From: Digestifier To: Linux-Misc@senator-bedfellow.mit.edu Reply-To: Linux-Misc@senator-bedfellow.mit.edu Date: Fri, 23 Sep 94 08:13:14 EDT Subject: Linux-Misc Digest #805 Linux-Misc Digest #805, Volume #2 Fri, 23 Sep 94 08:13:14 EDT Contents: SURVEY: Graphics card benchmarks under XFree86 (23 Sep 94) (Farrell McKay) Market survey: high-performance networking (Erann Gat) Re: Linux is a GNU system and the DWARF support (Chris Bitmead) Re: ext2 QUESTIONS (Unix answers) (Esther Heller) Re: Word Processor for Linux? (Philip Brown) Support for MediaVision Memphis (Bob Collie) Re: OpenStep on GNU or Linux? (David Jeske) Re: SB16 MCD and Mitsumi problem - Help (James CE Johnson) ---------------------------------------------------------------------------- From: fbm@thumbtack.bevc.blacksburg.va.us (Farrell McKay) Crossposted-To: comp.windows.x.i386unix,comp.benchmarks,comp.os.linux.help,aus.computers.linux,comp.sys.ibm.pc.hardware.video Subject: SURVEY: Graphics card benchmarks under XFree86 (23 Sep 94) Date: 23 Sep 1994 09:26:59 GMT Archive-name: xfree86-benchmarks Version: $Id: index.post,v 1.28 1994/09/15 05:29:29 fbm Exp fbm $ These are the latest results from a survey of xbench results under XFree86. The idea of the survey is to compile a comprehensive set of benchmark results, obtained by running xbench against all the main 386/486/586 graphics cards, and thereby place the graphics cards in an approximate performance ranking. Ideally all the results should be gathered using the same host machine, on the same day of the week, while wearing the same socks and jocks. Unfortunately no-one has yet come forward with such a database (although such databases are rumoured to exist). This survey is the next best thing. The summary results below distill the essence of the survey into single line entries. The entries are grouped by XFree86 release number, and ranked in decreasing 'xstone' order (last column) within each group. Ranking by xstones is a bit arbitrary. It is instructive to examine all the other columns too. The xbench survey has a home on the World Wide Web, at http://thumbtack.bevc.blacksburg.va.us/xbench/ If you have acess to a Web browser, check it out! It is at least 6 x 10^24 times better than this plain old Usenet posting, in lots of ways. New additions to this database are always welcome - especially entries for the latest servers or for unlisted graphics cards. If you use XFree86, you can contribute to this survey. Grab a copy of xbench, make sure it has the GraphicsExposures patch (see below), run it on your machine, fill in the blank form provided below, and mail it back to me. Please do not send the raw xbench output - just fill in the blank form! I will collate and post the results as they become available. XBench is available by ftp from hundreds of archive sites around the world. Check out archie for your nearest site. >>>>>> IT IS IMPORTANT TO USE THE CORRECT VERSION OF XBENCH. <<<<<< The correct version is the one with the "GraphicsExposures" patch. Check for this by looking at your source code - you should find calls to XSetGraphicsExposures(......, False); in the four files bitmapcpy.c complex1.c screencpy.c and scroll.c Some sites known to carry the patched version of xbench are: ftp.physics.su.oz.au:/xbench/xbench.tar.gz ftp.iastate.edu:/pub/netbsd/misc/xbench.tar.gz mcsun.eu.net:/os/linux/BETA/X_S3/xbench.tar.gz mcsun.eu.net:/os/linux/BETA/X_S3/xbench.tar.Z ftp.th-darmstadt.de:/pub/X11/other/xbench.tar.Z xbench.tar.gz file length=88785, /bin/sum=48788, /usr/5bin/sum=31503 xbench.tar.Z file length=156750, /bin/sum=21120, /usr/5bin/sum=40999 To run the benchmark, 1. If possible, switch into one of the more common display resolutions before starting, e.g. 800x600, 1024x768, 1280x1024 etc. 2. Kill off (i.e. exit from) all other programs on your machine, including your window manager! The only running program you should need is a single window from which to start xbench. XBench is a benchmark program. If it has to compete for memory or CPU, its results will suffer. 3. Read the XBench README file and edit the script.run file. 4. xset s off (turn off the screen saver) 5. xbench < script.run > results/YOURFILE 6. go have a coffee 7. awk -f scripts/xstones.awk < results/YOURFILE Farrell. -- Farrell McKay. Email: fbm@thumbtack.bevc.blacksburg.va.us (PGP key avail) ========================= Blank Form =========================================== CPU : (386/486 SX/DX etc) CPU Clock Speed : (MHz) M'board Memory : (MBytes) Card Vendor + Model : Card Bus : (ISA/EISA/VLB/PCI) Chipset : Video Memory : (MBytes) Memory Type : (DRAM/VRAM) Memory Speed : (ns) Clock Chip : RAMDAC : Operating system : XFree86 release : (e.g. 2.1.1) XFree86 server : (e.g. XF86_SVGA) Physical resoln : Virtual resoln : Dot-clock : (MHz) XBench lines : XBench fills : XBench blits : XBench arcs : XBench text : XBench complex : XBench xstones : >>>>>>>> Did you use the patched version of XBench ?? (Yes/No) <<<<<<<< XBench ftp'd from : Submitted by : Date stamp : ========================= Summary Results ====================================== --- Hware --- XFree Res Dot Px Display card Bus ln fl blt arc txt cplx xst CPU-Clk Mem Server P V Clk Sz Type --------- 000s ----------- =============================== XFree86 2.1.1 ================================= 486DX2-66 16 S3 N N 75 8 #9 GXE Level 12 V 442 80 65 6232 187 156 129 486DX-33 16 S3 E E 110 8 Actix Ultra+ I 323 68 59 2761 148 138 108 486DX-50 16 S3 N N 80 8 ELSA Winner 1000 I 303 63 54 3865 155 128 103 486DX-50 16 S3 C C 110 8 Metheus 928-4M V 310 67 50 3208 148 127 102 486DX-50 16 Mach32 N N 75 8 ATI Ultra Pro V 335 65 56 2845 142 84 101 486DX2-66 16 S3 N N 75 8 Diamond Stealth 24 V 340 54 45 5541 132 106 89 486SX-25 16 S3 P M 36 8 Spea V7 Mirage I 186 48 41 639 113 101 76 486DX2-66 20 S3 C C 135 8 #9 GXE Level 11 V 283 61 52 3345 56 121 74 486DX-33 16 S3 N N 72 8 Miro Crystal 10SD V 248 44 36 2256 104 87 71 486DX-33 8 S3 N N 75 8 Actix Grph Eng. 32 I 205 42 38 1504 101 89 70 486DX2-66 8 S3 N N 45 8 Generic S3-805 V 265 45 39 2259 66 95 67 486DX2-66 16 Mach32 c C 80 8 ATI Graphics Ultra V 271 36 31 3597 88 59 60 486DX2-66 32 SVGA K K 80 8 Generic Cirrus clg V 143 28 22 1862 281 44 51 486DX2-66 4 SVGA N N 75 8 VIA C.Logic clgd54 V 149 26 22 2097 238 42 49 486DX2-50 8 SVGA N H 75 8 Genoa 8500VL-28 V 113 23 23 1464 197 36 46 486DX2-66 20 SVGA D D 80 8 Generic Cirrus clg V 133 24 17 1708 254 40 42 386AM-40 8 Mach32 N N 65 8 ATI Graphics Ultra I 101 23 20 501 53 35 36 386DX-25 8 SVGA P L 36 8 Hercules Dynamite I 33 8 5 230 23 19 12 =============================== XFree86 2.1 ================================= 486DX2-66 32 S3 N N 77 8 STB Pegasus V 453 80 66 4318 195 156 131 486DX2-66 32 S3 N N 67 8 STB Pegasus V 461 79 66 4154 191 141 129 486DX2-66 32 S3 C C 125 8 STB Pegasus V 440 80 62 4240 193 153 127 Pnt-66 32 S3 N N 85 8 ELSA Winner 2000 V 494 75 57 6275 205 145 123 486DX2-66 16 S3 A A 185 8 #9 GXE Level 16 V 430 73 56 6833 182 139 117 486DX2-66 16 S3 F C 85 8 Spea V7 Mercury V 472 69 55 4287 171 136 113 486DX2-66 24 S3 C C 135 8 Spea V7 Mercury V 422 65 55 3404 162 132 109 486DX2-66 20 S3 N N 65 8 #9 GXE Level 11 I 374 65 55 4873 168 129 108 486DX2-66 8 S3 N L 75 8 Emca VGA-928-P P 370 64 55 3193 150 126 106 486DX-33 8 S3 C C 108 8 Spea V7 Mercury I 329 65 55 255 146 133 104 486DX2-66 16 S3 I I 96 8 ELSA Winner 1000 P 312 62 52 2907 148 128 101 486DX2-66 20 Mach32 H H 80 8 ATI Ultra Pro I 311 61 53 4336 144 133 100 486DX2-66 16 S3 H H 110 8 ELSA Winner 1000 P 298 60 52 2769 146 128 100 486DX-33 8 Mach32 O O 65 8 ATI Ultra XLR V 296 62 54 2723 129 83 95 486DX2-66 20 Mach32 N N 75 8 ATI Ultra Pro V 325 55 47 3437 129 75 88 486DX2-66 20 Mach32 H H 80 8 ATI Ultra Pro V 323 56 48 3682 118 77 88 486SX-33 16 S3 n N 45 8 VGA805-V from Uppe V 231 53 47 726 131 107 87 486DX-50 16 Mach32 C C 135 8 ATI Ultra Pro E 269 54 47 3038 123 74 85 486DX2-66 8 S3 N N 72 8 Actix Grph Eng. 32 V 328 48 40 3821 137 95 82 486DX-33 16 Mach8 N N 65 8 ATI Ultra (Mach8) I 255 47 42 4071 115 98 79 486DX-33 16 Mach32 N B 80 8 ATI Ultra Pro I 207 51 43 1600 106 61 76 486DX-33 16 S3 H H 62 8 ELSA Winner 1000 E 269 57 51 1720 50 130 69 Pnt-66 32 S3 C C 102 8 #9 GXE Level 12 P 389 72 61 4504 38 146 69 486DX-50 16 SVGA n N 45 8 STB Horizon VGA I 128 28 28 1767 224 55 56 486DX2-50 16 SVGA P H 45 8 ICL ValuePlus CL54 I 102 27 25 1131 145 38 48 486DX2-66 20 VGA256 N N 65 8 Cirrus Logic clgd5 V 61 28 20 1545 228 45 44 486DX-33 8 SVGA N N 65 8 Genoa Phantom 8900 V 131 28 12 2167 116 42 36 486DX2-66 32 SVGA N N 75 8 Generic Cirrus clg V 110 18 13 1474 229 30 33 486DX2-66 8 SVGA N N 75 8 Actix ProStar VL V 107 17 12 1563 205 27 30 486DX2-66 32 SVGA K K 80 8 Generic Cirrus clg V 102 16 12 1317 209 28 30 486DX2-50 16 SVGA H H 80 8 ICL ValuePlus CL54 I 76 14 12 880 140 20 26 486DX2-66 8 SVGA P N 36 8 Octek AVGA-20H I 80 11 5 891 54 17 14 486DX2-66 24 SVGA N N 65 8 Generic ET4000 I 91 8 3 1040 75 12 11 486SX-25 20 SVGA N L 45 8 STB PowerGraph I 67 8 4 416 52 14 11 486DX2-66 8 VGA16 Q P 26 8 Hedaka HED-622 I 20 3 3 471 7 7 5 386DX-25 8 SVGA P P 36 8 Trident TVGA 8900B I 18 2 1 156 12 2 3 =============================== XFree86 2.0 ================================= 486DX2-66 16 S3 N N 74 8 Diamond Stealth Pr V 451 92 78 3234 198 150 145 486DX2-66 16 S3 H H 85 8 #9 GXE Level 12 V 412 81 65 2872 196 157 130 486DX2-66 16 S3 N N 75 8 #9 GXE Level 12 V 468 80 64 4070 190 153 128 486DX-33 16 S3 N N 74 8 Diamond Stealth Pr V 278 83 75 2015 155 142 123 486DX-33 16 S3 N N 75 8 Diamond Stealth Pr V 348 76 66 2733 166 155 121 486DX-50 8 S3 N N 72 8 #9 GXE Level 11 V 395 71 62 3363 164 141 117 486SX-33 16 S3 C C 110 8 Actix Ultra+ I 293 73 62 3776 158 137 114 486DX-50 16 S3 N N 75 8 ELSA Winner 1000 E 316 64 55 2760 160 128 105 486SLC2-66 16 S3 N N 45 8 STB PowerGraph VL- V 282 60 50 1310 153 118 98 486DX-33 16 S3 N N 75 8 Orchid Fahrnht. VA V 266 48 40 2380 121 95 79 Pnt-60 16 Mach32 N N 75 8 ATI AX0 P 207 45 38 3438 126 68 73 486DX2-66 8 Mach32 G G 80 8 ATI AX0 P 183 39 34 3393 110 65 65 386DX-40 8 S3 N N 75 8 Generic S3-801 I 153 36 31 1153 84 72 57 486DX-50 16 SVGA Q L 25 8 ColorDesigner A3 V 126 23 12 1575 111 44 34 486DX-50 16 SVGA n N 45 8 STB Horizon VGA I 131 20 12 1551 130 47 33 386DX-40 8 S3 J J 95 8 Generic S3-801 I 122 23 20 1096 25 47 31 486DX-50 16 SVGA n L 45 8 ColorDesigner A3 V 115 18 10 1559 110 38 29 386DX-40 4 SVGA n G 45 8 Tseng Labs ET4000/ V 65 16 11 989 38 30 26 386DX-40 4 SVGA N G 72 8 Tseng Labs ET4000/ V 63 15 10 952 35 25 21 486DX2-50 16 SVGA P N 50 8 Diamond Speedstar I 76 16 7 927 61 39 21 486DX2-50 20 SVGA N N 85 8 Cirrus Logic clgd5 I 86 13 6 1203 68 33 18 486DX-33 8 SVGA N N 65 8 Diamond Speedstar+ I 87 9 4 1325 84 14 13 486DX2-66 8 SVGA P L 75 8 Octek AVGA-20H I 79 3 2 917 60 5 6 486SX-25 4 SVGA? Q Q 75 8 Sager NP840 notebo V 53 3 2 327 30 4 5 Key to Physical and Virtual resolutions (lower case=interlaced mode) A = 1600x1200 B = 1536x1024 C = 1280x1024 D = 1200x910 E = 1200x900 F = 1192x900 G = 1152x910 H = 1152x900 I = 1152x816 J = 1152x800 K = 1056x832 L = 1024x1024 M = 1024x900 N = 1024x768 O = 1024x767 P = 800x600 Q = 640x480 ------------------------------ From: gat@indra.caltech.edu (Erann Gat) Subject: Market survey: high-performance networking Date: Wed, 21 Sep 1994 14:57:41 -0800 Someone asked recently about ATM adapters for Linux. I am involved with a little startup company that is developing a new low-cost (<$1000) high-performance (>200 Mb/s) LAN. We are using Linux to do the system development, and I'd like to find out if there is a market for such a beast in the Linux community. Does anyone have any interest in a high-speed LAN with a Linux device driver? E. -- Erann Gat gat@indra.caltech.edu ------------------------------ From: chrisb@wombat.cssc-syd.tansu.com.au (Chris Bitmead) Crossposted-To: gnu.misc.discuss Subject: Re: Linux is a GNU system and the DWARF support Date: 23 Sep 94 14:48:23 In article <35huou$4ks@nkosi.well.com> gonzo@magnet.mednet.net (Patrick J. Volkerding) writes: >In article <35h334$g83@gandalf.rutgers.edu>, >Juana Moreno wrote: >>I like your idea. If you release something under this anti-GPL license, >>please let me know. I will take it, make a few modifications, sell it >>AND use the modified version along with GPL-ed programs. And if you >>really want to enforce your license, you'll need to make your wording as >>much or even more bizarre than the GPL one. So, just hire a good lawyer and >>do it, please. Maybe even someone could write an anti-(anti-GPL) license? > >This little blurb from the "file" source is about the closest I've even >seen. I'll reproduce part of it here for everyone's enlightenment: > >>This software is not subject to and may not be made subject to any >>license of the American Telephone and Telegraph Company (AT&T Inc.), >>UNIX System Laboratories (USL Inc.), Novell Inc., Sun Microsystems >>Inc., Digital Equipment Inc., Lotus Development Inc., the Regents of >>the University of California, The X Consortium or MIT, or The Free >>Software Foundation. > >Clearly there are people out there who aren't big fans of the GPL, and I >guess if I released something under a different (less restrictive?) license >and someone changed it a little and then GPLed it, I'm not sure how I'd >like it. FWIW, I have GPLed some of my own software before, but I think the >author has the right to say "derivatives of this software must *never* >be GPLed" just like the FSF says "derivatives of GPLed software must >*always* be GPLed". If you did this then I would just make a minor modification to the GPL, and call it say the "APL" and put you software under that. ------------------------------ From: eoh@raster.kodak.com (Esther Heller) Subject: Re: ext2 QUESTIONS (Unix answers) Date: 22 Sep 1994 14:42:06 GMT Reply-To: eoh@raster.kodak.com >I have a few questions and can't seem to find the answers to them: > > 1. I know that mke2fs reserves 5% of the disk space for the super user. > I even know how to set it to 0%. What I don't know is why 5% is > reserved in the first place? Does it hurt to set it to 0? > > 2. What is the purpose of the lost+found directory? > > 3. Why does ext2 have a bunch of options that don't do anything yet? > > >Thanks for your time. > >Michael. > >P.S. I have a hard copy of the ext2 FAQ in front of me. Please no cute > replies to read the FAQ ;-) You don't include a mail address so I can answer privately... Your questions are basically Unix, not Linux. 1. The reserved 5% of disc space gives you a very small amount of elbow room to get in and clean up things when your file system fills up for whatever reason. If the reason is some program merrily and unexpectedly writing a huge file of trash you will be grateful for the system stopping it before you don't even have enough space to look for it. As someone who has cleaned out file systems at 103% capacity, leave it on _all_ partitions. Trust me on this one! 2. There is a file system repair program called fsck that sometimes encounters a file that it can't figure out where belongs. Lost+found is a container that exists on all partitionss to dump them in. It should normally be empty, but I have had repair files land there that were perfectly human recognisable and useable. Yet another Unix perk. Fsck is what you use to fix things after a hard disc crash or someone pulled the power on a live file system, a truely amazing feature. 3. Don't know but would assume someone has kindly arranged for future features. I don't know the program but would look for a README file that came with it or for comments in the source since you say it isn't in the FAQ. Esther Heller eoh@raster.kodak.com Of course my opinions are my own! Interested in: Statistics, software QA, sewing, organic gardening, Hardanger, knitting, thread crochet, classical music, scratch cooking, woodworking... The 20th century version of the Proverbs 31 woman. ------------------------------ From: philb@cats.ucsc.edu (Philip Brown) Subject: Re: Word Processor for Linux? Date: 22 Sep 1994 16:15:12 GMT In article <35cd9h$8cm@superb.csc.ti.com>, Tony Cureington wrote: > I downloaded andrew last weekend and had it running in just a couple of >hours... >most of my time was spent reading the README's....I was very impressed! >To answere your >#2 quiestion below.....Yes, they do have just the WP part of andrew >available....this >was done by Terry Gliedt......Thanks Terry! Yeah, but I have a sparcstation, so I want a SOURCE distribution of such :-> but I am told that the authors are actually getting around to doing this themselves, so perhaps there's hope. -- ---------------------------------------------------------------------- Philip Brown, author of "kdrill", and "xmandel" philb@cats.ucsc.edu philb@soda.csua.berkeley.edu ------------------------------ From: collieb@iia.org (Bob Collie) Subject: Support for MediaVision Memphis Date: 23 Sep 1994 00:43:03 GMT Hello! I just was offered MediaVision's Memphis sound card/SCSI adapter combo for an excelent price. Could someone let me know if this card is supported in the Linux kernel (1.0.9) and Slackware (2.0)) Thanks! Bob Collie collieb@iia.org ------------------------------ From: jeske@ux4.cso.uiuc.edu (David Jeske) Crossposted-To: comp.sys.next.advocacy Subject: Re: OpenStep on GNU or Linux? Date: 23 Sep 1994 07:35:23 GMT dlj0@Lehigh.EDU (DAVID L. JOHNSON) writes: >They blew it from the beginning by not going with X. They didn't care about >compatibility, and paid the price. I don't see them changing heart. It >would be GREAT to see an X/NextStep combination (not an X emulator under NS), >but we'll have to wait and see. I don't think it is this simple at all. There are alot of compromises NeXT could have made to make their system more "standard". They "could" have used C++, they "could" have used X, they could have used more standard hardware. But then they would not be NeXT. Go run Solaris with a C++ development environment and X. Granted, NeXT needs to have open solutions to make it. However, there are ALOT of reasons that NeXT didn't use X which to them were more important than the reasons TO use X. Obviously, they have changed positions to try to make a bigger mark. OpenStep is going to be available on Solaris (on X) and the next version of NEXTSTEP is going to be OpenStep too. Therefore, you'll be able to do your NEXTSTEP stuff on a X machine. GNUSTEP (from what I understand) is just a project to get a GNU OpenStep compliant environment written. It will not be severly dependent on a given operating system, although it may require a certain lowest common denomenator. Also, someone mentioned something about "OpenStep" on "Windows 2000". There was actually some talk of plans for an "OpenStep" for Windows 95. Who knows if anything will actually come of it. -- David Jeske(N9LCA)/CompEng Student at Univ of Ill at Cham-Urbana/NeXT Programmer CoCreator of the GTalk Chat Software System - online at (708)998-0008 jeske@uiuc.edu (NeXTMail accepted) ------------------------------ From: jcej@tragus.atl.ga.us (James CE Johnson) Subject: Re: SB16 MCD and Mitsumi problem - Help Date: Fri, 23 Sep 1994 02:59:33 GMT Raymond Ho writes: > I have a Mitsumi FX001D controlled by a Sound Blaster 16 MCD. The > SB16 is their latest revision with the Mitsumi CD IO port set at > 0x230. The only options I have are 230, 250, 270 and 290, while with > the older board, the IO port can be set starting at 0x300. I was > trying to install the Yggdrasil Linux, the boot diskette insists on > looking for a Sony CD ROM drive, I guest it sees the IO port of 0x230 > being set, I believe that is where the Sony default. Can I use the > boot command to tell LILO that I have a Mitsumi CD at 0x230 and IRQ > 11? I think you've got one the same SB16/MCD that I have. Does it seem to have a serious lack of jumpers? If so then you've got the newer version. I called tech support on this and the deal is that the address of the CD will be 0x10 greater than the address of the SB. You have no choice in the matter. At the LILO prompt, try 'mcd=0xAAA,III' where 'AAA' is your CD's address and 'III' is it's IRQ. Once you get it all up and going, recompile the kernel to get rid of the Sony driver (unless you really do have a Sony CD:) [soapbox on] Evidently this is some great new "feature" that some standards group is trying to enforce. Personally, I'd rather that the cards allow you to choose ANY address (in the HW range of course) and ANY IRQ and ANY DMA. That way, you have to work really hard to get a conflict. With all the stuff I've got in my box, there is exactly ONE way to configure it. If I had more flexibility on a couple of cards then I might have more choices. As it stands, I can't add anything else because of conflicts. [soapbox off] Later, J ------------------------------ ** FOR YOUR REFERENCE ** The service address, to which questions about the list itself and requests to be added to or deleted from it should be directed, is: Internet: Linux-Misc-Request@NEWS-DIGESTS.MIT.EDU You can send mail to the entire list (and comp.os.linux.misc) via: Internet: Linux-Misc@NEWS-DIGESTS.MIT.EDU Linux may be obtained via one of these FTP sites: nic.funet.fi pub/OS/Linux tsx-11.mit.edu pub/linux sunsite.unc.edu pub/Linux End of Linux-Misc Digest ******************************