - which is faster as far as drawing speed is concerned? Using the T-mesh approximately halves the number of vertices that need to be specified, but twice as many polygons must be rendered... which of these is more costly? ---Marc Andreessen Materials Research, Univ of Illinois andreessen%uimrl.dnet@uimrl5.mrl.uiuc.edu   Received: from vmb.brl.mil by VMB.BRL.MIL id aa02851; 21 May 90 2:30 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa01977; 21 May 90 1:10 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa01967; 21 May 90 1:02 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa14872; 21 May 90 0:49 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA09390; Sun, 20 May 90 21:38:01 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 21 May 90 04:36:27 GMT From: Edward Vielmetti Organization: University of Michigan Math Dept., Ann Arbor MI. Subject: Re: [comp.sys.sgi] Backing up, possible solution. Message-Id: References: <12027@stag.math.lsa.umich.edu>, <107@fishpond.UUCP> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL [comparision of BRU and gnu tar] I'm not terribly familiar with BRU -- where can it be had from, what does it cost, etc. Can it be run a la rdump so that the disk to be dumped is not the system with the tape drive? Is source available? thanks. --Ed Edward Vielmetti, U of Michigan math dept comp.archives moderator   Received: from vmb.brl.mil by VMB.BRL.MIL id aa06204; 21 May 90 8:53 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa05638; 21 May 90 8:43 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa05608; 21 May 90 8:32 EDT Received: from aero4.larc.nasa.gov by VGR.BRL.MIL id aa17176; 21 May 90 8:16 EDT Received: Mon, 21 May 90 08:17:31 EDT by aero4.larc.nasa.gov (5.52/5.6) Date: Mon, 21 May 90 08:17:31 EDT From: "Brent L. Bates AAD/TAB MS361 x42854" Message-Id: <9005211217.AA25849@aero4.larc.nasa.gov> To: dftsrv!iris613.gsfc.nasa.gov!merritt@ames.arc.nasa.gov Subject: Re: Getting window id the mouse points to. Cc: info-iris@BRL.MIL I am guessing, but would a qread on "inptch" help. The value returned is the "current" window id. -- Brent L. Bates NASA-Langley Research Center M.S. 361 Hampton, Virginia 23665-5225 (804) 864-2854 E-mail: blbates@aero4.larc.nasa.gov or blbates@aero2.larc.nasa.gov   Received: from vmb.brl.mil by VMB.BRL.MIL id aa06683; 21 May 90 9:04 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab05638; 21 May 90 8:43 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa05611; 21 May 90 8:33 EDT Received: from aero4.larc.nasa.gov by VGR.BRL.MIL id aa17199; 21 May 90 8:23 EDT Received: Mon, 21 May 90 08:24:07 EDT by aero4.larc.nasa.gov (5.52/5.6) Date: Mon, 21 May 90 08:24:07 EDT From: "Brent L. Bates AAD/TAB MS361 x42854" Message-Id: <9005211224.AA25877@aero4.larc.nasa.gov> To: zaphod.mps.ohio-state.edu!usc!orion.oac.uci.edu!fredv@tut.cis.ohio-state.edu Subject: Re: IRIX support of BSD printers Cc: info-iris@BRL.MIL Try "mknetpr" there is a man page for it on the system. What this does is create an filter in the /usr/spool/lp/interface directory. When you send a file to the remote printer, the script rcp's the file to the machine with the remote printer, then rsh's the proper printer commands on the remote machine. On our 3130 it took me a while to modify what mknetpr produce into something that would work, but it gave me a good starting point. You will need the remote machine to "trust" your SGI in order for the remote commands to work. -- Brent L. Bates NASA-Langley Research Center M.S. 361 Hampton, Virginia 23665-5225 (804) 864-2854 E-mail: blbates@aero4.larc.nasa.gov or blbates@aero2.larc.nasa.gov   Received: from vmb.brl.mil by VMB.BRL.MIL id ab07160; 21 May 90 9:30 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab06683; 21 May 90 9:13 EDT Received: by VMB.BRL.MIL id ag06430; 21 May 90 9:02 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa05332; 21 May 90 8:15 EDT Received: from aero4.larc.nasa.gov by VGR.BRL.MIL id aa17125; 21 May 90 8:04 EDT Received: Mon, 21 May 90 08:05:10 EDT by aero4.larc.nasa.gov (5.52/5.6) Date: Mon, 21 May 90 08:05:10 EDT From: "Brent L. Bates AAD/TAB MS361 x42854" Message-Id: <9005211205.AA25757@aero4.larc.nasa.gov> To: dftsrv!iris613!merritt@ames.arc.nasa.gov Subject: Re: Why does a 3.5" floppy cost $1000?? Cc: info-iris@BRL.MIL irg@mace.cc.purdue.edu thinks SGI's prices are reasonable, and John H. Merritt thinks they aren't. We must compare apples with apples here. Purdue, being a uniniversity (more specific non U.S. Government) gets (I am taking an educated guess) a 50% discount on all hardware, while Mr. Merritt apparently working for the U.S. Government (gsfc, Goddard Space Flight Center-NASA) only gets about 23% "discount". From what I have heard from quite a few people, "everyone" not just educational institutions, get a 50% discount, except the U.S. Government. Now if I could get SGI hardware for 50% off, their prices would be reasonable to me too. However, as you can see where I work we don't usually and they aren't reasonable. (Standard disclaimers apply. My opinion.) -- Brent L. Bates NASA-Langley Research Center M.S. 361 Hampton, Virginia 23665-5225 (804) 864-2854 E-mail: blbates@aero4.larc.nasa.gov or blbates@aero2.larc.nasa.gov   Received: from vmb.brl.mil by VMB.BRL.MIL id aa07414; 21 May 90 9:40 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa07160; 21 May 90 9:30 EDT Received: by VMB.BRL.MIL id ai06821; 21 May 90 9:12 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab05332; 21 May 90 8:16 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa17134; 21 May 90 8:04 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA00585; Mon, 21 May 90 05:00:47 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 21 May 90 10:05:21 GMT From: Bruno Pape Organization: Silicon Graphics S.A., Zuerich, Switzerland Subject: Re: Compiler signal code ? (was:Re: GNU tar, compilation thereof.) Message-Id: <1990May21.100521.27359@sgzh.uucp> References: <1990May15.102109.729@urz.unibas.ch> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <1990May15.102109.729@urz.unibas.ch> doelz@urz.unibas.ch writes: > >I got it from there, no problem, and tried again; > ># make > cc -g -DUSG -I/usr/include/bsd -DDEF_AR_FILE=\"/dev/tape\" > -DDEFBLOCKING=20 -c tar.c > >Fatal error in: /usr/lib/ugen - core dumped > Signal 139 *** Error code 1 > >Stop. > > I suspect that the ugen is sick on 4D > 20. Mine is ># ls -ls /usr/lib/ugen > 609 -rwxr-xr-x 1 bin bin 311344 Dec 6 09:20 /usr/lib/ugen > >which is the primary installation date of 3.2 on our system. >Currently we're running 3.2.2 but obviously, things didn't change. > >- Reinhard > If anyone in Mountain View is following this and would like to look into it Reinhard is running a 4D/120 with the IP5 processors. The routine compiled correctly on a 4D/20 and a 4D/220, I think that's IP6 and IP7 processors respectively. I will forward this problem to customer support. Thanks, Bruno   Received: from vmb.brl.mil by VMB.BRL.MIL id aa11133; 21 May 90 13:26 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab10813; 21 May 90 13:15 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa10751; 21 May 90 13:06 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa00506; 21 May 90 12:50 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA16360; Mon, 21 May 90 09:44:57 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 21 May 90 16:22:14 GMT From: George Elkins Organization: Rutgers Univ., New Brunswick, N.J. Subject: getrusage Message-Id: Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Is there any SGI Irix equivalent to the "getrusage (2)" routine found in bsd unix? Here is part of the "man 2 getrusage" output from BSD Unix: --------------------------------------------------- NAME getrusage - get information about resource utilization SYNOPSIS #include getrusage(who, rusage) int who; struct rusage *rusage; DESCRIPTION Getrusage returns information describing the resources util- ized by the current process, or all its terminated child process. [...] --------------------------------------------------- I have some benchmark code which calls this routine. George Elkins   Received: from vmb.brl.mil by VMB.BRL.MIL id ab11133; 21 May 90 13:26 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ac10813; 21 May 90 13:15 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab10751; 21 May 90 13:06 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa00509; 21 May 90 12:50 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA16651; Mon, 21 May 90 09:49:05 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 21 May 90 16:26:31 GMT From: Vernon Schryver Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Re: bsd include files... Message-Id: <60683@sgi.sgi.com> References: <9005192210.AA18900@mcirps2.med.nyu.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL /usr/include/make/common{rules,defs} and /usr/sbin/mkdepend are what we use to maintain dependencies. It uses the -M arg. to cpp. You may not be able to use /usr/include/make directly. It might be profitable to see the results lf a large number of hours, even if only to see things to avoid. Besides "depend", there are bunch of other useful targets including "clean", "clobber", and "tags". A typical makefile for a command looks like the enclosed. This is a genuine example, with the name of the command and its .c files replaced with "foo" to avoid excitement. #ident "$Revision: 1.7 $" include $(ROOT)/usr/include/make/commondefs MKDEPFILE=Makedepend CFILES =foo.c bar.c other.c TARGETS =foo remotefoo I_FLAGS =-idb "foo.sw.foo" CI_FLAGS=-idb "foo.sw.foo config(suggest)" LLDLIBS =-lsun -lbsd # do not link with shared libraries, since says not to. SHDLIBC = default:$(TARGETS) include $(COMMONRULES) install:$(TARGETS) $(INSTALL) $(I_FLAGS) -F /usr/etc "foo" $(INSTALL) $(CI_FLAGS) -F /usr/etc "remotefoo" foo: $(OBJECTS) $(CCF) $(OBJECTS) $(LDFLAGS) -o $@   Received: from vmb.brl.mil by VMB.BRL.MIL id aa11271; 21 May 90 13:37 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa10813; 21 May 90 13:15 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa10719; 21 May 90 13:04 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa00440; 21 May 90 12:36 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA15432; Mon, 21 May 90 09:31:47 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 21 May 90 15:40:46 GMT From: Dave Olson Subject: Re: Why does a 3.5" floppy cost $1000?? Message-Id: <7925@odin.corp.sgi.com> References: <4830@mace.cc.purdue.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL irg@mace.cc.purdue.edu (Daniel L. Severance) writes: | Hi!, | First, let me say that I have found SGI to have fast machines at a great | price. Ok, now for the bad news: | Could someone from SGI or elsewhere tell me why a 3 1/2" floppy drive | for a personal IRIS lists for $1000, while the one in our SPARCstation I | lists (from SUN) for $250 ??? | | Even $300-$400 would be ok, but $1000?? There is NO WAY that | a $150 PC floppy with a SCSI adapter has to cost $1000! | My sales rep. in New Haven (we're moving to Yale in June) said he | talked to the product manager about the unreasonable price, and was | told that "well, we don't sell that many of them, anyway". Frankly, | I can't say I'm surprised. First, this is strictly unofficial, and my personal understanding of the issues; no promises SGI will do anything I say, etc., etc. Unfortunately, at the time qualification started, there were NO 3 1/2" floppy drives with an embedded SCSI controller (there were supposed to have been, but the mfgr's schedule was a bit, um, optimistic). We therefore had a drive and controller integrated by a distribution firm as an interim solution. Since the quantities were low, the price to us is high (VERY high for a peripheral of this sort). The same applies to the 5 1/4" floppy. Since then floppy drives with embedded SCSI have become available, and we will probably end up qualifying them. When we will release them is not known (by me). | There are 3 research groups here at Purdue Chem, and 1 central | facility which would all be interested in the SoftPC/AT emulator AND | floppy drives. The software isn't much use if you don't have easy | access to a floppy, and no one is willing to plunk down $1000 (- discount) | for the things. The floppies are also quite nice for archiving small | amounts of data. ... | In lieu of spending $1000, does anyone know how to make use of the | floppy in a SPARCstation I over the network with the SoftPC emulator | on the IRIS? What you can do is create a logical floppy, if I remember correctly. This is just a unix file that is an image of a floppy. Check the SoftPC documentation on this. If I am remembering correctly, you should just be able to rcp this image to a machine with a floppy and dd it into the floppy, and vice-versa. You can also set up a logical PC hard disk that is simply a unix directory (drive E:, I think), so you could simply rcp the actual MSDOS files you are interested in, and put them in that directory. -- Dave Olson Life would be so much easier if we could just look at the source code.   Received: from vmb.brl.mil by VMB.BRL.MIL id aa15516; 21 May 90 17:26 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa15373; 21 May 90 17:15 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa15296; 21 May 90 16:59 EDT Received: from TROUT.NOSC.MIL by VGR.BRL.MIL id aa00861; 21 May 90 16:49 EDT Received: from ucsd.edu by trout.nosc.mil (5.59/1.27) id AA07385; Mon, 21 May 90 13:48:38 PDT Received: from chema.ucsd.edu by ucsd.edu; id AA08692 sendmail 5.61/UCSD-2.0-sun via SMTP Mon, 21 May 90 13:48:34 -0700 for @nosc.mil:info-iris@brl.mil Received: by chem.chem.ucsd.edu (5.51) id AA22371; Mon, 21 May 90 13:47:57 PDT Date: Mon, 21 May 90 13:47:57 PDT From: Steve Dempsey Message-Id: <9005212047.AA22371@chem.chem.ucsd.edu> To: info-iris@BRL.MIL Subject: counting vertical retraces Does anybody know of a way to determine the number of vertical retraces that have occured between two points in time on a 4D? I am trying to interface some inexpensive ($50) stereo viewers to a 4D/25TG. The basic algorithm is to put the left view in one buffer and the right view in the other, and then enter a tight loop swapping buffers as fast as possible (60Hz) until some input event occurs. The first time into this loop the user may have to throw a switch to get the right image going to the right eye. An input event signals that new stereo images need to be generated, but if an odd number of vertical retraces occur while the new images are being rendered, the correct eye phasing will be lost and the depth effect will invert when the buffer swap loop is reentered. The problem I need to solve is to find a way to maintain the correct eye phasing after the initial manual sync. If I can determine that the number of vertical retraces that have occured since leaving the loop is odd, I can just wait for one more with 'gsync'. Of course, if the IRIS supported 'real' quad buffering I could have full resolution dynamic stereo at 30Hz like we had on our E&S PS1 15 years ago. Yes, I realize that the 30Hz flicker is noticeable, but it is not considered objectionable when compared to the benefits of dynamic stereo or the problems dealing with non-stereo windows which are stretched down the screen by the current quad buffer technique (StereoView). -------------------------------------------------------------------------------- Steve Dempsey (619) 534-0208 Dept. of Chemistry Computer Facility, B-014 INTERNET: sdempsey@ucsd.edu University of Calif. at San Diego BITNET: sdempsey@ucsd La Jolla, CA 92093 UUCP: ucsd!sdempsey   Received: from vmb.brl.mil by VMB.BRL.MIL id ab15516; 21 May 90 17:26 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab15373; 21 May 90 17:15 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa15298; 21 May 90 16:59 EDT Received: from SGI.COM by VGR.BRL.MIL id ab00866; 21 May 90 16:50 EDT Received: from tom.dallas.sgi.com by sgi.sgi.com via UUCP (5.52/900423.SGI) for info-iris@brl.mil id AA10079; Mon, 21 May 90 13:49:29 PDT Received: from tom.dallas.sgi.com by sgidal.dallas.sgi.com (5.52/891101.SGI) for sgi!BRL.MIL!info-iris id AA07227; Mon, 21 May 90 15:54:40 CDT Received: by tom.dallas.sgi.com (5.52/890619.SGI) (for @sgidal.dallas.sgi.com:info-iris@BRL.MIL) id AA09505; Mon, 21 May 90 15:54:38 CDT Date: Mon, 21 May 90 15:54:38 CDT From: "root@sgidallas" Message-Id: <9005212054.AA09505@tom.dallas.sgi.com> To: info-iris@tom.dallas.sgi.com Subject: Syquest removable drives HI All: Does anyone have any experience with using the Syquest 44MB removable fixed drive as a block device? Model # SQ555. Any experience with any Syquest drives at all? Thanks in advance......... -- Tom Reed SGI - Dallas email: treed@sgidal.dallas.sgi.com vmail: 8705 phone: 214-788-4122   Received: from vmb.brl.mil by VMB.BRL.MIL id aa15677; 21 May 90 17:48 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ac15516; 21 May 90 17:37 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa15433; 21 May 90 17:16 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa00947; 21 May 90 17:05 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA01955; Mon, 21 May 90 13:50:44 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 21 May 90 21:27:39 GMT From: Todd Michael Carrico Organization: Washington University, St. Louis MO Subject: Help building X11R4 libs on Personal Iris Message-Id: <1990May21.212739.8851@cec1.wustl.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL We have a Personal Iris under IRIX 3.2. We have the full distribution of X11R4. Has anyone build the full X11R4 for this machine, and if so could you answer a few questions. The sgi.cf file is there, but we are missing several files which are machine specific. Also, does anyone know of an ftp site which has X11R4 for the Personal Iris which works? Any comments or suggestions would be helpful. Thanks in advance... ---------------------------------------------------------------- Send reply via Email to todd@wucs1 Washington University St. Louis, MO   Received: from vmb.brl.mil by VMB.BRL.MIL id ab15677; 21 May 90 17:48 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ad15516; 21 May 90 17:37 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab15433; 21 May 90 17:16 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa00950; 21 May 90 17:05 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA02600; Mon, 21 May 90 14:00:39 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 21 May 90 20:57:21 GMT From: Gerry Roston Organization: Carnegie-Mellon University, CS/RI Subject: Help with terminfo Message-Id: <9355@pt.cs.cmu.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Can someone please tell me where I can find the program infocmp. I have searched the disk, but to no avail. Is it possibly on the distribution tape and I didn't load it onto my disk? We have an Iris 4d. thanx gerry -- gerry roston, field robotics center robotics institute, carnegie mellon university pittsburgh, pennsylvania, 15213 (412) 268-6557 gerry@cive.ri.cmu.edu   Received: from vmb.brl.mil by VMB.BRL.MIL id aa16104; 21 May 90 18:18 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa15762; 21 May 90 18:07 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa15736; 21 May 90 17:53 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa01107; 21 May 90 17:35 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA04320; Mon, 21 May 90 14:28:14 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 21 May 90 21:19:12 GMT From: Mark Moraes Organization: Department of Computer Science, University of Toronto Subject: Re: DWB, LaserWriter support and the screen Message-Id: <90May21.171808edt.2626@smoke.cs.toronto.edu> References: <9005181608.AA00871@castor.wustl.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL weinhous@CASTOR.WUSTL.EDU ("Martin S. Weinhous") writes: >We just purchased DWB and LaserWriter support software from SGI. >How does one preview a postscript document on the screen? If you don't embed custom PostScript in documents, only want to preview the output of ditroff, have X running on your display, can compile X programs, and have the X fonts that come standard with MIT X.V11R3/R4, then xtroff should do a reasonable job. I know it compiles and works on SGIs because we use it, and someone with a display says that it does seem to work with Xsgi (or whatever -- we don't have Iris displays, so we don't use SGI's X distribution) You can get xtroff from the X.V11R4 contrib distribution in clients/ or from your friendly neighbourhood X archive site. (eg. expo.lcs.mit.edu in contrib/xtroff.tar.Z, or cs.toronto.edu in pub/X/xtroff.tar.Z) The R4 version should compile and work with R3. In general, ditroff and dvi previewers seem to be much faster than PostScript preview of psdit/dvi2ps output. Mark.   Received: from vmb.brl.mil by VMB.BRL.MIL id aa16494; 21 May 90 19:40 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa16378; 21 May 90 19:30 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa16355; 21 May 90 19:16 EDT Received: from SGI.COM by VGR.BRL.MIL id aa01486; 21 May 90 18:56 EDT Received: from oni.sgi.com by sgi.sgi.com (5.52/900423.SGI) for info-iris@brl.mil id AA15512; Mon, 21 May 90 15:56:19 PDT Received: from sonata.sgi.com by relay.sgi.com (5.52/891101.SGI) for @sgi.sgi.com:info-iris@brl.mil id AA01693; Mon, 21 May 90 15:56:15 PDT Received: by sonata.sgi.com (5.52/890923.SGI) for @relay.sgi.com:info-iris@brl.mil id AA01475; Mon, 21 May 90 15:56:13 PDT Date: Mon, 21 May 90 15:56:13 PDT From: Rolf van Widenfelt Message-Id: <9005212256.AA01475@sonata.sgi.com> To: info-iris@BRL.MIL Subject: Re: lmcolor and normals (and more on meshes) > The man page for lmcolor led me > to believe that after the call to cpack the material color in > the hardware should now be the color specified in cpack. However, > experience shows that you have to specify the vertex normal after > the cpack call before the color takes effect. Is this the correct > behavior? Why isn't the already specified normal used? You didn't mention what specific graphics you have, so I'll assume you have a GT. You're right.. the normal needs to be specified after the color is set but before the vertex is called. Although this behavior may seem strange, consider the alternative: to make your case work, we should evaluate lighting at the cpack call and also at the normal call. Unfortunately, cases which have non-flat meshes would run slower because lighting is being evaluated twice for each vertex, the first of which is unnecessary. The non-flat mesh is probably the more common case, and so the current implementation is probably a better choice. By the way, our lighting performance benchmarks typically assume that a normal is provided with each vertex, so specifying normals at each vertex is not so unusual. next question... > Between the two ways of rendering a planar surface in 3D - that is, > specifying each 4-vertex polygon at a time vs. using a T-mesh - which is > faster as far as drawing speed is concerned? For an average surface in 3D, quads can be a more efficient primitive than a triangle mesh on GT/GTX. The GTX benchmarks (as I remember them) of 100,000 quads/s or 135,000 tmeshes/s indicate this. This is really just a suggestion because performance depends on many things, such as the amount of data being sent to the Geometry Engines and the number of pixels that are filled. This guideline is not true of PowerVision, which is better tuned for triangles, but quads still perform quite well. next... > When rendering a quad mesh does the VGX (PowerVision) break it up into triangles? > For example, I use the quad mesh to render a square with a different > color at each vertex. Is the pixel in the center of the square an > average of all the colors at the four verticies or is it an average > of two diagonal verticies? The underlying area filling primitive of PowerVision is the triangle, so your quad becomes two triangles. A quad may not be planar in color, and so we don't define what the color in the center should be. Actually, the center of some quads is not even inside the quad. If the vertexes of your quad have radically different colors, further subdivision is suggested to avoid mach banding artifacts. Now, a triangle is planar, so the color in the center is roughly the average of the vertex colors. I hope this helps a little... Rolf van Widenfelt SGI/ASD Graphics   Received: from vmb.brl.mil by VMB.BRL.MIL id aa16692; 21 May 90 20:11 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa16561; 21 May 90 20:01 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa16545; 21 May 90 19:48 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa01604; 21 May 90 19:36 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA11962; Mon, 21 May 90 16:31:34 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 21 May 90 23:28:29 GMT From: Tim Hall Organization: Boston University Subject: Re: counting vertical retraces Message-Id: <57608@bu.edu.bu.edu> References: <9005212047.AA22371@chem.chem.ucsd.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Use the timer devices. If you do a "qdevice( TIMER0 )", this will generate an event every 1/60th of a second. I don't remember if it is directly tied into the screen retrace though. -Tim tjh@bu-pub.bu.edu   Received: from vmb.brl.mil by VMB.BRL.MIL id aa16994; 21 May 90 20:32 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab16692; 21 May 90 20:21 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa16674; 21 May 90 20:06 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa01659; 21 May 90 19:50 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA12658; Mon, 21 May 90 16:43:22 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 21 May 90 20:01:47 GMT From: sgi!shinobu!odin!inferno.esd.sgi.com!rmr@ucbvax.berkeley.edu Subject: Re: DWB, LaserWriter support and the screen Message-Id: <7935@odin.corp.sgi.com> References: <9005182126.AA18797@aero4.larc.nasa.gov> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL >How does one preview a postscript document on the screen? If you have DWB and Laser Printer support, you can use the -t option to psroff to create a postscript file: psroff doc.troff -t > doc.ps You can then use psview to preview the document: psview doc.ps There are several flags to psview that can help you scale the window properly (check the man page). psview doesn't work great on all postscript files, but works fine on those created with troff, eqn, tbl, and (I'm pretty sure) pic. Robert Reimann ESD Developer Documentation   Received: from vmb.brl.mil by VMB.BRL.MIL id ab16994; 21 May 90 20:32 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ac16692; 21 May 90 20:22 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab16674; 21 May 90 20:06 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa01661; 21 May 90 19:50 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA12856; Mon, 21 May 90 16:45:56 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 21 May 90 23:36:09 GMT From: Kurtis MacFerrin Organization: Schreiber Group (Harvard Chemistry Department) Subject: Anyone use MultiNet software? Message-Id: <3016@husc6.harvard.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Hi, Does anyone out there use MultiNet software? We have a IRIS 4D/85GT without a tape drive, and a VAXstation 3500 with a TK70 tape drive, and we'd like to back the iris onto the vax tape drive. The iris runs Irix 3.2.1, and the vax runs VMS 4.7 with MultiNet 2.1 software to handle TCP/IP stuff. Supposedly the MultiNet software supports the RMT protocol, but I haven't been able to get the two machines to talk rmt to each other. Everything else (such as ftp, telnet, finger, mail, nfs, and lpd) works fine. Does anyone out there have any experience with MultiNet and IRISes? I will summarize for the net any reponses I get. Thanks. Kurtis MacFerrin macferrin@slsvax.harvard.edu or @huche1.bitnet   Received: from vmb.brl.mil by VMB.BRL.MIL id ac16994; 21 May 90 20:32 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ad16692; 21 May 90 20:22 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ac16674; 21 May 90 20:06 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa01666; 21 May 90 19:51 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA12699; Mon, 21 May 90 16:43:44 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 21 May 90 23:06:48 GMT From: Ian Clements Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Re: Why does a 3.5" floppy cost $1000?? Message-Id: <60713@sgi.sgi.com> References: <4830@mace.cc.purdue.edu>, <7925@odin.corp.sgi.com> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <7925@odin.corp.sgi.com>, olson@anchor.esd.sgi.com (Dave Olson) writes: > What you can do is create a logical floppy, if I remember correctly. > This is just a unix file that is an image of a floppy. Check the > SoftPC documentation on this. If I am remembering correctly, you should > just be able to rcp this image to a machine with a floppy and > dd it into the floppy, and vice-versa. Actually, this is an image of a hard disk. These files can be rcp'ed over the net to another IRIS. Another trick is to NFS mount the directories that contain a number of these 'container' files where ever you use SoftPC. The caveat is that most PC applications do not support file locking so access by more than one user must be controlled. DOS 'container' files can be anywhere from 1 to 32 Megs in size. > You can also set up a logical PC hard disk that is simply a unix > directory (drive E:, I think), so you could simply rcp the actual MSDOS > files you are interested in, and put them in that directory. This works well too. However, applications such as Time-Line which have the ability to run both singley and on network drives will often not run because of licensing conflicts (ie; you're running on the network drive therefore you must be running more than one copy). Both forms of disk access, container files and network drives, offer about the same performance characteristics (wrt SoftPC/AT). Personally, I prefer running things from the network drive since that way, I can use IRIX utilities to move/sort PC files. Cheers, Ian A friend with weed is a friend indeed.   Received: from vmb.brl.mil by VMB.BRL.MIL id aa17145; 21 May 90 20:50 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ad16994; 21 May 90 20:39 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa16954; 21 May 90 20:22 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa01755; 21 May 90 20:09 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA13560; Mon, 21 May 90 16:56:21 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 21 May 90 21:38:08 GMT From: Archer Sully Subject: Re: Getting window id the mouse points to. Message-Id: <7944@odin.corp.sgi.com> References: <2101@dftsrv.gsfc.nasa.gov> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <2101@dftsrv.gsfc.nasa.gov> merritt@iris613.gsfc.nasa.gov (John H Merritt) writes: >In a multiwindow program, >I can't seem to find the routine which returns the window >id for the window pointed to by the mouse. This example >program opens up 2 windows and attempts to have two different >popup menus for each window, however no matter which window >the mouse is in the last window id is returned by 'winget'. >Without going through a list of currently opened windows, is there >a routine which returns the window id of the pointed to window? >I do not want to have to do a winset and check the origin and >dimension of each opened window to determine if the mouse is with >in it. >---------- cut -------------------- >/* > * Multiwindow multiple popup menu example > * > * Can we have different popup menus in different windows? > * > * Answer is: YES > */ Lot's o' stuff delted.... winget merely gets the id of the window that you last did a winset to, so clearly it isn't the beast you want. What you should look at is the INPUTCHANGE event. The value associated with INPUTCHANGE is the window id of the window receiving input focus, or zero if the program is loosing focus. Thus... switch(event = qread(&val)) case INPUTCHANGE: cur_win_id = val; break; -- Archer Sully | I'm 27 years old. That's 54 in Nerd Years (archer@esd.sgi.com) | -- Keith Rienzi   Received: from vmb.brl.mil by VMB.BRL.MIL id ab17145; 21 May 90 20:50 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ae16994; 21 May 90 20:39 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa16961; 21 May 90 20:22 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa01767; 21 May 90 20:10 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA13529; Mon, 21 May 90 16:56:02 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 21 May 90 21:28:54 GMT From: Thant Tessman Organization: Silicon Graphics Inc. Subject: Re: Getting window id the mouse points to. Message-Id: <7942@odin.corp.sgi.com> References: <2101@dftsrv.gsfc.nasa.gov> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <2101@dftsrv.gsfc.nasa.gov>, merritt@iris613.gsfc.nasa.gov (John H Merritt) writes: > In a multiwindow program, > I can't seem to find the routine which returns the window > id for the window pointed to by the mouse. This example > program opens up 2 windows and attempts to have two different > popup menus for each window, however no matter which window > the mouse is in the last window id is returned by 'winget'. There is a pseudodevice called INPUTCHANGE. When the mouse moves into a window, the window id is put on the queue. When the mouse moves out of the window, a zero is put on the queue; > > wait_for_quit() > { > int event; > short val; > int menu; > int cur_win_id; > > > do > { > switch(event = qread(&val)) case INPUTCHANGE: cur_win_id = val; break; > case MENUBUTTON: > if (val) > { > /* > * THE MOUSE IS IN WHICH WINDOW ? > * --- winget is not enough > */ > cur_win_id = winget(); ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Don't do this. Winget returns current graphics id. > printf("current window: %d\n",cur_win_id); > if (cur_win_id == win1) > { > menu = dopup(menu1); > printf("MENU1 item: %d\n",menu); > } > else if (cur_win_id == win2) > { > menu = dopup(menu2); > printf("MENU2 item: %d\n",menu); > } > else > printf("Not in valid window for menu.\n"); > } > } while (event != WINQUIT); > > } > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > John H. Merritt # Yesterday I knew nothing, > Applied Research Corporation # Today I know that. > merritt@iris613.gsfc.nasa.gov # thant   Received: from vmb.brl.mil by VMB.BRL.MIL id aa17706; 21 May 90 21:50 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa17526; 21 May 90 21:39 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa17515; 21 May 90 21:31 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa02374; 21 May 90 21:21 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA18230; Mon, 21 May 90 18:11:19 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 21 May 90 21:33:56 GMT From: "Louis M. McDonald" Organization: The Aerospace Corporation, El Segundo, CA Subject: sigvec function Message-Id: <73659@aerospace.AERO.ORG> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL I am trying to rehost a program from a Sun system that makes use of the "sigvec" function. This is currently causing me great pains to figure out how to get around it. "sigvec" is both a structure and a routine. If anyone has been successful in move a "sigvec" program to the Iris, let me know. --- from man page on BSD system --- SIGVEC(2) UNIX Programmer's Manual SIGVEC(2) NAME sigvec - software signal facilities SYNOPSIS #include struct sigvec { int (*sv_handler)(); int sv_mask; int sv_flags; }; sigvec(sig, vec, ovec) int sig; struct sigvec *vec, *ovec; Printed 8/26/86 January 8, 1986 5 -- Louis McDonald Internet: louis@aerospace.aero.org The Aerospace Corporation 213-336-8914   Received: from vmb.brl.mil by VMB.BRL.MIL id aa18826; 22 May 90 0:14 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa18497; 22 May 90 0:03 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa18495; 21 May 90 23:55 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa03303; 21 May 90 23:35 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA25928; Mon, 21 May 90 20:22:36 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 22 May 90 02:40:27 GMT From: Jesse Rendleman Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Re: Help with terminfo Message-Id: <7964@odin.corp.sgi.com> References: <9355@pt.cs.cmu.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <9355@pt.cs.cmu.edu> gerry@cive.ri.cmu.edu (Gerry Roston) writes: >Can someone please tell me where I can find the program > infocmp. >I have searched the disk, but to no avail. Is it possibly >on the distribution tape and I didn't load it onto my disk? > >We have an Iris 4d. > >thanx >gerry The infocmp program (as well as some extra terminfo files, etc..) is on tape 1 (aka the "execution only environment, tape 1" tape), in the terminf subsystem. To load it, you can use the manual feature of inst, and select "eoe1.sw.terminf". See your release notes for complete instructions on how to do this. To save space, it is not loaded by default.   Received: from vmb.brl.mil by VMB.BRL.MIL id aa19591; 22 May 90 2:17 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa19457; 22 May 90 2:07 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa19412; 22 May 90 1:55 EDT Received: from cunyvm.cuny.edu by VGR.BRL.MIL id aa03829; 22 May 90 1:39 EDT Received: from DDATHD21.BITNET by CUNYVM.CUNY.EDU (IBM VM SMTP R1.2.2MX) with BSMTP id 4934; Tue, 22 May 90 01:38:08 EDT Received: from BR2.THD.DA.D.EUROPE by DDATHD21.BITNET via GNET with RJE ; 22 May 90 07:11:05 Date: Tue, 22 May 90 07:09:31 +0200 (Central European Summer Time) From: Knobi der Rechnerschrat Subject: Kernel Error Messages To: info-iris@BRL.MIL X-VMS-To: X%"info-iris@brl.mil" Message-ID: <9005220139.aa03829@VGR.BRL.MIL> Hi Netlanders, a friend of mine without network connections asked me about the following error message that happens toappears on the console of his 80GTB running 3.2.2: ERROR #31 un kernel: severity=2, max retrace events:20 I couln't help him, because I don't know this message and I also don't know where to look (ok, I/He could ask SGI, but I'm just curious !!!). So: can somebody explain this message and its possible causes, and can somebody tell me where to get a "Novice Users Guide through the Kernel Messages"? Martin Knoblauch PS: A "Gurus Guide ..." would also do   Received: from vmb.brl.mil by VMB.BRL.MIL id aa27288; 22 May 90 12:47 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa25530; 22 May 90 11:14 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa25396; 22 May 90 10:59 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa06000; 22 May 90 10:37 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA26574; Tue, 22 May 90 07:22:55 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 22 May 90 14:21:00 GMT From: Tim Hall Organization: Boston University Subject: more on lighting... Message-Id: <57630@bu.edu.bu.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL >You didn't mention what specific graphics you have, so I'll assume you >have a GT. 4D 240/GTX >You're right.. the normal needs to be specified after the color is set >but before the vertex is called. >Although this behavior may seem strange, consider the alternative: >to make your case work, we should evaluate lighting at the cpack call and >also at the normal call. I have no knowledge of the inner workings of the geometry engine but... I would expect that you would evaluate the lighting at the vertex call. The only time you could get away with evaluating the lighting at the normal call is when you have infinite light sources and a non-specular material. (Which does happen to be my case.) Otherwise the lighting is dependent on both the normal and the positionn of the vertex. -Tim Hall tjh@bu-pub.bu.edu   Received: from vmb.brl.mil by VMB.BRL.MIL id ab27288; 22 May 90 12:47 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ac26999; 22 May 90 12:36 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa26860; 22 May 90 12:21 EDT Received: from cunyvm.cuny.edu by VGR.BRL.MIL id aa06815; 22 May 90 12:12 EDT Received: from MCCLB0.MED.NYU.EDU by CUNYVM.CUNY.EDU (IBM VM SMTP R1.2.2MX) with BSMTP id 0677; Tue, 22 May 90 12:10:43 EDT Received: from mcirps2.med.nyu.edu by MCCLB0.MED.NYU.EDU; Tue, 22 May 90 12:11 EDT Received: by mcirps2.med.nyu.edu (5.52/890607.SGI) (for @mcclb0.med.nyu.edu:info-iris@brl.arpa) id AA28234; Tue, 22 May 90 12:17:25 DSD Date: Tue, 22 May 90 12:17:25 DSD From: karron%mcirps2.med.nyu.edu@cunyvm.cuny.edu Subject: Re: Help with terminfo To: cive.ri.cmu.edu!gerry@pt.cs.cmu.edu Cc: gerry@civt.ri.cmu.edu, info-iris@BRL.MIL Message-id: <9005221917.AA28234@mcirps2.med.nyu.edu> X-Envelope-to: info-iris@brl.arpa Gerry Roston writes about Help with terminfo >Can someone please tell me where I can find the program infocmp. >I have searched the disk, but to no avail. Is it possibly >on the distribution tape and I didn't load it onto my disk? The answer is YES. Check your distribution tape with the program "inst". You have to ask for some sections of the tape to be unloaded. This is because all of the man pages can fill up your disk. I use infocmp all too often, and I know it is on the iris, at least in v3.2. There is a terminfo package on the tape that may also include the man pages. -- +-----------------------------------------------------------------------------+ | karron@nyu.edu Dan Karron | | . . . . . . . . . . . . . . New York University Medical Center | | 560 First Avenue \ \ Pager <1> (212) 397 9330 | | New York, New York 10016 \**\ <2> 10896 <3> | | (212) 340 5210 \**\__________________________________________ | +-----------------------------------------------------------------------------+   Received: from vmb.brl.mil by VMB.BRL.MIL id aa27658; 22 May 90 13:08 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab26999; 22 May 90 12:36 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab26857; 22 May 90 12:21 EDT Received: from cunyvm.cuny.edu by VGR.BRL.MIL id aa06776; 22 May 90 12:06 EDT Received: from MCCLB0.MED.NYU.EDU by CUNYVM.CUNY.EDU (IBM VM SMTP R1.2.2MX) with BSMTP id 0591; Tue, 22 May 90 12:04:26 EDT Received: from mcirps2.med.nyu.edu by MCCLB0.MED.NYU.EDU; Tue, 22 May 90 12:05 EDT Received: by mcirps2.med.nyu.edu (5.52/890607.SGI) (for mcclb0.med.nyu.edu!uunet.uu.net!wuarchive!cec2!news) id AA28195; Tue, 22 May 90 12:11:04 DSD Date: Tue, 22 May 90 12:11:04 DSD From: karron%mcirps2.med.nyu.edu@cunyvm.cuny.edu Subject: Re: Help building X11R4 libs on Personal Iris To: wuarchive!cec2!news@uunet.uu.net Cc: info-iris@BRL.MIL, todd@wucs1 Message-id: <9005221911.AA28195@mcirps2.med.nyu.edu> X-Envelope-to: info-iris@brl.arpa, wuarchive!cec2!news@uunet.uu.net Where did you get your X11R4 distribution ? I thought sgi sold an X11 developers package, and if that is your case, they should fix it. I am looking to start an X11 development project, that is why I am interested. dan. -- +-----------------------------------------------------------------------------+ | karron@nyu.edu Dan Karron | | . . . . . . . . . . . . . . New York University Medical Center | | 560 First Avenue \ \ Pager <1> (212) 397 9330 | | New York, New York 10016 \**\ <2> 10896 <3> | | (212) 340 5210 \**\__________________________________________ | +-----------------------------------------------------------------------------+   Received: from vmb.brl.mil by VMB.BRL.MIL id aa28072; 22 May 90 13:25 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab27658; 22 May 90 13:14 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa27580; 22 May 90 13:03 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa07180; 22 May 90 12:51 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA04865; Tue, 22 May 90 09:37:20 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 22 May 90 16:24:24 GMT From: Gary Tarolli Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Re: lmcolor and normals Message-Id: <60746@sgi.sgi.com> References: <57555@bu.edu.bu.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <57555@bu.edu.bu.edu>, tjh@bu-pub.bu.edu (Tim Hall) writes: > In displaying a multi-colored planar mesh I use the routine > lmcolor( LMC_DIFFUSE ) inorder to use the routine "cpack" to > specify the vertex colors. The man page for lmcolor led me > to believe that after the call to cpack the material color in > the hardware should now be the color specified in cpack. However, > experience shows that you have to specify the vertex normal after > the cpack call before the color takes effect. Is this the correct > behavior? Why isn't the already specified normal used? (Since > my mesh is planar I had dreams of only specifing the normal once.) > > -Tim Hall > tjh@bu-pub.bu.edu According to the man page: NAME lmcolor - change the effect of color commands while lighting is active C SPECIFICATION void lmcolor(mode) long mode; PARAMETERS mode the name of the mode to be used. Possible modes are: LMC_COLOR, RGB color commands will set the current color. If a color is the last thing sent before a vertex the vertex will be colored. If a normal is the last thing sent before a vertex the vertex will be lighted. LMC_COLOR is the default mode. ... the vertex will be colored to the exact color specified with cpack, ie it wont be lit. This allows you to draw un-lit objects that are pseudo-colored even while lighting is on. To get the vertex lit, you have to specify a normal after the color, as you noted. This does make it necessary to re-specify the normal for a planar surface after each vertex. On the other hand, it does allow a mechanism for by-passing the lighting model... Gary Tarolli   Received: from vmb.brl.mil by VMB.BRL.MIL id aa28915; 22 May 90 13:53 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa28666; 22 May 90 13:42 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa28411; 22 May 90 13:30 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa07364; 22 May 90 13:23 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA06946; Tue, 22 May 90 10:11:42 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 22 May 90 17:10:53 GMT From: Tim Hall Organization: Boston University Subject: man page for lmcolor Message-Id: <57639@bu.edu.bu.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Me.. In article <57555@bu.edu.bu.edu>, tjh@bu-pub.bu.edu (Tim Hall) writes: >> In displaying a multi-colored planar mesh I use the routine >> lmcolor( LMC_DIFFUSE ) inorder to use the routine "cpack" to ^^^^^^^^^^^ >> specify the vertex colors. Gary Taroli replies >According to the man page: > >PARAMETERS > mode the name of the mode to be used. Possible modes are: > > LMC_COLOR, RGB color commands will set the current color. If a ^^^^^^^^^ > color is the last thing sent before a vertex the vertex will be > colored. If a normal is the last thing sent before a vertex the > vertex will be lighted. LMC_COLOR is the default mode. The part of the man page I paid most attention to is.... LMC_DIFFUSE, RGB color commands will set the DIFFUSE color ^^^^^^^^^^^ property of the current material. Alpha, the fourth color component specified by RGB color commands will set the ALPHA property of the current material. No mention of respecifing the normal.... -Tim Hall tjh@bu-pub.bu.edu   Received: from vmb.brl.mil by VMB.BRL.MIL id aa29232; 22 May 90 14:03 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa26999; 22 May 90 12:36 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa26857; 22 May 90 12:21 EDT Received: from SGI.COM by VGR.BRL.MIL id aa06774; 22 May 90 12:06 EDT Received: from oni.sgi.com by sgi.sgi.com (5.52/900423.SGI) for info-iris@brl.mil id AA11423; Tue, 22 May 90 09:05:42 PDT Received: from palladium.sgi.com by relay.sgi.com (5.52/891101.SGI) for @sgi.sgi.com:info-iris@BRL.MIL id AA05374; Tue, 22 May 90 09:05:39 PDT Received: from sgihub.corp.sgi.com by palladium.corp.sgi.com (5.52/891101.SGI) for @relay.sgi.com:info-iris@BRL.MIL id AA14285; Tue, 22 May 90 09:05:34 PDT Received: by sgihub.corp.sgi.com (5.52/891101.SGI) for @palladium.corp.sgi.com:info-iris@BRL.MIL id AA20929; Tue, 22 May 90 09:05:31 PDT Received: from atl280.atlanta.sgi.com by sgiatl.atlanta.sgi.com (5.52/891101.SGI) for sgihub!BRL.MIL!info-iris id AA09613; Tue, 22 May 90 11:33:56 EDT Received: by atl280.atlanta.sgi.com (5.52/891101.SGI) for info-iris%BRL.MIL@sgiatl.atlanta.sgi.com id AA04707; Tue, 22 May 90 11:33:45 EDT Date: Tue, 22 May 90 11:33:45 EDT From: George Smith SGI Atlanta Message-Id: <9005221533.AA04707@atl280.atlanta.sgi.com> To: info@atl280.atlanta.sgi.com Subject: Display lib. HI:: Does any one have a GL version of the old graphics library Display ? Thanks.. -- /-----------------------------\ /-------------------------------------------\ | George Smith | | | | Systems Eng. | | ph.(404)392-1333 | | | | Email: | | Silicon Graphics Inc. | | georges@.sgi.com | | 1100 Abernathy Road N.E. | | {decwrl,pyramid,ucbvax,gatech}!sgi!georges| | Building 500, Suite 1120 | | | | Atlanta,Ga. 30328 | | | \-----------------------------/ \-------------------------------------------/   Received: from vmb.brl.mil by VMB.BRL.MIL id aa01416; 22 May 90 15:47 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa00636; 22 May 90 15:12 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa00407; 22 May 90 14:56 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa07856; 22 May 90 14:37 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA12126; Tue, 22 May 90 11:34:50 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 22 May 90 17:58:53 GMT From: Jerre Bowen Organization: Silicon Graphics Inc. Subject: Re: sigvec function Message-Id: <7987@odin.corp.sgi.com> References: <73659@aerospace.AERO.ORG> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL > From: louis@aerospace.aero.org (Louis M. McDonald) > Subject: sigvec function > Message-ID: <73659@aerospace.AERO.ORG> > Date: 21 May 90 21:33:56 GMT > Organization: The Aerospace Corporation, El Segundo, CA > > I am trying to rehost a program from a Sun system that makes use of > the "sigvec" function. This is currently causing me great pains to > figure out how to get around it. "sigvec" is both a structure and > a routine. If anyone has been successful in move a "sigvec" program > to the Iris, let me know. > > > > -- > Louis McDonald Internet: louis@aerospace.aero.org > The Aerospace Corporation > 213-336-8914 > The BSD signal routines are more powerful than the SysV versions; therefore there is no way to truly simulate them in IRIX 3.2 (which has only SysV signals). The main problem is the inability to atomically block sets of signals during the execution of handler routines. The next release of IRIX will have the full set of BSD signal routines, plus (better yet) POSIX signals. Jerre Bowen (bowen@sgi.com)   Received: from vmb.brl.mil by VMB.BRL.MIL id aa01843; 22 May 90 16:05 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab01416; 22 May 90 15:54 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa01297; 22 May 90 15:37 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa07960; 22 May 90 15:07 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA13444; Tue, 22 May 90 11:56:07 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 22 May 90 18:48:23 GMT From: "Spencer W. Thomas" Organization: University of Michigan EECS Dept Subject: Re: Rotations about an arbitrary axis Message-Id: References: <9005171647.AA14106@baby.swmed.utexas.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <9005171647.AA14106@baby.swmed.utexas.edu> rose@BABY.SWMED.UTEXAS.EDU (Rose Oguz) writes: [Problem of rotation about an arbitrary axis.] > Basically, I have two vectors coming > from the same point and one of the vectors needs to be rotated into the > other. I'm cross-posting this to comp.graphics, as I am sure it will be of interest to readers there, too. My solution illustrates a method for directly constructing rotation matrices that is not usually emphasized in graphics texts. You can construct the required matrix directly by the following procedure: You want to rotate the vector V1 into the vector V2. We can do this in two steps: (1) rotate V1 to the X axis and then (2) rotate the X axis to V2. (Why do this, you say? Because each step is easy.) I will denote arbitrary vectors with capital letters, and the corresponding unit vectors with lowercase. I.e., v1 = V1 / |V1| 1. N = V1 x V2 (cross product) 2. m = n x v1 (v1, m, n are now orthogonal unit vectors) 3. Form a matrix R1 with v1, m, n as its columns: ( v1x, mx, nx ) R1 = ( v1y, my, ny ) ( v1z, mz, nz ) This matrix will rotate v1 to the positive x unit vector (1,0,0) and n to the positive z unit vector (0,0,1) (and m to the positive y unit vector (0,1,0)). 4. p = n x v2 (so v2, p, n are orthogonal unit vectors) 5. Form a matrix R2 with v2, p, n as its rows: ( v2x, v2y, v2z ) R2 = ( px, py, pz ) ( nx, ny, nz ) This rotation matrix takes the x vector to v2 and the z vector to n (and the y vector to p). 6. The desired matrix is then R1 R2: ( v1x v2x + mx px v1x v2y + mx py v1x v2z + mx pz ) ( + nx nx + nx ny + nx nz ) ( v1y v2x + my px v1y v2y + my py v1y v2z + my pz ) ( + ny nx + ny ny + ny nz ) ( v1z v2x + mz px v1z v2y + mz py v1z v2z + mz pz ) ( + nz nx + nz ny + nz nz ) This rotates V1 into V2 about N (and incidently carries m into p). As an example, consider the problem of rotating V1 = (1,2,2) into V2 =(2,1,2). Then v1 = (1/3,2/3,2/3), v2 = (2/3,1/3,2/3). N = (2,2,-3), n = (2,2,-3)/sqrt(17). m = n x v1 = (10,-7,2) / (3*sqrt(17)) p = n x v2 = (7,-10,-2) / (3*sqrt(17)) ( 1/3 10/3sqrt(17) 2/sqrt(17) ) R1 = ( 2/3 -7/3sqrt(17) 2/sqrt(17) ) ( 2/3 2/3sqrt(17) -3/sqrt(17) ) ( 2/3 1/3 2/3 ) R2 = ( 7/3sqrt(17) -10/3sqrt(17) -2/3sqrt(17) ) ( 2/sqrt(17) 2/sqrt(17) -3/sqrt(17) ) ( 140/153 -47/153 -40/153 ) R1*R2 = ( 55/153 140/153 28/153 ) ( 28/153 -40/153 145/153 ) -- =Spencer (spencer@eecs.umich.edu)   Received: from vmb.brl.mil by VMB.BRL.MIL id ac01843; 22 May 90 16:11 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ac01416; 22 May 90 15:55 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa01315; 22 May 90 15:39 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa08028; 22 May 90 15:24 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA14746; Tue, 22 May 90 12:17:13 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 22 May 90 18:53:03 GMT From: Thomas Russo Organization: University of Texas at Austin, Center for Nonlinear Dynamics Subject: Nawk script causes nawk to drop core Message-Id: <30324@ut-emx.UUCP> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL We're running a 4 Server 8 with Irix 3.2: So, any clues why /usr/bin/nawk should drop core on this script (lifted pretty much verbatim from the nawk manual): END { while ("who" | getline x) print x } when I run it: /usr/bin/nawk -f test.awk test.awk without ever producing any output? We had a simliar problem with gawk-2.11.1 too, and that had a bug where do_getline was calling free with a null pointer. Could nawk be doing something like that too? This script fails too: END { while ("who" | getline) print $0 } Thomas Russo Center for Nonlinear Dynamics, University of Texas at Austin russo@chaos.utexas.edu or phib421@utchpc.bitnet ------   Received: from vmb.brl.mil by VMB.BRL.MIL id aa02278; 22 May 90 16:21 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab01843; 22 May 90 16:10 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa01660; 22 May 90 15:54 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa08217; 22 May 90 15:51 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA16165; Tue, 22 May 90 12:40:49 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 22 May 90 19:19:24 GMT From: "Bruce R. Holloway" Organization: Silicon Graphics, Inc., Mountain View, CA Subject: more on quads Message-Id: <60767@sgi.sgi.com> References: <9005212256.AA01475@sonata.sgi.com> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <9005212256.AA01475@sonata.sgi.com>, rolf%sonata@SGI.COM (Rolf van Widenfelt) writes: > > In article <57560@bu.edu.bu.edu>, tjh@bu-pub.bu.edu (Tim Hall) writes: > > > > When rendering a quad mesh does the VGX break it up into triangles? > > For example, I use the quad mesh to render a square with a different > > color at each vertex. Is the pixel in the center of the square an > > average of all the colors at the four verticies or is it an average > > of two diagonal verticies? > > The underlying area filling primitive of PowerVision is the triangle, so > your quad becomes two triangles. > A quad may not be planar in color, and so we don't define what the > color in the center should be. > Actually, the center of some quads is not even inside the quad. > If the vertexes of your quad have radically different colors, further > subdivision is suggested to avoid mach banding artifacts. > Now, a triangle is planar, so the color in the center is roughly the > average of the vertex colors. > > I hope this helps a little... Since the quad gets split into two triangles, the color of the center pixel is the average of two diagonal vertices. When Rolf says the center might not be in the quad, he is referring to the possibility of a concave or bowtie quad (when projected on the screen). Regards, bruceh   Received: from vmb.brl.mil by VMB.BRL.MIL id aa04277; 22 May 90 18:05 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab04092; 22 May 90 17:54 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa04063; 22 May 90 17:40 EDT Received: from cunyvm.cuny.edu by VGR.BRL.MIL id aa08690; 22 May 90 17:29 EDT Received: from MCCLB0.MED.NYU.EDU by CUNYVM.CUNY.EDU (IBM VM SMTP R1.2.2MX) with BSMTP id 4295; Tue, 22 May 90 17:27:13 EDT Received: from mcirps2.med.nyu.edu by MCCLB0.MED.NYU.EDU; Tue, 22 May 90 15:06 EDT Received: by mcirps2.med.nyu.edu (5.52/890607.SGI) (for @mcclb0.med.nyu.edu:info-iris@brl.arpa) id AA00520; Tue, 22 May 90 15:12:20 DSD Date: Tue, 22 May 90 15:12:20 DSD From: karron%mcirps2.med.nyu.edu@cunyvm.cuny.edu Subject: Makefile madness... To: info-iris@BRL.MIL Message-id: <9005222212.AA00520@mcirps2.med.nyu.edu> X-Envelope-to: info-iris@brl.arpa The purpose of this note is to report on my resolution of makefile madness. My design objective is to make a makefile for a large, multisource file program that would not need a lot of hand editing of the makefile to keep track of the dependency. I want only to edit a list of .c source files in the make file and let it (make) figure out the rest. The old way was to add/delete dependency as I wrote the program. I found that some of the most difficult runtime bugs were ultimately caused by bad or outdated dependencies in the makefile. My first attempt to clean up this problem was to try to write a rule to get make to test a .c file as a dependent on a .h file, or to get a .o file to depend on a .c or a .h file. My own programming style is to put all .h and header stuff in a .h file for each .c file. This lead me on a through education in make, but not a clean solution. The real problem was that I did not ask the proper question. The solution is to get the actual dependency information out of the source files and make the explicit dependency rules from that. Then the make file would not have to infer that all .c files depend on .h , or all .o depends on .c or .h. I needed a program to spell it out. Various replies over the net came with ideas and code that I incorporated into my makefiles. The main idea is that cpp -M will make a list of dependencies, and there are various awk scripts to massage it into something you can make with. The best script is in /usr/sbin as mkdepend, and is part of the sgi distribution. I made some minor changes to my taste, but in effect mkdepend will go through your list of .c sources and make the complete dependencies rules and edit it into your makefile. I made some changes to sort the system and full path (/usr/include) files at the end of the list, and shorten the full path names for readability. I have not tested this with various -I include file search path options, and I have not had and experience with archives. The dependencies will change with the -I options and the -D options, so the cc (cpp) flags should be shown for a set of explicit dependencies. Question: How do it get the .mkdepend rule to work without an explicit "make depend". Can I make a rule to keep the dependencies current, rebuilding the makefile as I go ? Perhaps it would be better to include the depencies as a makefile include file ? Emacs would not know to update the makefile buffer as the makefile file changes under it in the emacs compile-it facility. What I would like to see: The .o dependent list should be sorted with the .c files first, the local .h files next, and the //.h files last, with an option to eliminate the /usr/include files, as they only will change in a operating system upgrade. If one has completed work on a program, and the only changes are likely to come from an operating system change, then one would want system include files in the makefile list. 1)Appended below is a makefile I am using, incorporating a mkdepend. 2)Below that is my version of mkdepend. dan. 1)---------Makefile below-------------------------------------------------- # THis is the ultimate makefile. It knows all, and does only what is needed. CSOURCES = Main.c 3space.c TableTop.c SpaceCursor.c DataFile.c Pups.c \ Math.c SandRstates.c Triad.c FileNames.c Modes.c PupDir.c DataDisplay.c \ Amoeba.c Align.c Numerical.c TARGET = Main OBJECTS = $(CSOURCES:.c=.o) LIBS = -lgl -lm .SUFFIXES : .SUFFIXES : .c .o DRECK = .B* .C* a.out core CFLAGS=-g2 -DDEBUG $(TARGET) : $(OBJECTS) @echo "$(TARGET) RULE $(@) : $(?)" $(CC) $(CFLAGS) -o $(TARGET) $(OBJECTS) $(LIBS) .c.o : @echo ".c.o RULE $(@) : triggered by $(?)" $(CC) $(CFLAGS) -c $(<) depend : $(CSOURCES) mkdepend -c "$(CC) $(CFLAGS)" Makefile $(?) clean : rm -f $(OBJECTS) $(TARGET) $(DRECK) # DO NOT DELETE THIS LINE -- START machine edit marker for mkdepend # Edited in by mkdepend : on Tue May 22 14:43:34 DSD 1990 # by /karron/bin//mkdepend : DBK version 4938 May 22 14:43 # CC SETTINGS are cc -g2 -DDEBUG -M IN = /usr/include Main.o: DatabaseInCore.h Globals.h Inits.h Main.c Main.h ModeDefines.h \ access.h getopts.h $(IN)/device.h $(IN)/gl.h $(IN)/stdio.h 3space.o: 3space.c 3space.h Globals.h ModeDefines.h $(IN)/ctype.h \ $(IN)/fcntl.h $(IN)/gl.h $(IN)/stdio.h $(IN)/string.h \ $(IN)/sys/fcntl.h $(IN)/sys/ioctl.h $(IN)/sys/termio.h TableTop.o: ColorRamp.h Globals.h ModeDefines.h TableTop.c TableTop.h \ $(IN)/gl.h $(IN)/stdio.h SpaceCursor.o: SpaceCursor.c SpaceCursor.h $(IN)/gl.h $(IN)/stdio.h DataFile.o: DataFile.c DataFile.h DatabaseInCore.h Globals.h Math.h \ ModeDefines.h access.h $(IN)/gl.h $(IN)/math.h \ $(IN)/sgimath.h $(IN)/stdio.h $(IN)/string.h $(IN)/sys/stat.h \ $(IN)/sys/types.h $(IN)/values.h Pups.o: DatabaseInCore.h Globals.h ModeDefines.h Pups.c Pups.h \ $(IN)/device.h $(IN)/get.h $(IN)/gl.h $(IN)/math.h \ $(IN)/sgimath.h $(IN)/stdio.h $(IN)/values.h Math.o: Math.c Math.h $(IN)/gl.h $(IN)/math.h $(IN)/sgimath.h \ $(IN)/stdio.h $(IN)/values.h SandRstates.o: DatabaseInCore.h Globals.h SandRstates.c SandRstates.h \ $(IN)/gl.h $(IN)/stdio.h Triad.o: Triad.c Triad.h $(IN)/gl.h $(IN)/stdio.h FileNames.o: FileNames.c FileNames.h $(IN)/stdio.h Modes.o: DatabaseInCore.h Globals.h Math.h ModeDefines.h Modes.c Modes.h \ $(IN)/gl.h $(IN)/math.h $(IN)/sgimath.h $(IN)/stdio.h \ $(IN)/values.h PupDir.o: Globals.h PupDir.c PupDir.h $(IN)/device.h $(IN)/dirent.h \ $(IN)/gl.h $(IN)/stdio.h $(IN)/sys/dirent.h $(IN)/sys/stat.h \ $(IN)/sys/types.h DataDisplay.o: DataDisplay.c DataDisplay.h DatabaseInCore.h Globals.h \ Math.h ModeDefines.h $(IN)/gl.h $(IN)/math.h $(IN)/sgimath.h \ $(IN)/stdio.h $(IN)/values.h Amoeba.o: Amoeba.c Amoeba.h $(IN)/math.h $(IN)/sgimath.h $(IN)/stdio.h \ $(IN)/values.h Align.o: Align.c Align.h DatabaseInCore.h Globals.h $(IN)/gl.h \ $(IN)/math.h $(IN)/sgimath.h $(IN)/stdio.h $(IN)/values.h Numerical.o: Numerical.c Numerical.h $(IN)/malloc.h $(IN)/math.h \ $(IN)/sgimath.h $(IN)/stdio.h $(IN)/values.h # DO NOT DELETE THIS 2nd LINE -- END machine edit for mkdepend SNIP-------END of Makefile------------------------------------------------SNIP Here is my version of mkdepend, with my changes annotated. 2)SNIP----mkdepend--------------------------------------------------------SNIP #! /bin/sh # NAME # # mkdepend - compute header file dependencies # # SYNOPSIS # # mkdepend [-c compilehow] [-e sedprog] [-f force] [-i] depfile file ... # # DESCRIPTION # # Mkdepend infers make dependencies from source containing C #include # directives. Given a shell command compilehow which consists of cc # followed by options, mkdepend processes its file arguments and edits # the generated dependency information into depfile, which may be a # makefile or a make include file. # # The -e flag passes an immediate program to sed, which is applied to # raw dependency information of the following form: # # target: dependent # # Thus one may substitute pathname prefixes with envariable parameters, # for example. # # The -f flag causes mkdepend to add a dependent named force to each # target file's dependency list. Using -f '$(FRC)' and setting FRC=FRC # in a make's environment, one may rebuild certain objects without first # removing them. # # Normally, old dependencies are deleted from the depfile. The -i # option causes mkdepend to preserve old dependencies. When invoked # from a makefile, the following rule enables incremental updates to # the ".mkdepend" dependency database: # # .mkdepend: $(CFILES) # mkdepend -c "$(CC) $(CFLAGS)" -i $@ $? # version="$0 : DBK version `ls -l $0| tr -s ' ' |cut -d' ' -f5-8`" echo $version depgen="cc -M" incr=no # # Process options and arguments. We put quotes around the edit (-e) arguments # and use eval "sed -e ..." later, to preserve spaces in the edit command. # while test $# -gt 0; do case $1 in -c) depgen="$2 -M" shift;; -e) sedprog="$sedprog -e '$2'" shift;; -f) force="$force $2" shift;; -i) incr=yes;; *) break esac shift done case $# in 0) echo \ "usage: $0 [-c compilehow] [-e sedprog] [-f force] [-i] depfile [file ...]" ;; 1) echo "$0 : no source file arguments, so nothing to do" exit 0 esac newmakefile="$1" oldmakefile=".B$$""$1" shift # # An awk script to compress dependencies. # awkprog=' BEGIN { FS = ":" INDENT = "\t" INDENTSIZE = 8 MAXLINE = 72 MAXINDENTEDLINE = MAXLINE - INDENTSIZE - 1 } # Folds up dependency lines from cpp -M into # "target1 : dep1, ... , depN" # "target2 : dep1, ... , depN" # For each line of the form # "target1: dependent1, ... ,targetN: dependentM", where the # spaces are literal blanks, do the following: { if ($1 != target) { if (depline) print depline target = $1 depline= $0 "'"$force"'" lim = MAXLINE } else { if (length(depline) + length($2) > lim) { print depline " \\" depline = INDENT lim = MAXINDENTEDLINE } depline = depline $2 } } END { print depline }' # # Save the old file in case of problems. Use a temporary file for incremental # make depend. Clean up if interrupted. # deptmp=/tmp/incdep.$$ if test $incr = yes; then trapcmds="rm -f $deptmp;" fi if test -f $newmakefile; then trapcmds="$trapcmds mv -f \\$oldmakefile \\$newmakefile;" mv -f $newmakefile $oldmakefile fi trap "$trapcmds exit 0" 1 2 15 if test ! -f $oldmakefile; then echo "Neither $oldmakefile nor $newmakefile exists." exit 1 fi # # Remove any old dependencies from the makefile. # firstline='# DO NOT DELETE THIS LINE -- START machine edit marker for mkdepend' lastline='# DO NOT DELETE THIS 2nd LINE -- END machine edit for mkdepend' sed -e "/^$firstline/,/^$lastline/d" $oldmakefile > $newmakefile # # Delimit the beginning of the dependencies. Save old dependences in # incremental mode. # shorten long full path names starting with incl. (DBK) in="/usr/include" IN='IN' echo $firstline >> $newmakefile echo "# Edited in by `basename $0` : on `date` by $version" >> $newmakefile echo "# CC SETTINGS are ${depgen}" >> $newmakefile echo "${IN} = ${in}" >> $newmakefile IN='$('"${IN}"')' if test $incr = yes; then # # XXX Workaround for this sed bug: 1,/pat/d deletes 1,$ if line 1 # XXX and only line 1 matches pat. # (echo '# food for sed bug to eat'; cat $oldmakefile) | sed -e "1,/^$firstline/d" -e "/^$lastline/"',$d' >> $deptmp awkinc='/^[^ ].*$/ { deleting = 0; }' for f in $* do basename=`expr ./$f : '.*/\(.*\)\..*'` awkinc="$awkinc /^$basename\./ { deleting = 1; }" done awkinc="$awkinc { if (deleting == 0) print \$0; }" awk "$awkinc" $deptmp >> $newmakefile rm -f $deptmp fi # # Process source files - the sed commands and their functions are: # /^[^ ]*\..*:$/d delete SVR0 cc multi-file blather # from stdout # s:[^\./][^\./]*/\.\./::g rewrite "dir/../path" to be "path" # s: \./: : rewrite " ./path" to be " path" # Sort local files first in the depend list (DBK) ( for i in $* do $depgen $i | eval "sed -e '/^[^ ]*\..*:$/d' \ -e :loop -e 's:[^\./][^\./]*/\.\./::g' -e tloop \ -e 's: \./: :' \ $sedprog" | sort -ud | # sort .c first, .h second, then the full path stuff.... later (DBK) sed -e "s:${in}:${IN}:" done ) | awk "$awkprog" >> $newmakefile echo $lastline >> $newmakefile rm -f $oldmakefile # the return value should never be left undefined in a program run by make. exit 0 ------------END OF mkdepend-------------------------------------------------- -- +-----------------------------------------------------------------------------+ | karron@nyu.edu Dan Karron | | . . . . . . . . . . . . . . New York University Medical Center | | 560 First Avenue \ \ Pager <1> (212) 397 9330 | | New York, New York 10016 \**\ <2> 10896 <3> | | (212) 340 5210 \**\__________________________________________ | +-----------------------------------------------------------------------------+   Received: from vmb.brl.mil by VMB.BRL.MIL id aa04839; 22 May 90 19:58 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa04774; 22 May 90 19:47 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa04748; 22 May 90 19:27 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa09155; 22 May 90 19:22 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA29549; Tue, 22 May 90 16:10:13 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 22 May 90 20:45:22 GMT From: Harry Sharp Subject: C++ on Personal Iris Message-Id: <3260@usceast.UUCP> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Again, I come to the net for knowledge.... We have three Personal Irises here at USC on which we would like to run C++ on, but so far have had no luck in porting the software. In building the munch version of the translator, I get an error complaining about an undefined function _semgetc. I cannot seem to find it in any of the libraries on the system, or on any other system for that matter. Has anyone ported the AT&T C++ system to the Irises? If so, may I have a copy of the notes? On a related note, has anyone used the SGI version of the C++ system? Comments would be appreciated. Harry Sharp sharp@usceast.cs.scarolina.edu -- [=============================================================================] USMail : Dept of Mathematics, USC, Columbia, SC 29208 E-mail : gatech!hubcap!usceast!sharp sharp@usceast.cs.scarolina.edu GO COCKS! You can't lick our cocks! Univ. of South Carolina   Received: from vmb.brl.mil by VMB.BRL.MIL id aa04985; 22 May 90 20:19 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab04721; 22 May 90 19:30 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab04710; 22 May 90 19:16 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa09125; 22 May 90 19:10 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA28759; Tue, 22 May 90 15:56:49 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 22 May 90 22:25:30 GMT From: Mark Callow Organization: Silicon Graphics Inc., Entry Systems Division Subject: Re: named and NeWS Message-Id: <8022@odin.corp.sgi.com> References: <9005180312.AA00594@poly1.nist.gov> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <9005180312.AA00594@poly1.nist.gov>, rbriber@POLY1.NIST.GOV writes: |> |> "getsocketpeername: cant find name for 129.6.154.1 ..." in SYSLOG from |> init.ps. I don't know specifically what is wrong. Maybe the following info. can help you track down your problem. getsocketpeername does a getpeername (which is presumably returning 129.6.154.1) followed by a gethostbyaddr. gethostbyaddr looks up the host through named if directed to by the file /etc/resolv.conf otherwise it uses /etc/hosts. I believe named maintains 2 separate databases one indexed by name and the other by address. |> |> I've also made a few other observations. For example, after I login I can get |> the chests to come up by starting them from the command line in the CONSOLE |> WINDOW ONLY. If I try to start them from a wsh window I get nothing or |> "Connection reset by peer". I should also note any other PostScript processes |> (such as/usr/demos/bin/eye.ps or anything in /usr/NeWS/demo) also fail and |> can only be started in the console window. This is currently not a |> satisfactory solution. This is happening because the console window is not a child of the NeWS server. Consequently it doesn't have a NEWSSERVER environment variable so clients fall back on connecting to "localhost". Every process launched by the NeWS server has a NEWSSERVER environment variable which gives the internet address of your machine as returned by a gethostbyname(hostname) call made when the server first starts. Clients started by these processes consequently to connect to whatever that port number is and run into the database problem you are having. |> I use nslookup I can find the address for any hosts I |> query about but the default server announced when the program starts is: |> |> Default Server: localhost.nist.gov |> Address: 127.0.0.1 |> |> and not my machine's name (poly1.nist.gov, 129.6.154.1). If I then ask for the |> address associatied with my name (poly1) I get the non-autoritative answer: |> 129.16.154.1, which is OK. BUT if I do |> |> 'nslookup -q=ptr 129.16.154.1' to ask for the name associated with my address |> I get the error: |> |> ** No domain name pointer information is available for 1.154.6.129.in-addr.arpa I'm no named expert but it looks to me as if your /etc/resolv.conf file is incorrect and is making you look up things on your own machine. |> |> I've had a number of suggestions which involved setting NetSecurityWanted in |> init.ps to false. Is this dangerous? The server will call getsocketpeername regardless of the setting of NetSecurityWanted so it won't help. The danger level depends upon your environment. -- From the TARDIS of Mark Callow msc@ramoth.sgi.com, ...{ames,decwrl}!sgi!msc "There is much virtue in a window. It is to a human being as a frame is to a painting, as a proscenium to a play. It strongly defines its content."   Received: from vmb.brl.mil by VMB.BRL.MIL id aa05041; 22 May 90 20:29 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa04721; 22 May 90 19:29 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa04710; 22 May 90 19:16 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa09123; 22 May 90 19:09 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA29164; Tue, 22 May 90 16:03:56 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 22 May 90 22:33:58 GMT From: Mark Callow Organization: Silicon Graphics Inc., Entry Systems Division Subject: Re: DWB, LaserWriter support and the screen Message-Id: <8023@odin.corp.sgi.com> References: <9005182126.AA18797@aero4.larc.nasa.gov> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <9005182126.AA18797@aero4.larc.nasa.gov>, blbates@AERO4.LARC.NASA.GOV ("Brent L. Bates AAD/TAB MS361 x42854") writes: |> |> You don't. (Preview PostScript files on the screen) You can use psview(1). In release 3.2 due to bugs and NeWS 1.1's incomplete PostScript implementation you can preview maybe 60% of the PostScript files you are likely to run into. That number includes all files produced by the dwb tools and Transcript[TM] (unless you are including some unusual PostScript files directly into your troff documents). In release 3.3 you can preview probably 90% of the PostScript documents you might run into. The NeWS 1.1 PostScript implementation is still incomplete particularly in the areas of charpath and charstrings. [TM] Transcript is, I'm sure, a trademark of Adobe Systems, Inc. -- From the TARDIS of Mark Callow msc@ramoth.sgi.com, ...{ames,decwrl}!sgi!msc "There is much virtue in a window. It is to a human being as a frame is to a painting, as a proscenium to a play. It strongly defines its content."   Received: from vmb.brl.mil by VMB.BRL.MIL id aa05268; 22 May 90 21:16 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa05196; 22 May 90 21:06 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa05184; 22 May 90 21:00 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa09399; 22 May 90 20:52 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA05028; Tue, 22 May 90 17:40:30 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 22 May 90 23:44:01 GMT From: Wen-ching Chow Organization: Silicon Graphics, Inc. Subject: HP Laserjet III Message-Id: <8034@odin.corp.sgi.com> References: <7987@odin.corp.sgi.com> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL -------- Hi, Everyone HP recently announced the new Laserjet III with Postscript support. I am wondering if anyone has ever tried to hook one up to an IRIS through either the serial line or the parallel connection. And I am also interested in knowing if the SGI DWB package would work with the HP LJIII. Thanks! Wen   Received: from vmb.brl.mil by VMB.BRL.MIL id aa05532; 22 May 90 21:50 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab05381; 22 May 90 21:39 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa05377; 22 May 90 21:31 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa09555; 22 May 90 21:22 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA06972; Tue, 22 May 90 18:14:51 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 23 May 90 01:13:46 GMT From: yan xiao Organization: Univ.of Toronto, Dept. of Indus. Eng. Subject: 3120's paretab.h Message-Id: <1990May23.011346.16088@gpu.utcs.utoronto.ca> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL We accidently lose a file in /usr/people/gifts/mextools/portlib called: paretab.h (It is a 3120) Does anybody out there still have this "gifted" files Thanks Keywords:   Received: from vmb.brl.mil by VMB.BRL.MIL id aa06149; 22 May 90 23:38 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa06079; 22 May 90 23:27 EDT Received: from smoke.brl.mil by VMB.BRL.MIL id aa06068; 22 May 90 23:21 EDT Received: from SMOKE.BRL.MIL by SMOKE.smoke.brl.mil.MIL id aa11410; 22 May 90 23:11 EDT Received: by gould (5.52/(IC 19May87-13:37)) id AA03329; Tue, 22 May 90 20:40:52 EDT Received: from swbatl.UUCP by uunet.uu.net (5.61/1.14) with UUCP id AA22378; Tue, 22 May 90 03:20:12 -0400 Received: by swbatl.sbc.com (/\=-/\ Smail3.1.16.1 #16.9) id ; Tue, 22 May 90 02:21 CDT Received: by texbell.sbc.com (Smail3.1.18) id ; Tue, 22 May 90 02:14 CDT Received: by redsim (5.52/121989) id AA13586; Tue, 22 May 90 02:02:36 CDT Date: Tue, 22 May 90 02:02:36 CDT From: "Ronald B. Adams II" Message-Id: <9005220702.AA13586@redsim> To: info-iris@BRL.MIL Subject: InterViews Has anybody brought up InterViews on a 4-D? Does anybody like it (X not withstanding)? We've got InterViews rev. 2.5 (and C++, of course). Is there an "sgi.cf" file available? Thanks. Ron Adams Hughes Simulation Systems Inc. -- (formerly Rediffusion Simulation Inc.) redsim!adams (uucp) "If it was real, we wouldn't adams@redsim.lonestar.org (internet) call it simulation"   Received: from vmb.brl.mil by VMB.BRL.MIL id aa07171; 23 May 90 1:29 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ac06648; 23 May 90 0:27 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa06618; 23 May 90 0:16 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa10314; 23 May 90 0:07 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA15802; Tue, 22 May 90 20:51:35 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 23 May 90 03:30:42 GMT From: Andrew Myers Organization: Silicon Graphics, Inc., Mountain View, CA Subject: SGI version of the "ultimate makefile" Message-Id: <8044@odin.corp.sgi.com> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Here's how we'd do this makefile here at SGI. I think you'll find it's somewhat shorter and easier to read. Note that "commondefs" defines most of the stuff you want in a standard way. It also causes the dependencies to be read in automatically. You can override definitions in this file with a "localdefs" file. I don't know why you want to customize mkdepend; I'd advise against it. Your dependencies will be put in a file called "Makedepend". # ------------- SGI version of the ultimate makefile -------------- include $(ROOT)/usr/include/make/commondefs TARGETS = Main CFILES = Main.c 3space.c TableTop.c SpaceCursor.c DataFile.c Pups.c \ Math.c SandRstates.c Triad.c FileNames.c Modes.c PupDir.c \ DataDisplay.c Amoeba.c Align.c Numerical.c LCOPTS = -DDEBUG # "local C options" LLDLIBS = -lgl -lm # local libraries (do you want -lgl_s instead?) OPTIMIZER = -g # -g2 is default for -g, I believe default: incdepend $(TARGETS) # incdepend=incremental dependency computation include $(COMMONRULES) # clean and other rules supplied by COMMONRULES $(TARGETS): $(OBJECTS) $(CCF) $(OBJECTS) $(LDFLAGS) -o $@ # -------------- end of file ------------------- Andrew   Received: from vmb.brl.mil by VMB.BRL.MIL id aa08159; 23 May 90 4:10 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa08094; 23 May 90 3:59 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa08060; 23 May 90 3:47 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa11327; 23 May 90 3:38 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA27727; Wed, 23 May 90 00:32:27 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 22 May 90 22:42:50 GMT From: Tom Weinstein Organization: Silicon Graphics Inc. Subject: Re: InterViews Message-Id: References: <9005220702.AA13586@redsim> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <9005220702.AA13586@redsim>, uunet!swbatl!texbell!redsim!adams@gould.UUCP ("Ronald B. Adams II") writes: > Has anybody brought up InterViews on a 4-D? Does anybody like > it (X not withstanding)? > We've got InterViews rev. 2.5 (and C++, of course). Is there an "sgi.cf" > file available? I got the latest InterViews off of interviews.stanford.edu a few weeks ago, and it has support an sgi.cf. The changes I had to make amounted to no more than commenting out some of the prototypes in some of the C++ .h files that come with InterViews. > Ron Adams > Hughes Simulation Systems Inc. -- (formerly Rediffusion Simulation Inc.) > redsim!adams (uucp) "If it was real, we wouldn't > adams@redsim.lonestar.org (internet) call it simulation" -- Tom Weinstein Silicon Graphics, Inc., Entry Systems Division, Window Systems tomw@orac.esd.sgi.com Any opinions expressed above are mine, not sgi's.   Received: from vmb.brl.mil by VMB.BRL.MIL id aa05381; 22 May 90 21:33 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab05268; 22 May 90 21:22 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa05245; 22 May 90 21:11 EDT Received: from grace.waikato.ac.nz by VGR.BRL.MIL id aa09442; 22 May 90 21:08 EDT Received: from aukuni.ac.nz by waikato.ac.nz; Wed, 23 May 90 13:07 +1200 Received: from ccu1.aukuni.ac.nz by aukuni.ac.nz; Wed, 23 May 90 13:07 Y Received: by ccu1.aukuni.ac.nz (5.52/890607.SGI) (for @ccvcom.aukuni.ac.nz:info-iris@brl.mil) id AA27339; Wed, 23 May 90 13:04:20 NZT Date: Wed, 23 May 90 13:04:20 NZT From: russell@ccu1.aukuni.ac.nz Subject: exabyt tape problems To: info-iris@BRL.MIL Message-Id: <9005230104.AA27339@ccu1.aukuni.ac.nz> Greetings, I am relatively new to UNIX but have extensive experience on other systems. We now operate a 4D/240s as our main campus research machine. My current problem involves exabyte tapes. What I want to do is write ansi labels on the tapes that we use for backup so that the system can verify that the correct tapes have ben mounted. I have written a small C program ( part of which is appended to this message) that reads, writes or verifies Ansi labels. I works fine on the 1/2" tapedrive (xmt0d0...) but always fails on the exabyte drives. label: Error writing dev: '/dev/mt/tps0d6' label: Invalid argument I have hunted through the man pages and seen reference to the distinction between fixed and variable blocked devices. If I try the variable device tps0d6v it then appears to hang. I suspect it is spacing down the tape because when I ^C it it takes 20 or 30 sec to terminate. We have also tried to put more than one tar file on to the tape but have so far failed. Again this works as expected with the 1/2" tape. Any help and/or explainations will be greatly appreciated. -------------- extract from label.c ------------------------------------ #include #include #include #include #include #include /* sundry declarations deleted */ int tapectl(tape, cmd) /*==================*/ int tape, cmd ; { struct mtop op ; op.mt_op = cmd ; op.mt_count = 1 ; return ioctl(tape, MTIOCTOP, &op ) >= 0 ? 0 : errno ; } int main(argc, argv) int argc; char *argv[]; { char *p; int i, tape, errors = 0, lread = 0, lwrite = 0, lverify = 0; tapedev[0] = '\0'; /* argument parsing deleted */ if ( lread || lverify) { /* read and verify code deleted */ }else { if ((tape = open(tapedev, O_WRONLY)) < 0) { perror("tape open") ; exit(1) ; } tapectl(tape, MTREW) ; memset(&vol, ' ', sizeof vol); memcpy(vol.lid, "VOL1", 4); memcpy(vol.id, volid, sizeof vol.id); memcpy(vol.owner_id, owner, sizeof vol.owner_id); vol.standard = '3'; /* ansi x3.27-1978 */ if( write(tape, (char *)&vol, sizeof vol) != sizeof vol ) { fprintf(stderr,"label: Error writing dev: '%s'\n", tapedev); perror("label"); tapectl(tape, MTREW); exit(1); } if( tapectl(tape, MTWEOF) != 0 || tapectl(tape, MTWEOF ) != 0 ) { perror("Error writing tape mark"); exit(1); } tapectl(tape, MTBSF); /* position between tape marks */ } }/* label */ -------------------- end of label.c ------------------------------ -- +-----------------------------------------------------------------+ | Russell Fulton 'phone +64 9 737-999 x 8955 | | Computer Centre domain rj_fulton@aukuni.ac.nz | | University of Auckland fax +64 9 303-2467 | | Private Bag time gmt -12 (13) (summer time)| | Auckland, New Zealand. | +-----------------------------------------------------------------+   Received: from vmb.brl.mil by VMB.BRL.MIL id ab06817; 23 May 90 0:37 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa06648; 23 May 90 0:26 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa06588; 23 May 90 0:15 EDT Received: from cunyvm.cuny.edu by VGR.BRL.MIL id aa10304; 23 May 90 0:05 EDT Received: from MCCLB0.MED.NYU.EDU by CUNYVM.CUNY.EDU (IBM VM SMTP R1.2.2MX) with BSMTP id 6752; Wed, 23 May 90 00:02:50 EDT Received: from mcirps2.med.nyu.edu by MCCLB0.MED.NYU.EDU; Wed, 23 May 90 00:04 EDT Received: by mcirps2.med.nyu.edu (5.52/890607.SGI) (for @mcclb0.med.nyu.edu:info-iris@brl.arpa) id AA02757; Wed, 23 May 90 00:10:59 DSD Date: Wed, 23 May 90 00:10:59 DSD From: karron%mcirps2.med.nyu.edu@cunyvm.cuny.edu Subject: Re: Makefile madness... To: andru%rhialto.esd@sgi.com Cc: info-iris@BRL.MIL Message-id: <9005230710.AA02757@mcirps2.med.nyu.edu> X-Envelope-to: info-iris@brl.arpa, andru%rhialto.esd@sgi.com Thanks for your more ultimate makefile. I need to look more at /usr/include/make for a better feel for how you guys do things, and convince myself that that is a better way. My only observations about mkdefs was that I would like the depend rules to be sorted the way I would list them, e.g, subr.o : subr.c include1.h include2.h include3.h /usr/include/stdio.h I would like the .c first, the local .h second, then the system files last, and optional. They would not change in the few weeks that a program is germinating. mkdefs should also exit with a defined value in a makefile. (especially in a make file). If I understand correctly, mkdefs will incrimentally update depends by rules given in /usr/include/make/common... I have found that errors in makefiles have cost me many hours tracking down obscure runtime bugs. I am all for a uniform and absolutely ultimate method to write makefiles that would eliminate any margin for error. Our practice here is chaotic and erratic. I am glad that there is a proper way. dan. -- +-----------------------------------------------------------------------------+ | karron@nyu.edu Dan Karron | | . . . . . . . . . . . . . . New York University Medical Center | | 560 First Avenue \ \ Pager <1> (212) 397 9330 | | New York, New York 10016 \**\ <2> 10896 <3> | | (212) 340 5210 \**\__________________________________________ | +-----------------------------------------------------------------------------+   Received: from vmb.brl.mil by VMB.BRL.MIL id aa13603; 23 May 90 9:43 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa12842; 23 May 90 9:22 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa12597; 23 May 90 9:06 EDT Received: from aero4.larc.nasa.gov by VGR.BRL.MIL id aa15022; 23 May 90 9:00 EDT Received: Wed, 23 May 90 09:00:53 EDT by aero4.larc.nasa.gov (5.52/5.6) Date: Wed, 23 May 90 09:00:53 EDT From: "Brent L. Bates AAD/TAB MS361 x42854" Message-Id: <9005231300.AA03545@aero4.larc.nasa.gov> To: cs.utexas.edu!news-server.csri.toronto.edu!utgpu!xiao@tut.cis.ohio-state.edu Subject: Re: 3120's paretab.h Cc: info-iris@BRL.MIL Are you sure of the files location and existence? We have a 3130 and I couldn't find the stated file: /usr/people/gifts/mextools/portlib/paretab.h I did find a file parsetab.h in /usr/people/gifts/mextools/include, could that be the file you are looking for? -- Brent L. Bates NASA-Langley Research Center M.S. 361 Hampton, Virginia 23665-5225 (804) 864-2854 E-mail: blbates@aero4.larc.nasa.gov or blbates@aero2.larc.nasa.gov   Received: from vmb.brl.mil by VMB.BRL.MIL id aa15524; 23 May 90 10:23 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa11340; 23 May 90 8:29 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa11223; 23 May 90 8:11 EDT Received: from aero4.larc.nasa.gov by VGR.BRL.MIL id aa13757; 23 May 90 7:52 EDT Received: Wed, 23 May 90 07:53:36 EDT by aero4.larc.nasa.gov (5.52/5.6) Date: Wed, 23 May 90 07:53:36 EDT From: "Brent L. Bates AAD/TAB MS361 x42854" Message-Id: <9005231153.AA03356@aero4.larc.nasa.gov> To: sgi!shinobu!odin!ramoth.esd.sgi.com!msc@ucbvax.berkeley.edu Subject: Re: DWB, LaserWriter support and the screen Cc: info-iris@BRL.MIL psview works on LESS than 1% of the files I have tried it on, no where close to 60%. psview only worked on EXTREMELY simple files, I wrote manual. None of the PostScript files generated by any of the software we have worked in psview, yet ALL worked when sent to laser printers. I hope psview in release 3.3 is an improvement, but, when you are at the bottom the only place to go is up. SGI isn't the only ones with a psview that doesn't work, Sun's is equally bad. -- Brent L. Bates NASA-Langley Research Center M.S. 361 Hampton, Virginia 23665-5225 (804) 864-2854 E-mail: blbates@aero4.larc.nasa.gov or blbates@aero2.larc.nasa.gov   Received: from vmb.brl.mil by VMB.BRL.MIL id aa00092; 23 May 90 11:04 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab13603; 23 May 90 9:51 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa13437; 23 May 90 9:35 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa15213; 23 May 90 9:22 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA15108; Wed, 23 May 90 06:20:31 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 23 May 90 13:18:03 GMT From: Scott Kahn Organization: University of Illinois at Urbana Subject: Font manager information Message-Id: <1990May23.131803.749@ux1.cso.uiuc.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL How does one access font characters having glyth values >255 (eg in "TimesRoman")? Is there a two character sequence that is used, and where in the manuals is it mentioned? Scott Kahn (kahn@uxh.cso.uiuc.edu)   Received: from vmb.brl.mil by VMB.BRL.MIL id aa04838; 23 May 90 14:28 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa02773; 23 May 90 13:12 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa02688; 23 May 90 12:55 EDT Received: from PIG.DREA.DND.CA by VGR.BRL.MIL id aa16914; 23 May 90 12:43 EDT Received: Wed, 23 May 90 13:45:02 ADT by pig.drea.dnd.ca (5.52/5.6) Date: Wed, 23 May 90 13:45:02 ADT From: Jim Diamond Message-Id: <9005231645.AA16424@pig.drea.dnd.ca> To: info-iris@BRL.MIL, unix-emacs@vm.tcs.tulane.edu Subject: quit signal, gnu emacs, Iris workstations Hi: I have gnu emacs 18.55 running on both an Iris 3130 workstation and an Iris 4D/50 workstation. TeX-mode works fine on the 3130, but on the 4D system when one attempts to use TeX-kill-job (which is called on every invocation of TeX-buffer (or TeX-region) except the first) rather than happily returning, the error "Quit" appears. This apparently terminates the execution of TeX-region, negating the usefulness of this function. Has anyone else (a) noticed this problem, or (b) fixed it? Obviously, I could rewrite TeX-mode.el to avoid calling TeX-kill-job, but this is working around the problem, rather than addressing it. Thanks. Jim Diamond zsd@pig.drea.dnd.ca   Received: from vmb.brl.mil by VMB.BRL.MIL id ab04838; 23 May 90 14:28 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa03129; 23 May 90 13:35 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa03050; 23 May 90 13:15 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa17066; 23 May 90 13:08 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA28204; Wed, 23 May 90 10:01:52 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 23 May 90 16:58:38 GMT From: Vernon Schryver Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Re: Makefile madness... Message-Id: <60829@sgi.sgi.com> References: <9005230710.AA02757@mcirps2.med.nyu.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <9005230710.AA02757@mcirps2.med.nyu.edu>, karron@MCIRPS2.MED.NYU.EDU writes: > My only observations about mkdefs was that I would like the depend rules > to be sorted the way I would list them, e.g, > > subr.o : subr.c include1.h include2.h include3.h /usr/include/stdio.h > ... We don't do much to sort dependencies here. They tend to be too big and change too often. As long as the right things happen, one need not examine the dependencies, any more than one needs to review the u-code generated by the compilers. Our files containing dependencies are now considered objects rather than sources. (See, even I've reformed!) Large projects might benefit from the "incdepend" target in /usr/include/make/common*. It is intended to allow incremental updating of dependencies. In other words, it re-runs `ccp -M` only on those source files that have changed. Some people here use incdepend instead of the hour or so needed to regenerate all dependencies for the IRIX kernel. Vernon Schryver Silicon Graphics vjs@sgi.com   Received: from vmb.brl.mil by VMB.BRL.MIL id aa09312; 23 May 90 17:35 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab09011; 23 May 90 17:24 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa08840; 23 May 90 17:07 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa18543; 23 May 90 16:38 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA11567; Wed, 23 May 90 13:26:14 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 23 May 90 19:35:48 GMT From: sgi!shinobu!odin!inferno.esd.sgi.com!rmr@ucbvax.berkeley.edu Subject: Re: DWB, LaserWriter support and the screen Message-Id: <8077@odin.corp.sgi.com> References: <9005231153.AA03356@aero4.larc.nasa.gov> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL >None of the PostScript files generated by any of the >software we have worked in psview... I don't know what tools you are using, but I *can* vouch for the fact that psview works fine for troff, tbl, and eqn documents that don't import images or other non-DWB format items. If you haven't found this to be the case, I'd like to know. Robert Reimann Entry Systems Developer Documentation rmr@sgi.com   Received: from vmb.brl.mil by VMB.BRL.MIL id aa10333; 23 May 90 19:39 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa10167; 23 May 90 19:03 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa10070; 23 May 90 18:48 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa19214; 23 May 90 18:40 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA19305; Wed, 23 May 90 15:25:17 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 23 May 90 22:02:13 GMT From: Rob Warnock Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Re: Nawk script causes nawk to drop core Message-Id: <60865@sgi.sgi.com> References: <30324@ut-emx.UUCP> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <30324@ut-emx.UUCP> russo@chaos.utexas.edu (Thomas Russo) writes: +--------------- | We're running a 4 Server 8 with Irix 3.2: +--------------- I'm running "nearly the next release", with the same effect. (I'll pass it own to the internal bugs followers...) +--------------- | So, any clues why /usr/bin/nawk should drop core on this script | (lifted pretty much verbatim from the nawk manual): | | END { | while ("who" | getline x) | print x | } | | when I run it: | /usr/bin/nawk -f test.awk test.awk | | without ever producing any output? +--------------- If you run it as: nawk -f text.awk Organization: Silicon Graphics, Inc. Subject: Re: Help !! Message-Id: <8071@odin.corp.sgi.com> References: <1990May17.210342.17949@ux1.cso.uiuc.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <1990May17.210342.17949@ux1.cso.uiuc.edu>, gautamm@kandinsky.ncsa.uiuc.edu (Gautam Mehrotra) writes: > > > WHile running a program, I got this message in my console > window : > > > NOTICE: File table overflow > > > The program failed to find some files ( which were there ). The > disks were not full and I could see no other abnormality ..... > I reran the program immediately afterwards and it ran just fine. > > Any clues to whats happening ? Why ? > There is a global table in the kernel that contains one entry for every open file in the system (essentially). You have filled up that table. The size of the table is static for any particular kernel and is controlled by the kernel configuration parameter NFILE in the file /usr/sysgen/master.d/kernel. There are two possibilities: either you legitimately need more file table entries than are configured by default in your kernel or there are some processes out there running amok and opening lots of files that they don't really need to open. You can monitor the current allocation state of the file table, process table and inode table by using the command 'sar -v'. For example: sar -v 2 100 will report the state of the tables every 2 seconds for 100 iterations. If you want to increase the size of your file table, modify the formula for NFILE and reconfigure your kernel by executing the script /etc/init.d/autoconfig. In 3.2 software, NFILE is expressed in terms of NPROC (the size of the process table): #define NFILE (16*NPROC/5+80) You can change this definition to suite the needs of your system.   Received: from vmb.brl.mil by VMB.BRL.MIL id aa12329; 24 May 90 1:52 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa11897; 24 May 90 0:18 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa11736; 24 May 90 0:02 EDT Received: from grace.waikato.ac.nz by VGR.BRL.MIL id aa20204; 23 May 90 23:53 EDT Received: from aukuni.ac.nz by waikato.ac.nz; Thu, 24 May 90 15:52 +1200 Received: from ccu1.aukuni.ac.nz by aukuni.ac.nz; Thu, 24 May 90 15:51 Y Received: by ccu1.aukuni.ac.nz (5.52/890607.SGI) (for @ccvcom.aukuni.ac.nz:info-iris@brl.mil) id AA29046; Thu, 24 May 90 15:48:04 NZT Date: Thu, 24 May 90 15:48:04 NZT From: russell@ccu1.aukuni.ac.nz To: info-iris@BRL.MIL Message-Id: <9005240348.AA29046@ccu1.aukuni.ac.nz> An update on my problems with exabyte tapes: Dave Olson pointed out that there is a bug in 3.2 that stops tar,bru etc from writing multiple files on a tape. Thanks Dave. On the tape label program the problem was one of patience. It takes over 2 min to write the label and about 1min to read it. This seems inordinately slow to me. I assume that unlike 1/2" tape the drive does not automatically search for the 'load point' when the tape is loaded so that the first read or write then has the get the tape into postition before actually doing the I/O. Somebody asked me what type of drive we have. I don't know. There is no id on the drive its self and we didnot get any details from SGI when we brought the machine. -- +-----------------------------------------------------------------+ | Russell Fulton 'phone +64 9 737-999 x 8955 | | Computer Centre domain rj_fulton@aukuni.ac.nz | | University of Auckland fax +64 9 303-2467 | | Private Bag time gmt -12 (13) (summer time)| | Auckland, New Zealand. | +-----------------------------------------------------------------+   Received: from vmb.brl.mil by VMB.BRL.MIL id aa12758; 24 May 90 3:32 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa12494; 24 May 90 2:19 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa12435; 24 May 90 2:03 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa20573; 24 May 90 1:56 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA11925; Wed, 23 May 90 21:39:28 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 24 May 90 04:09:45 GMT From: Tom Mitchell Organization: Silicon Graphics Computer Systems, Mountain View CA. 94039 Subject: Re: SG vs Sun Message-Id: <8111@odin.corp.sgi.com> References: <9005010243.AA15404@physics.phy.duke.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <9005010243.AA15404@physics.phy.duke.edu> rgb@PHY.DUKE.EDU ("Robert G. Brown") writes: * Dear Mike, and netsters in general, * * Geeze-oh-Goshen, that (or rather this) is the LAST time I'll ever * stick my neck out on a Publick Platform. Let me reiterate: * * a) We are a .edu subnet, e.g. -- a University. That means ..... Oh, I hope not. Your letter/posting about your environment and what you want to do in it was informative to me. Thanks, mitch -- Thomas P. Mitchell -- mitch@sgi.com or mitch%relay.csd@sgi.com "All things in moderation; including moderation."   Received: from vmb.brl.mil by VMB.BRL.MIL id aa16821; 24 May 90 9:33 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa15119; 24 May 90 8:20 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa15010; 24 May 90 8:06 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa21913; 24 May 90 7:56 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA00505; Thu, 24 May 90 04:49:34 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 24 May 90 10:01:46 GMT From: Scum Organization: Carnegie-Mellon University, CS/RI Subject: pointer to bgnpolygon, etc. Message-Id: <9388@pt.cs.cmu.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL I have an element in a structure which can point to bgnpolygon and bgnclosedline. Also, it has pointers to the appropriate end functions. The declaration of the element structure is: ... void (*startform)(); ... I tried to replace my calls to bgnpolygon(); with (self->startform)();, but the code now crashes. Is there some reason why I can't use pointers to these functions, or what? I treat them just like I do any other pointer to a function. Please, help. Thanks, -- Chris. -- -- Chris. (cycy@isl1.ri.cmu.edu) "People make me pro-nuclear." -- Margarette Smith   Received: from vmb.brl.mil by VMB.BRL.MIL id aa20784; 24 May 90 14:42 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa20397; 24 May 90 14:11 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa20389; 24 May 90 14:04 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa23686; 24 May 90 13:54 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA22036; Thu, 24 May 90 10:46:44 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 24 May 90 17:41:49 GMT From: Mark Rosenstein Organization: Bellcore, Morristown, NJ Subject: Memory for 320s Message-Id: Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL We are getting a 320 and I am trying to figure out what to do about memory. If someone actually has a 300 series machine and has put memory in it, especially 3rd party memory I'd be most interested in their experiences. (Also data from 200 series is useful.) Not having a machine yet, I can't verify the following, but would be happy to be corrected by folks who know. ---rumors follow ------ There are 2 simm carriers for the machine. Each carrier can hold 32 simms. Currently SGI is currently only selling 2mb Simms for the 320s The first board can only be populated with 2mb Simms. Third party vendors are selling 4mb simms and possibly 8mb simms. There is an upper bound on the amount of memory the machine can address, which is less than a fully populated machine with 8mb simms. The usual confusion about whether you can mix or match different types of simms. --- end of rumors ------ Any clarification with real data (no guesses, please!) would be helpful. Thanks. Mark. Life: The art of carefully throwing away opportunities.   Received: from vmb.brl.mil by VMB.BRL.MIL id aa22016; 24 May 90 16:05 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab21322; 24 May 90 15:22 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa21253; 24 May 90 15:10 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa24612; 24 May 90 14:55 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA26339; Thu, 24 May 90 11:52:07 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 24 May 90 18:45:36 GMT From: Scott Kahn Organization: University of Illinois at Urbana Subject: Re: Memory for 320s Message-Id: <1990May24.184536.6797@ux1.cso.uiuc.edu> References: Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL The absolute max for memory on 300 series machines is 256Mb, rather than the 128Mb limit on 200 series machines. My understanding is that the 300 series machines will access the larger memories using 4Mbit chip technol- ogy rather than adding two more memory boards. I also know that the direct address space on the R3000 is 512 Mbyte, so larger memories must await a newer technology from MIPS. Scott Kahn (uxh.cso.uiuc.edu)   Received: from vmb.brl.mil by VMB.BRL.MIL id aa22396; 24 May 90 16:37 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa21322; 24 May 90 15:22 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa21244; 24 May 90 15:09 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa24180; 24 May 90 14:41 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA23609; Thu, 24 May 90 11:09:44 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 24 May 90 18:09:40 GMT From: Frederick S Hann Organization: Univ. of Pittsburgh, Comp & Info Services Subject: SLATEC woes Message-Id: <24506@unix.cis.pitt.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL I am attempting to install the SLATEC math library, and need the machine dependent constants contained in r1mach.f, d1mach.f and i1mach.f. Anyone who has done this could save me a lot of time. Thanks. fred or fshst1@unix.cis.pitt.edu   Received: from vmb.brl.mil by VMB.BRL.MIL id aa23686; 24 May 90 17:40 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa21644; 24 May 90 15:55 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa21556; 24 May 90 15:34 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa24778; 24 May 90 15:30 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA27934; Thu, 24 May 90 12:17:12 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 24 May 90 19:15:03 GMT From: Jim Hollan Organization: Bellcore, Morristown, NJ Subject: Maximum Cable Lengths Message-Id: Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL I sent the following message to comp.sys.sgi Does anyone have experience locating monitors and keyboards more than 75 feet from the machine? What is the limit? We are interested in data for both Personal Irises and Power Series machines I append the responses I received. Jim ====== From: Steve Lamont Don't know about > 75 feet but ours work just dandy at 75 feet. We have a Sun 4/490 with 100 feet of keyboard and monitor cable and it works great too. I talked to one of the field engineers from SGI about this subject and he said that there should be no problem at 100 feet. spl ====== From: tarnoff@cme.nist.gov (Nicholas Tarnoff) We have a 4D/25GT with a regular monitor/keyboard cable and a 4D/220 GTXB with a 75ft extension. I don't see any difference and we have had no problems. We have no experience with longer cables. According to SGI they will not support longer extensions. I imagine that you can use a longer cable as long as it has extra shielding. -Nicholas ====== From: pswhip@avelon.lerc.nasa.gov (Dan Whipple) I can provide some info on this subject but not for the PI and 4D systems. We have a mix of 3030, PI, and 4D workstations and about two years ago had to remote the 3030 CPU box from the monitors because we were taxing the HVAC systems. I never got a good answer from SGI but the NAS people at NASA-Ames had done some testing and gave me 400 feet of a special Belden cable they had ordered (it may be a normal product now). I seem to remember it had been checked to 175 feet and worked. It has the coax lines plus enough twisted pairs to run the kybd and serial ports. (the kybd runs at 600 baud so no prob with length. The handbook says 50 feet is the limit for 9600 baud serial but I have run 200 ft. of 20 ga. shielded multiconductor cable at 9600 with no problems). The key characteristics of the coax lines are: impedance 75 ohms capacitance 17.4 pf/ft velocity 78% delay 1.3 ns/ft attenuation...db/100 ft 5 mhz 1.2 10 mhz 1.6 20 mhz 2.2 50 mhz 3.4 100 mhz 4.7 500 mhz 10.0 I am currently running 130 ft. lengths of this cable on two 3030 systems (1024 x 780 pixels). The only effect visible is a slight loss of brightness on the monitors. No smearing or loss of sharpness is visible. I don't know if the folks at NAS have tested any 4D systems yet but if needed I could try to hunt up a name for you to contact...let me know. Also let me know if you get any good technical data as sooner or later we will be faced with the same problem as the building HVAC can't handle all the workstations people want to buy here. dan ====== From: trb%bettysue.dallas@sgi.com (Tom Barton) We currently have two Power series machines that are at the end of 100' of monitor and keyboard cables. The monitors are a littel dimmer than they are on the short cables. Also be sure to make the RGB cable lengths as equal as possible in order to avoid syncronization problems between the RG & B signals. We ran into problems with the keyboard cables, due to improper grounding. It seems that someone with a static charge that touches the mouse or keyboard can cause the system to reboot. In order to solve this, we are have ordered the 65' keyboard cables that SGI makes. As I understand it, you need to have metal D connectors instead of the plastic ones we used. The metal connectors are somehow attached (or grounded) to the cable. I don't know how. SGI officially recommends that you do not exceed the 75' cable length. However, we do have the occasional customer who is successful with >75'. This oppinions are my own, and not SGI's ====== From: baskett%forest@sgi.com You can do more than 75 feet but the picture gets fuzzy. The pixels spread out, you know. I used to be able to explain the physics of why but I've forgotten. There are some small companies that make fiber optic repeaters that can give you several thousand feet of cable with sharp pictures. I think their primary market so far has been for government type installations so the volumes have been low and the prices are still high ( ~$4000 - last time I checked). Forest Baskett Silicon Graphics ====== From: rutgers!beaver.cs.washington.edu!ssc-vax!voodoo!zombie@bellcore.bellcore.com (Mike York) We've been running 23 4D/70G's at lengths between 125' and 189' for a year and half now. SGI makes no promises at lengths greater than 75', but we haven't had any problems. --   Received: from vmb.brl.mil by VMB.BRL.MIL id aa23977; 24 May 90 18:11 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab23686; 24 May 90 17:50 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa23605; 24 May 90 17:34 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa25449; 24 May 90 17:25 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA05520; Thu, 24 May 90 14:11:49 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 24 May 90 20:44:15 GMT From: Wiltse Carpenter Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Re: counting vertical retraces Message-Id: <60934@sgi.sgi.com> References: <9005212047.AA22371@chem.chem.ucsd.edu>, <57608@bu.edu.bu.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <57608@bu.edu.bu.edu>, tjh@beefeaterbu.edu (Tim Hall) writes: > Use the timer devices. If you do a "qdevice( TIMER0 )", this will > generate an event every 1/60th of a second. I don't remember if > it is directly tied into the screen retrace though. They are not tied to the screen retrace and they run at something more like 67hz. -Wiltse Carpenter   Received: from vmb.brl.mil by VMB.BRL.MIL id aa24603; 24 May 90 20:13 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa24244; 24 May 90 19:10 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa24195; 24 May 90 18:59 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa25871; 24 May 90 18:54 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA11448; Thu, 24 May 90 15:43:13 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 24 May 90 17:36:34 GMT From: Michael Zeitlin Organization: Texaco Houston Res. Cntr Hou, Tx Subject: attatch a process to a "window" on startup Message-Id: <420@texhrc.UUCP> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL When I log in and the window manager starts up, how can I attach a process to one of the subwindows (i.e. console) and invoke it (i.e. xinit) while maintaining the window Iconic?? Everytime I try to do this, all windows (wsh.....) invoke the command instead of just the console.... Is there an easy way to do this via the startup.ps or user.ps ???   Received: from vmb.brl.mil by VMB.BRL.MIL id aa24719; 24 May 90 20:34 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa24401; 24 May 90 19:37 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa24351; 24 May 90 19:22 EDT Received: from [129.112.1.12] by VGR.BRL.MIL id aa25963; 24 May 90 19:19 EDT Received: from baby.swmed.utexas.edu by utsw.swmed.utexas.edu with SMTP; Thu, 24 May 1990 18:19:39 CDT Received: by baby.swmed.utexas.edu (5.52/890619.SGI) (for utsw.swmed.utexas.edu!pt.cs.cmu.edu!isl1.ri.cmu.edu!cycy) id AA04985; Thu, 24 May 90 18:24:46 CDT Date: Thu, 24 May 90 18:24:46 CDT From: Rose Oguz Message-Id: <9005242324.AA04985@baby.swmed.utexas.edu> To: info-iris@BRL.MIL, pt.cs.cmu.edu!isl1.ri.cmu.edu!cycy@baby.swmed.utexas.edu Subject: Re: pointer to bgnpolygon, etc. I've done something similar. This is what worked for me: void (*bgn_ptr)(); ... bgn_ptr = bgnpolygon; ... (*bgn_ptr)(); My first and third lines seem like what you did. How did you assign the pointer to bgnpolygon()? I don't know if this will help, but at least you know it should work. Rose   Received: from vmb.brl.mil by VMB.BRL.MIL id aa24945; 24 May 90 21:05 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa24516; 24 May 90 20:02 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa24504; 24 May 90 19:56 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa26031; 24 May 90 19:40 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA14727; Thu, 24 May 90 16:33:56 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 24 May 90 23:32:15 GMT From: Scum Organization: Carnegie-Mellon University, CS/RI Subject: Re: pointers to bgnpolygon, etc. Message-Id: <9405@pt.cs.cmu.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Sign. Do I ever feel like an idiot. Nevermind my last question... I made a really stupid mistake, which I just discovered. Please, just pretend I never posted anything. I'm so embarrassed, I can't even bring myself to post it. Ever Redfaced, -- Chris. -- -- Chris. (cycy@isl1.ri.cmu.edu) "People make me pro-nuclear." -- Margarette Smith