Re: how much processor time is allocated to a program
On 23 Feb, Frankie wrote: I am sure that when I first started using debian/linux I had a program (I think part of another packages), and if you ran thisprogram anotherprogram then it would rum anotherprogram and tell you exactly how much time was allocated to the running of that program. I don't know where to look in dselect, so can anyone remember what it was called? [maybe this would be a good time to suggest that dselect lets you search the descriptions as well] cheers, frankie time? -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: Floppy Drive Alignment Software
On 16 Feb, Person, Roderick wrote: Hi All, My 3.5 floppy is out of alignment, I know for other OSs there is software to help you align your drive is there any for Debian or Linux in general. thanks Rod Alignment? Are you sure? I thought that alignment problems needed an oscilloscope or at the least an expensive specially-made floppy. Anyway, I'm pretty sure I've never seen such a program for Linux. -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: dpkg and configuring packages
On 16 Feb, J K wrote: Hello, I was wondering how to configure packages that I installed with dpkg -i filename.deb. After installing a package, I got a message saying package depends on another file, and that that file was not installed. I have installed the other package now, and wonder how to configure the first package. Any help would be appreciated. Try dpkg --pending --configure That (attempts to) configure every package that has been installed but not configured yet. -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: PCMCIA cards fail under 2.2.1
On 11 Feb, Paul Nathan Puri wrote: I just rebooted my new 2.2.1 drive on my laptop. Where my bios is set to defaults (i.e., PnPOS), it locks up at boot. When I change my bios settings to enable serial port, and enable parallel port it boots up, but gives the following warning: cs: warning: no high memory available! cs: Request Window [something something] When pcmcia-cs attempts to detect my modem and nic it fails. Why would this be? Thank You NatePuri [EMAIL PROTECTED] Are you running version 3.0.8 of pcmcia-cs? Thre release notes for both kernel 2.2 and pcmcia-cs 3.0.8 indicate that you need 3.0.8 to work correctly under 2.2.x -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: NCR53C8xx Problem
On 11 Feb, Tim Buller wrote: A bit more info... It looks like the problem is being caused by the hot-swappable SCA bay that is part of the Intel Astor Chasis in which this system is installed. Info from the Symbios boot-up sequence: SCSI ID = 0 6 Vendor = ESG SHV Device = SCA HSBP M5 Sync = No On Thu, 11 Feb 1999, Tim Buller wrote: I'm trying to install on a Dual PII 450 Intel 440BX-based system. This MB has an integrated Symbios SYM53C876 SCSI controller, which uses the NCR53C8xx SCSI driver. The 2 9GB drives are recognized by the BIOS, and by the integrated Symbios SCSI utility. Booting from the rescue disk (I've tried the current disks from hamm, slink and potato), the kernel recognizes both drives, and then goes into fits of the following until it is rebooted or turned off: ncr53c875-0: copying script fragments into the on-board RAM SCSI host 0 abort (pid 23) time out - resetting SCSI bus is being reset for host 0 channel 0 ncr53c8xx_reset: pid 23 reset_flags=2 serial_number=$INC \ serial_number_at_timeout=$INC ncr53x875-0: restart (scsi reset) $INC is an integer that starts out 0 and increments by 1 each time this sequence is printed. Any ideas what I can pass to the NCR538xx driver to get it to work? Or other ideas? TIA, Tim I'm not so sure that it's caused by the bay; I have a Diamond Fireport 40 (SYM53c875J-0), also in a dual PPro, and I get the same messages with any of the 2.2 kernels, but not with 2.0.34. I haven't tried anything between 2.0.34 and 2.2.0. I've tried shuffling my devices around, even removing most of them, and it hasn't made any difference. -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: my kernel does not assemble!
On 5 Feb, Mike Marsman wrote: Hey all ... I think I have a question for ya (hopefully it's not completly stupid) I'm running deb w/ the 2.0.34 kernel installed -- I want to recompile the kernel so I can add network support, but I've come accross a problem. I can switch to /usr/src/linux and run 'make config' ... the I run 'make dep' and finally 'make zImage'. The config/dep work fine, and all the components of the kernel compile too.. but then the make zImage falls out on me after trying to run 'as86 ...', as86 being some program that's not on my machine! I've tried to search for as86, and have come up with nothing. I've even decompressed the kernel-2.0.34 source again, but same problem. Am I doing something stupid here, or is this legit? I'm gonna try 2.2.1, and maybe that'll work okay. It's just the as86 only appears in the Makefile for /usr/src/linux/arch/i386/kernel that has the as86 defined in it. I've tried replacing it with just plain 'ol as, but the switches for as86 don't work -- and it won't compile w/o the switches.. do I have to run something like 'make archdep'? .. many thanks from a confused user .. ..Mike I'm guessing (no, I'm positive) that you don't have the bin86 package installed. The long description for kernel-source-2.0.xx indicates that on intel platforms, you also need the bin86 package if you actually intend to compile the source. Just grab that and install it, and you'll be fine. Also, unless you're compiling a kernel for a non- Debian system, check out kernel-package. It provides a script (make-kpkg) that builds the kernel for you and makes a new Debian package out of it, with a revision number that you specify; this prevents dselect/apt/whatever from trying to replace your custom kernel with a prepackaged kernel. -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: kernel 2.2.0
On 28 Jan, Anthony Campbell wrote: On a slightly different theme, will there be patches available to upgrade 2.0.36 to 2.2.0, or do we have to start afresh? As this would be a 12 Meg download, it would take a long time and be expensive :( Anthony I'm afraid that enough has changed that you'd have a 12 Meg patch. OTOH, rumor on the developer list has it that the 2.2 kernel source will be included in slink, so it should be available if you get a slink cd (if you can wait that long :-) -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: MogoBIPS.
On 25 Jan, Chris Wong wrote: Hmm, Is there a certain # I should get from different CPUs? I mean, my P200MMX is getting like, 400 or so, and my dual P2-300 is getting a total of like 700 or so. Is this right? Chris Wong | [EMAIL PROTECTED] AD Digital Media Inc. (c) 1998 http://addm.com/ Yes, that's right. Pentium MMX chips should get roughly 2xclock speed. Pentium Pro / Pentium II chips get 0.99xclock speed (all per CPU, of course). Check the BogoMIPS HOWTO (in /usr/doc/HOWTO, if you've installed that package). -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: cdrom
On 5 Jan, ktb wrote: Hi, I've been hard at work getting nothing done as usual. I have three things I'm trying to accomplish but lets start with #1 and see if I can't get this to work. I've been trying to figure out how to use my cdrom. I have been reading the cdrom HOWTO. This is what I've done so far: crossyourfingers:/home/kent# ln -s /dev/hdd /dev/cdrom crossyourfingers:/home/kent# chmod 664 /dev/hdd crossyourfingers:/home/kent# ls -l /dev/hdd brw-rw-r-- 1 root disk 22, 64 Jul 20 19:45 /dev/hdd I can use the commands cdplay and cdeject just fine as a regular user. I can listen to a music cd through my headphones. I have an extras cd with my debian package and I would like to look at what it contains. I put the cd in, used the command mount /mnt/cdrom and got a message back that said there wasn't anything in the /etc/fstab file (there was another file listed but I forget what it was) so I followed the directions in the HOWTO. I put the following line in the file, /dev/hdd /mnt/cdrom iso9660 user,noauto,ro Tried again and got the error mount: mount point /mnt/cdrom does not exist Went back into the file and removed the following line /dev/hdd /cdrom iso9660noauto,nouser,ro 0 0 Tried to mount again same error mount: mount point /mnt/cdrom does not exist Does this directory (/mnt/cdrom) exist? Make it, put that line back in /etc/fstab and try again. So that is what I have done. What I would like to do is be able to listen to music cd's through my speakers. I put a cd in I start it, I verify it is working through my headphones, I take the headphone jack out and no sound. When I type an error (like try to backspace when there is no more space to travel) I get a beep that comes through my speakers. I don't understand much about sound cards. I don't have one in a slot. I looked in windows and came up with Crystal PnP Audio System maybe that is my sound card. It seems to me if Linux can use my speakers in one instance but not another there is a linkage problem? This sounds to me like a hardware problem - i.e., you need to have an audio cable going from your CDROM to your sound card. Since you don't seem to have an actual sound card, you should check to see if there is a slot on the motherboard to plug an audio CD cable into. The Crystal PNP Audio System is in fact your sound card; IIRC, that particular model is usually built into the motherboard. Where do your speakers plug in? There might be a connector near there. Also, does playing a CD through the speakers work in Windows? If that's all right, then check the volume setting in linux; after that, I'm stuck. I would also like to view what is on a particular cd without having to install what is on that cd. You should be able to look at the cd after you get the mounting bit fixed. thanks, Kent -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: can't use editor when su... why?
On 22 Dec, Kirk Hogenson wrote: The problem is that you own the X session, root doesn't. The easiest way to get this to work is to type xhost + localhost before you do your su. This means that you'll let anyone from the host localhost (ie, your computer) connect to your X. However, I recall there were some security risks associated with using xhost like this... maybe someone else will point them out. If you aren't connected to a network (or just dial up occasionally using, eg, ppp) then you should have no problems. (Using xhost + localhost helps, lots of people just use xhost +, which allows *anyone* from *anywhere* access -- bad idea.) If you don't trust people who might be logging in to your machine remotely, you probably don't want to do this. Kirk Jesse Evans wrote: Hi, folks! I like to use fte as my default editor, however, when I su to modify my system I cannot use it. I get the following messages: Xlib: connection to :0.0 refused by server Xlib: Client is not authorized to connect to Server Could not open display: :0.0 However, if I log in as root (as opposed to su from my users account, it works just fine. Any ideas as to what's up? xhost + localhost allows anyone logged in to open new windows on the display, capture keystrokes, etc. An alternate method (without the security problems of xhost + __) is to do export XAUTHORITY=/home/_user_currently_logged_in_/.Xauthority after the 'su'. This gives root the same permissions that the currently logged in user has over the X display, without extending those permissions to anyone else at the same time. -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: card selection in X for HP Kayak XC
On 17 Dec, Richard E. Hawkins Esq. wrote: I installed debian on a professor's machine yesterday, but have hit a barrier: we don't know what the video card that it uses is. Unix/ linux tech support in our department is non-existent (which is why he asked me for help). It can't even tell us what card type to use in configuration. The best he could come up with is that matrox usually works. However, we've hung the machine (not just X) three times this way. Does anyone know what setting to use? rick Is SuperProbe installed? Can you use that? SuperProbe is supposed to identify the chipset for you. It's in xbase. HTH, -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: Accupoint Mouse Pointer
On 16 Dec, Austanners Wet Blue Pty Ltd wrote: I have a Toshiba t211x series laptop with a pointer (accupoint) on the keyboard. Does anyone know how (if) I can configure this in X11? Regards, Stephen Lavelle I think it looks like a PS/2 mouse, software-wise. Try ps2 as the mouse type, and /dev/psaux as the mouse device. I'm not familiar with your exact model laptop, but most laptops have the built-in pointing device work that way. HTH, -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: wmaker
On 2 Nov, Debian-User wrote: can anyone tell me how to make images your background in wmaker? I dloaded some of those themes from e.themes.org which are .jpg format then i converted them to xpm with the gimp. But i can't figure out how to set it to my background. Can anyone help? Am I doing anything wrong? Thanks in advance. Don't bother converting them. Just dump the .jpg's in your ~/GNUstep/Library/WindowMaker/Backgrounds/ directory, then choose it from the Workplace-Appearance-Background-Images menu. HTH, -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: linux crash when RAM 128MB
On 1 Nov, Oz Dror wrote: Hi I have hust installed debian 2.0 on new P-II machine. MB ASUS P2B-S (with aic7890) I have 2 SDRAMS 128MB each. When I try to boot linux with more then 128M linux crashes. Is this a hardware or software problem. -Oz My guess is hardware. I have 256M and it works fine. One other thing to try, first, though: I'm assuming that you have the proper append=mem=256M line in your lilo.conf. Try setting that to 255M instead. Some boards borrow a little bit of RAM and it makes Linux very unhappy to not have all the ram it thinks it has. I've seen a couple of people suggest setting the mem= line to a little bit lower to compensate for this. You might also try swapping the DIMMS(?) to see if that changes anything. HTH, -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: lilo (linux only hd - dos only hd), no solution possible?
On 29 Oct, Kevin Grant wrote: Thanks for the help I've gotten on this so far. Unfortunately it doesn't seem to be working. I seem to be caught in a catch-22 situation as far as getting the system to behave as I want it to. Here's the situation: I have one dos/windows only hd. I have one linux only hd. I want to be able to boot up linux and then use lilo to move to dos if I desire. The earlier problem was that I had the linux hd as /dev/hda and the dos drive as /dev/hdb and lilo couldn't get dos to boot off of /dev/hdb. Apparently the dos hd has to be /dev/hda (drive C in dos terms) to boot. So I could only get into dos by telling my bios to swap the boot sequence around and boot off of drive D (apparently this caused the system to treat the dos drive as if it were drive C). To solve this I swapped the hd's around and made the dos drive /dev/hda and the linux drive /dev/hdb. However I still want to boot off of the linux drive, so I set the bios boot sequence so that it would boot from drive D (/dev/hdb). Unfortunately this seems to cause the bios to consider drive C not to be the master drive anymore so that when I try to use lilo to get into dos I get exactly the same error message as before, a sort of disk not available or non-system disk in drive sort of message. The catch-22 seems to be: the only way to boot linux from a linux-only hd as the default when the machine is turned on is to make the linux-only hd /dev/hda, or to alter the bios boot sequence so that the system treats it that way (no, I don't want to use a boot floopy to boot up linux either). At the same time, in order for lilo to boot up a dos-only hd, the dos-only hd has to be /dev/hda. Unfortunately both hd's cannot be /dev/hda at the same time. So if my analysis is correct, and unless lilo has some way to make the system change the hd that it considers the primary drive, there is no way to accomplish what I wish to accomplish. The problem is that I am a linux newbie and can't tell if I'm making sense here. Comments and possible solutions, if any exist, are requested. Oh, and my latest attempt has netted me a cannot start virtual terminal sort of error message on the linux side, followed by a freeze, so that I can't boot linux at all. This raises another interesting issue. Since the lilo.config I'm experimenting with now contains lines like boot = /dev/hdb1, if I change the drives back the way they were (with the linux drive as /dev/hda) presumably I won't be able to boot linux because the lilo.config stuff will point to the wrong place. On the other hand, I can't boot into linux now to change the lilo.config file. So how do you get into linux to change a lilo.config file when you cannot get into linux until after you've changed the file? :) I'm guessing I've got to resort to the boot floppies at this point, right? Even if this is the case, there should be a sort of general solution to this problem, to let you experiment with the contents of lilo.config safely. I.E. to write a new one that doesn't work, and then to be able to get into linux anyway, without having to resort to a boot floppy, by some command that tells lilo to boot up using a safe backup config file that you know works. Is this option around and being new I'm just not aware of it? If you've actually read this far thanks for your attention, -Kevin Whew! I'm not sure what to tell you as to how to get from where you are to where you want to be, but: Dos/Win really must boot from the first hard drive, i.e. hda. So that one has to go first, and linux has to go on hdb. When you have the BIOS switch things around, dos gets confused, so that won't do either. What you need to do is to put the hard drives in with the dos drive as hda and the linux drive as hdb, with the BIOS set to boot from hda. From here, what you would like to have happen is to have lilo boot from hda and boot into linux on hdb or dos on hda. I think your /etc/lilo.conf should look something like: boot=/dev/hda(this line is the key) root=/dev/hdb1 (or whatever) .(regular linux boot stuff) (and any other stuff you need or want) other=/dev/hda1 (or whichever partition dos is really on) label=dos The first line tells lilo to install itself on the master boot record of hda, so that it will be called first in the boot sequence. The rest just tells it where everything lives, so that it can find it appropriately. HTH, -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: couple of question..
On 7 Oct, tracheotomy_bob wrote: First the most important question. Does anyone have experience running a Matrox Millennium G200? I assume it needs special drivers as they're not standard in XFree86, so where do I get them from? Can anyone recommend a truly excellent card for X. Not a G200, but I do have the original Millenium. However, I there are Matrox drivers for the G200 that have been tweaked by S.U.S.E. and donated back to XFree86. Try their web site (http://www.suse.de/XSuSE/XSuSE_E.html). There are a number of people (including me) who swear by Matrox cards, especially the Millenium line. If I install just the binary CD (cheapbytes) and run XF86Setup. Setup runs ok, but when I exit, I have no visible text when I type I have to log out and then everything is ok. It's almost as though the normal vga driver is being screwed. Has anyone else had this happen? (I've got a Videologic GrafixStar 600) Dunno 'bout this one, sorry. I'll keep ya posted... -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: some WindowMaker questions ...
On 5 Oct, Nuno Carvalho wrote: On Mon, 5 Oct 1998, Marcelo E. Magallon wrote: Make an ~/.xsession file and put exec wmaker on it. Then chmod +x it. I just commented the line with xterm on Xsession global (/etc/X11/) and it worked ! :) The file is /etc/X11/WindowMaker/menu.hook, but you don't want to change that. If you want to modify an entry, copy it from /usr/lib/menu to /etc/menu, edit it, and run update-menus. I'll do it ! :) Thanks for the help ! Best regards, Nuno Carvalho BTW: I want to start some aplication when wmaker starts. How can I do it ? I already tried autostart file but it didn't worked ! ;( Is there any sintax for it ? $ cat ~/GNUstep/Library/WindowMaker/autostart /usr/local/bin/xradiotrack $ My ~/GNUstep/Library/WindowMaker/autostart is a shell script; i.e., starts with #! /bin/sh and permissions are set to 744 (i.e. owner read/write/execute, everyone else read-only. HTH, -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: CPU temp/fan tachometer for linux?
On 2 Oct, Peter S Galbraith wrote: [EMAIL PROTECTED] wrote: Try lm_sensors. http://appindex.freshmeat.net/view/904554804/ Cool. Have you compiled this? (Which do you pick? lm78-1.2.3.tgz or lm_sensors-1.4.6.tar.gz) I get: # insmod ./lm_sensors.o ./lm_sensors.o: kernel-module version mismatch ./lm_sensors.o was compiled for kernel version 2.0.33 while this kernel is version 2.0.35. It must be getting the wrong kernel version 2.0.33 from /usr/include/linux/version.h $ grep linux/version *.c lmsysctl.c:#include linux/version.h smbus.c:#include linux/version.h $ dpkg -S /usr/include/linux/version.h libc6-dev: /usr/include/linux/version.h If I force it: # insmod -f ./lm_sensors.o Warning: kernel-module version mismatch ./lm_sensors.o was compiled for kernel version 2.0.33 while this kernel is version 2.0.35 ./lm_sensors.o: init_module: Device or resource busy So no luck so far. (This is why I love .debs !) Jeff, I hope this answers your question as well. Let's see, I have a lm_sensors-1.4.2.tgz. I have compiled it, and I have also gotten the same error; I have also fixed this. In the Makefile, change the line CFLAGS = -D__KERNEL__ -DMODULE -I. -O2 -fomit-frame-pointer to read CFLAGS = -D__KERNEL__ -DMODULE -I/usr/src/kernel-source-2.0.34/include -I. -O2 -fomit-frame-pointer instead. (the above should be all one line; it'll probably wrap or something in between here there.) Change the kernel-source-2.0.34 to reflect whatever kernel you really are running. After all this, I've found that my board is one of the few P6-based boards that *doesn't* have these sensors on board. :-( I'd love to be proven wrong, but the lm_sensors module claims there isn't any such thing present in my system (Tyan Titan Pro S1662D, FWIW). -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: CPU temp/fan tachometer for linux?
On 1 Oct, Paul Reavis wrote: My latest motherboard has a CPU temp sensor and fan tachometer feature. Are there any linux drivers/programs for reading these? Try lm_sensors. http://appindex.freshmeat.net/view/904554804/ -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: pcmcia scsi
On 22 Sep, Giuseppe Sacco wrote: Hi all, I would like to buy a pcmcia scsi card for my laptop. Can someone suggest me a model that is know it works with Debian linux? There is a complete list of supported cards at http://hyper.stanford.edu/~dhinds/pcmcia/SUPPORTED.CARDS Unfortunately, there is no information as to how well each one works. Anything listed as a supported card is supported by the pcmcia package(s) distributed as part of Debian, though you may have to recompile the pcmcia modules to match your kernel. Anything listed as supported by a contributed driver is not (yet) supported by the Debian packages, though you can probably add the contributed drivers in yourself fairly easily. I have a recent New Media Bus Toaster, which is only supported by a contributed driver; however, it was fairly easy to rebuild the pcmcia package to include the new driver, and it has worked well for me. I've used a CDROM, a Syquest EZ-drive and a Seagate tape backup with it. The Bus Toaster was also relatively inexpensive (I got mine at an auction), which made a big difference in my decision. HTH, -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: Modules problems
On 19 Sep, Remo Badii wrote: [snip] Manoj Srivastava wrote: Possibly you did not build and install the pcmcia modules? The pcmcia modules are added on to the kernel, and come separately. Install the pcmcia source package and look for instructions on using make-kpkg to create pcmcia modules. I could not find instructions for the creation of pcmcia modules: have they been created in the original Debian setup while installing pcmcia-cs? If this is the case, I repeat my question above: must pcmcia-cs be reinstalled with the new kernel running? [snip] Try /usr/doc/pcmcia-source/README. -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: Proposed-updates subdir?
On 18 Sep, Randy Edwards wrote: Could someone tell me exactly what the dists/proposed-updates subdirectory is for? Yes, I know it's for proposed updates ;-), but what for what dist? Hamm? Slink? Does this subdir function as sort of a new incoming type of subdir so that the files are checked and then moved into Hamm or Slink? Any details on what this is used for would be appreciated; thanks in advance. For now, hamm. The idea is that unstable is, well, unstable, and changing all the time. Changes for unstable just get tossed into the archive. Stable, however, has been released and now *shouldn't* be changing. If it were to change without changing the version number of the release as a whole, there would be a flame fest on debian-devel and six more developers would quit over it*. So proposed-updates holding tank, if you will, to hold the changes to stable until there are enough of them or there is a sufficiently severe bug fixed to warrant releasing 2.0 r1 (or whatever they've decided to call it**). * this has happened before. ** And there have been people threatening to quit over what to call the new version number as well. -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: gimp /or xv resize function?
On 16 Sep, [EMAIL PROTECTED] wrote: Hi everybody, I just installed gimp and was trying to resize an image from 640x480 to 1024x768. I can't seem to do it. In gimp if you use the resize the image it resizes the window and leaves the image at 640x480. Do you have any suggestions. I can't seem to get this task done. There are no features that I can find with xv to resize an image to a bigger size. I have been to the gimp and xv site and could not find the answers there. I am not an artist just a guy trying to customize his enlightenment desktop. If any one can help me, I would really appreciate it and be greatful. Paul :-) Just did this, trying to customize my WindowMaker desktop Try m, for maximize in xv, or (right click on image) - Image - Scale in gimp. You may need to uncheck the Constrain Ratio box to get it to exactly the size you want. -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
RE: debian 2.0 install
On 16 Sep, Shane S. wrote: Thanks for the response. So do you think I should forget about recompiling the kernel and just install it as if there is 1 processor? -Original Message- From: Nathan E Norman [mailto:[EMAIL PROTECTED] Sent: Wednesday, September 16, 1998 11:31 AM To: Shane S. Cc: 'debian-user@lists.debian.org' Subject: Re: debian 2.0 install On Wed, 16 Sep 1998, Shane S. wrote: : : : : I just downloaded debian 2.0 linux and I want to install it : on a SMP Wyse7000i machine. This machine has 2 486 processor. : After reading the documentation, I have a question. : After downloading, do I install debian 2.0 first, and : then re-compile the kernel to take advantage of the SMP, : or should I recompile in the begining? I don't believe Linux SMP code supports dual 486 processors (due to the lack of standards for such machines). Pentium and above multi-processor systems are supported. Before you give up: My impression is that Linux SMP doesn't support dual 486SX boxes, because of something involving floating point support, but that it does support some 486 dual processor boxes. From the Linux SMP page (way out of date, but presumably correct as of whenever it was written: * Intel MP v1.1 compliant 486, Pentium and Pentium Pro hardware. * Intel MP v1.4 compliant Pentium and Pentium Pro hardware. * Multiprocessor Sun4m sparc machines. * 2.1 development on Alpha, PowerPC, Sun4d, Ultrasparc. I would suggest installing as usual, pretending you have a single processor box, which should work. Then install the kernel source and compile a custom kernel which supports SMP, and put it in LILO as a secondary kernel (second menu entry), so that you can try it out but have something to fall back on (the single processor kernel) in case it really doesn't work. If it does work, shuffle the LILO configuration around to boot the SMP kernel by default. I've been really lucky with SMP support, but I haven't tried it with 486s. I think I've run into one SMP related bug, and got around that by not using the SCSI port on my (now replaced) PAS16 sound card. -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: 2nd Try: Ctrl-Alt-Minus on Laptop
On 14 Sep, Kent West wrote: Anyone know how to change resolution/screen size on a laptop that doesn't have a separate numeric keypad? I can lock the number keys on that are imbedded in the alpha keys, but that still doesn't let me do the Ctrl-Alt-Minus key combo. Nothing I've tried has worked. Is there another way to change res without this key combo? Are you using the minus key in the embedded numeric keypad when you try this? (Mine is on the 'p' key, and it works for me.) Reorder the resolutions in your XF86Config if that doesn't work. Maybe you'll have to remove all but the resolution you want from the Screen section. -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: KDE problems.
On 14 Sep, Person, Rod wrote: Anyone using KDE, do you have this problem, can you help.. When I start KDE as the default window manager it takes 1/2 hour or more to load. Then once it does load it is take just as long to do anything. I added kdmdesktop to my Xsetup_0 (as the last line, not sure if this is correct placement or not.) I made kde primary in the window-manger file. Other than that I made no modifications, I am thinking the is a problem in the search paths, but I am clueless to which config file to check. Any idea? By the way, please reply to me directly at this address, for some reason I don't get the journal anymore. I think that my system admin, locked it out!!! Rod. I'm replying to the list also, so that other people can follow along correct me if necessary. I don't have this problem. OTOH, I can't run KDE on the slowest system I have, because it doesn't have a graphics card installed. (Monochrome text only!) How fast is your system? Which version of KDE do you have installed? How much RAM do you have? Swap space? KDE takes a fair amount of memory, last time I checked. I think my PPC box (PPC 604/150, 16M) goes from 11M used to 28M used when I start X KDE, which is bad because I have only 16M RAM, but it isn't anywhere near as slow as what you're describing. For starters, check /var/log/kern.log, /var/log/daemon.log and /var/log/messages. I'm not sure what else to check. -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: KDE's Kedit and Kview
On 12 Sep, Michael Beattie wrote: I know that KDE is still in the development stages, but one small bug that bugs me is that when using the K file manager to open text, .c or image files, The respective viewers/editors take the file's path and location to be seperate arguments. For example, Kedit says Cant open directory with the first two program instances, and then opens the file in the third instance when passed the filename and path of /home/userfoo/bar.txt The image viewer opens with %m and [filename] as available images to view. I dont know if anyone understnds what I am describing, I haven't been motivated enough to ask until today.. Anyway, Is there a temporary fix for this, or do I have to wait till the next packaged release? I hope I am not the only one with these problems! I had this problem with the older versions of KDE. It went away when I installed the version from slink. (1.0) You can get it (probably temporarily only) from slink, or from the KDE web site (they have the .deb's someplace there). -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: Help! CD-RW Media
On 9 Sep, [EMAIL PROTECTED] wrote: Hi, i'm trying to mount some CDRW media, but its not working. :(. SNIP [EMAIL PROTECTED]:/home/timothy# mount -t iso9660 /dev/hdc /cdrom mount: block device /dev/hdc is write-protected, mounting read-only mount: wrong fs type, bad option, bad superblock on /dev/hdc, or too many mounted file systems /SNIP .. So I suppose its not iso9660? Now, man mount says that without the '-t' option, it will probe the filesystem type. But SNIP [EMAIL PROTECTED]:/home/timothy# mount /dev/hdc /cdrom /dev/hdc: Input/output error mount: block device /dev/hdc is write-protected, mounting read-only /dev/hdc: Input/output error mount: you must specify the filesystem type /SNIP Anyone happen to know whats wrong here? What kind of filesystem CDRW drives normally use. (This one was made in an 'omniwriter')? Or some other way that I can probe the filesystem and see what it is? You wouldn't happen to be trying to mount this one in a standard CDROM drive, would you? Not all CDROM drives are capable of reading CDRW media, only some recent models. In fact, not all CDROM drives are capable of reading even CDR media, as I recently discovered to my chagrin. I get similar errors trying to mount a CDR in my (admittedly ancient) CDROM, but if I put that CDR disc in my Mac, I can mount and read it, no problem. I forget where I first heard about this, but I think it has something to do with how reflective the metal layer inside the media is. -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: swap size
On 2 Sep, Max wrote: I'm about ready to setup a machine with 512 MB of RAM and I'm wondering how much swap space I should allocate. I've read about the 2x rule, but 1 GB of swap seems somewhat excessive. I've also been told that Linux will not use more than 128 MB of swap. So, how much should I allocate? Can I change it later without having to reinstall everything? The machine is a dual-processor workstation that will be used by 4-8 people concurrently for number crunching and all sorts of simulations. Also, is it possible to put /tmp within swap? Is that the default? Thanks, Max 1GB isn't excessive if you're going to use it all. A better rule is to try to figure out what you will need*, and make sure that real + swap is at least as large as what you need. Of course, it really does help if real average need and you only need to hit swap on rare occasions. If it turns out that you only need 480MB, max., then you probably don't need any swap at all. OTOH, if you are running something really huge, then you might need 2GB swap, although that might turn out to be simply too large for your system. I think there is some benefit to having some swap available, but I'm not enough of a kernel guru to know exactly why. I did have a situation recently where I didn't have any swap, and I ran out of RAM, and the system shut down sshd, breaking my connection. Dunno why. * - don't forget the kernel, all the daemons, etc.etc.etc. At any rate, so long as you have disk space available, you can change the amount of swap by running mkswap and swapon with suitable partitions, and you can do this on the fly. You can even do this with swap files in the regular filesystem, but I understand this isn't recommended, for performance reasons and security reasons. I haven't done this, so I don't know how, but I'm sure it's in a HOWTO or a man page someplace. Worst case, you could add another hard drive and use it for swap space if it turns out you need more. The correct way to set up swap involves separate partitions, with a different partition type and a separate format; /tmp needs to be part of the regular filesystem, mounted in the usual manner, so I doubt you can share space between /tmp and the swap space. HTH, -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: Sound card question, mouse question
On 29 Aug, Christopher M. Wesneski wrote: I recently learned how to setup my Plug-and-Pray modem using pnpdump and isapnp. pnpdump /etc/isapnp.conf (then select the correct settings) isapnp /etc/isapnp.conf setserial /dev/ttyS1 port 0x2f8 irq 3 uart 16550 Everything works great. My question is, pnpdump also listed my sound card. When I edited isapnp.conf to select the correct settings and ran isapnp everything looks ok but the card doesn't work. Is there something similar to setserial that I have to do after I run isapnp? Also I went out and bought a three button mouse so I could middle-click cut-n-paste in xterms but I can't seem to get it to work. I have gpm running and it works while not in X (at the console). But when I am in X the middle button does nothing. It is a $0.50 serial port mouse (GoldenImage) and it has a little switch on the bottom (which I have no clue what it does) with one side labeled MS AM (which is the only way it will work in Windows) and the other PC AT. Any suggestions? You might get a better response with better information in your question. What model is your sound card? Are the settings listed for it in /etc/isapnp.conf correct? Do you get any error messages when trying to use the sound card or does it just fail silently? I get a Interrupt test on IRQ5 failed - device disabled when my sound card doesn't work, meaning that Plug-and-Pray has in fact introduced an interrupt conflict into the system. Most importantly, does the kernel you're running have sound support? (I always compile custom kernels, so I don't remember if the default Debian kernel has sound support or not.) You may need to recompile your kernel with sound support on. As for your mouse, the switch on the bottom changes which protocol the mouse uses; have you read the 3-Button-Mouse HOWTO? (It's in /usr/doc/HOWTO/mini/, and may have the information you need to decide where that switch belongs.) Also, you need to make sure that the information in your XF86Config matches your mouse hardware. You should tell X that you have whatever kind of mouse you have set the switch to. -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: Custom boot Disk Creation?
On 28 Aug, Young, Ed wrote: I'll bet this has been covered, and I feel shame for not knowing and not being able to find out how, but How do I create a custom boot disk? Is there a utility, or just a sequence of steps? Thanx, Ed The package you want it boot-floppies. Be warned, there are some things you need that are not listed in the dependencies for the package. Among other things, the kernel you are running when you build the boot disks needs to have loopback support, minix filesystem support, and possibly a couple of other things I've now forgotten. Worse, an upgrade to the boot-floppies package replaced the Makefile I had customized with a standard one, so I'm not positive about everything I changed to make it work. You'll also need a local copy of the Debian archive, probably on CD-ROM (I NFS mounted our local mirror and that worked, too). The boot-floppies package isn't as easy to use out of the box as many of the other Debian packaged items are, but it does make nice boot disks (I built a set for my laptop when PCMCIA support on the boot disks was broken). HTH, -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: Beowulf cluster
On 28 Aug, Ossama Othman wrote: from what I remember Beowulf uses PVM, and that is already packaged. They basically seemed when I read their pages to be makeing add ons which make PVM more powerfull From what I recall, PVM is being superseded by MPI. There is already a Debian packaged implementation of MPI called mpich. I'd suggest using MPI instead of PVM. By the way, has anyone adopted MPICH since it was orphaned (poor little mpich :))? If not, I'll adopt it. -Ossama My impression, from having looked into them both, is that PVM is better at some things and MPI is better at others, so both are still used. BTW, in answer to the original question (which I've since deleted), is that you can use SMP in a Beowulf cluster. At least, there isn't any technical reason that I know of why you couldn't. Simply build the Beowulf cluster out of SMP-capable systems, install multiple processors in each one, and make sure to compile SMP into the kernels when you set it up. My guess is that most Beowulf class clusters don't do this, because the limiting factor on computation speed in most clusters is communication speed. Having more than one processor in the same box only exaggerates the communication speed problem. Also, most PC-based SMP implementations have difficulty with memory bandwidth. I followed the linux-smp mailing list for a short while; one of the things that came up was that there were many SMP boxes which ran two memory intensive processes slower in parallel than they did sequentially. That is, it was actually faster to not use the second processor! My guess is that you haven't seen anything on SMP Beowulfs preciesly because of the memory and network bandwidth problems, but I've never built a Beowulf, so I don't really want to put words in the mouth of anyone who has. -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: sound card
On 28 Aug, Brian Morgan wrote: Running debian 2.0 / 2.0.34 (stable) and having trouble getting my sound card to work. I can get the cd rom player in X to run, but not through the sound card. I didn't see anything in the device drivers setup when I installed debian, and wasn't sure how else to get it to work. It's a creative labs soundblaster awe64e. I installed all the sound files in the debian setup (using the easy to use custom menu for setting up: WORKS GREAT!!!), but it still isn't working. Any thoughts? Thanks, Brian Morgan BTW, nice job to all you developers out there who made vast improvements to the debian 2.0 setup. So much easier now than in the frozen version! I installed everything I needed right from the start (including x, samba, and mucho other packages . . .no problem!) Look out NT! You'll need to patch the kernel source and compile a custom kernel; fortunately, this is a) not hard, and b) well documented in the Sound Blaster AWE HOWTO, by our very own Marcus Brinkmann and c) there are Debian packages for everything you'll need. /usr/doc/HOWTO/mini/Soundblaster-AWE.gz (or search Yahoo for Sound Blaster AWE HOWTO - the first 22 matches are to copies of the HOWTO) This HOWTO made the difference for me - I needed to replace my sound card, as the old one bumps into my SIMM slots. The options were yank half of my RAM (which I had just bought) or replace the sound card, so I went shopping, and Marcus' excellent HOWTO was the deciding factor in my getting a SoundBlaster 64 AWE. Read the HOWTO, follow it, and yell for help if you get stuck. HTH, -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: Rescue disk is SLOW to load
On 26 Aug, Mike Roberts wrote: I and a friend both have the same problem on our older 386 machines - the resue disk for hamm takes about 15 to 20 minutes to load the root filesystem and the kernel image. These same floppies work just fine on newer systems. We've both already completed the installs, but why does it take so long? Granted, the processors are slower, but we didn't see this problem with 1.3. Which rescue disk? Are you using resc1440-fast.bin or resc1440.bin? The resc1440.bin is known to be atrociously slow, but safe. resc1440-fast.bin is *much* faster at loading. -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: e2fsck error on boot
On 15 Aug, Mark Wagnon wrote: Hi all- I've managed to install the base system via floppies and I've spent several hours downloading packaged with dselect, and all seemed to be progressing nicely. I've only rebooted the system twice, and after downloading and installing something to do with the libc5 libraries, dselect instructed me to reboot when it was finished. So after everything was installed and configured, I rebooted the machine with Ctrl-Alt-Del, but I didn't stop to think if that's okay with Debian (it's posed no problem with any other dist that I've installed), so the system rebooted and when fsck started to check my filesystems it barfed up the following: ---Begin fsck message--- Couldn't find ext2 super-block while trying to open /dev/hda The superblock could not be read or does not describe a correct ext2 filesystem. If the device is valid and it really contains an ext2 filesystem (and not swap or ufs or something else), then the superblock is corrupt, and you might try running e2fsck with an alternate superblock: e2fsck -b 8193 device fsck failed. Please repair manually and reboot. Please note that the root filesystem is currently mounted read-only. To remount it read-write: #mount -n -o remount,rw / ---End fsck message --- I entered my root password and typed e2fsck -b 8193 /dev/hda1 at the ^ prompt. fsck ran and asked me to fix a few things, and exited without any error messages. I rebooted (this time with shutdown -r now) and was presented with the same error message. I tried again, this time mounting the root filesystem read-write as mentioned above, but then fsck complained that the system was mounted and asked me if it should continue. Not sure of the consequences, I declined. The only thing that I can think of that affected the mounting of filesystems was I edited the /etc/fstab to include my cdrom, cdwriter, and floppy drive. I cleaned it up a little as it was ugly, but I didn't modify any of the info on lines already present execept to format them nicer with tabs. I've never been presented with an error like this before, so I don't know what to do or where to look. I can't even bring up a man page on e2fsck. Am I screwed, or is there a solution to my dilemna? I would surely be disappointed if I had to trash everything I had accomplished up to this point. Heck I was proud of myself, because last year I couldn't figure out dselect! Thanks for any help, Curious question - does the bootup message really say /dev/hda rather than /dev/hda1? If so, that may be your problem. Check to see if in editing your /etc/fstab, you didn't lose a 1 in the first line. -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College
Re: PC Card Modem and Printer Recommendations
On 28 Jul, Justin Liu wrote: Hi - I have an IBM Thinkpad 560 with Debian 1.3.1 (among other OS's) installed on it and I was looking for some equipment recommendations. I plan an upgrade to hamm sometime soon. I want to get a 56K PC card modem, hopefully one that will also work with OS/2 and Windoze if it needs to. The Xircom 10/100 56K combo card looks interesting, just because it combines so much into a small package, but cost and ease of installation is more imporatant to me than throughput. I'm also looking for a low-cost ink-jet with good quality, especially for the occasional color output. I'll be a student again this fall, so I'll have use of the school's laser printers for black and white output, but I want something to turn out drafts at home. I might consider a portable printer if the quality and price don't drop off that much. Again OS compatibility is a major factor and speed is not. The HP and Epson look OK. Are multifunction devices worth it, or should I get separate scanner/fax/copiers? http://hyper.stanford.edu/HyperNews/get/pcmcia/home.html is the definitive source of information on PC Card anythings under Linux. Combo PC Cards are problematic under Linux, because there is no standard for them in the way that there is a standard for single-function PC Cards. All of them come with drivers for Windows, and most also come with drivers for OS/2 and NT, so Linux compatibility is probably the hardest thing to come by. Apparently, nearly any single-function PC Card modem will work, as will nearly any single-function PC Card Ethernet adapter. However, for combo cards, if the EXACT model is not listed on the above web page, it WILL NOT WORK. One of my friends found this out the hard way when he bought an Ethernet/28.8 modem combo card that was not supported; the painful thing was that the Ethernet/33.6 combo card from the same manufacturer is supported. Check the exact model number against the list on the web page, since there are some Xircom combo models that are supported and some Xircom combo models that are not supported. Somebody else will have to answer the printer question, since I am a student, using the school's laser printers for black and white output, but my home printer is a 9 pin dot-matrix :-) -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null
Re: problems with X in hamm
On 17 Jul, Peter S Galbraith wrote: Carlo U. Segre wrote: I have been trying to install a new machine with the frozen hamm distribution and I am running into 2 problems with X11. 1. No user outside root is allowed to start X11. This may be a configuration issue but I know that this did not happen in Debian 1.3 2. I use a S3 ViRGE video card and I am noticing that I have a corrupted text console screen when I exit X (running it as root, of course). This happens both when I am using the SVGA server and the S3V server. I am about to drop back to 1.3 out of frustration but wondered if these problems are just my incompetence... A similar problem was reported by Stephen Ryan in the debian-devel list. If he figures it out perhaps he can help you. For the first problem, I only know of two things - the fact that /usr/bin/X11/X should be suid root, and the /etc/X11/Xserver should have Console as the second line, to allow anyone logged into the console to start the X server. Anything beyond that, and I'm lost. Sorry. For the second problem, I don't have a fix so much as a workaround, discovered in trying to troubleshoot the cause for my problem - nobody can run the X server, not even root, because it crashes every time with a Bus Error. This appears to be a bug related to some PCI based video cards and certain types of PCI bus controllers. In the course of trying to troubleshoot this, Gregory Stark passed on this neat hack: copied from a post by Gregory in the debian-devel list Incidentally here's a useful hack: in /etc/kbd/default.map bind Spawn_Console to some unused key: control keycode 127 = Spawn_Console in /etc/inittab: kb::kbrequest:restoretextmode -r /var/tmp/restoretext.regs and run: restoretextmode -w /var/tmp/restoretext.regs Then you can press that key combination to restore text mode after an svgalib app, or now X, has messed it up. On a multi-user system /var/tmp might not be a good choice for security though. I haven't tried this little hack for fixing corrupted text mode screens yet, as I'm still rebuilding my system from the effects of the HP recovery disk, which recovered my system by erasing all the partitions; it looks to me like you might need the svgatextmode package to use it. HTH, -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null
Re: drm-1804x
On 19 Jun, Babs wrote: Hiya People, I work for an isp and am fairly new to linux still, but have some experience with recompiling the kernel and module installation and configuration. I'm looking for a drv and/or module for a drm-1804x scsi external. Have not had much luck finding even a hit of info. Hoping someone knows what direction to point me in?? Any info would be just great! What sort of device is this? Who makes it? As long as your controller card is supported, then you shouldn't have much trouble with the devices hanging off it. There are options in the kernel configuration for SCSI disk, SCSI CD-ROM, SCSI tape, and SCSI generic (for CD (re)writables and scanners, among other things). Also, since it's external, you should probably make sure it is powered on when you boot up, so that the controller will recognize it. -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Building Kernel for Sound
On 6 Jun, Doug Thistlethwaite wrote: Thanks for the input George, I have looked all through the bo archive (using dselect) for a package called make-kpkg. I do not see one on my mirror by that name. ^ Try kernel-package instead. The program to run (after you have installed kernel-package) is indeed called make-kpkg. G Kapetanios wrote: Try the make-kpkg package. Then make menuconfig make-kpkg --revision (version nunber) buildpackage in /usr/src/linux and dpkg -i kernel-imagedeb and you are done It is much simpler than the standard procedure George HTH, -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Printing to appletalk printer?
On 26 Apr, Asher Haig wrote: Is it possible to make linux print to an appletalk printer? I see papd can spool from appletalk to an lp printer. Can it work the other way? I want to print to an HP 6MP hooked up via a Jet Direct Plus to 10bT ethernet. It does NOT have a specific IP address assigned to it -- I only have 6 IPs on my subnet and they're all in use. They cost money b/c they're for my ISDN router for internet access -- I don't need the printer on the net, just my internal network. (Although Theoretically I could set up a second subnet and put it on that with the printer and have the printer have an IP on that subnet -- but I'd rather not if I don't have to) What's the best way to go about doing this? http://www.giub.unibe.ch/~eugster/appleprint.html Install the netatalk package, then follow half the directions on this web page. You don't need the part about setting the Linux box up as a print spooler for the Macs, just the part about getting your Linux box to find and use the Appletalk printer. Works great here! -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Help on installing Deb 1.3.1 on Toshiba laptop
On 21 May, Philip Restuccia wrote: Hello. I sent a request for help on this topic a few days ago, but have received no responses as of yet. In case I wasn't clear enough on my needs the first time, I'm trying again. If you saw the first request and just don't know of anything that would help me, then I apologize for the duplication! :-) I'm sorry. I didn't see your question the first time. I do try to read most laptop-related questions. I purchased the Debian 1.3.1 binary/source CD set from LSL several months ago, and am finally getting to the actual installation (been *very* busy). What I would *really* like to be able to do is install Linux on a 1.5GB cartride in my Syquest SyJet removable drive, which I have attached to my Toshiba Satellite 105cs laptop via an Adaptec APA-1460A PCMCIA SCSI adapter, to which is also attached a Teac SCSI 16x CD-ROM drive. My laptop has a 504MB hard disk that has only the single partition it came with on it, and that has only about 170MB left on it; the Win95 stuff on it cannot be removed. The laptop also has 24MB of RAM. The hardware sounds like it should be good; AFAIK everything is supported by Linux. Is there any way possible that I can accomplish the installation on the SyJet? I have no real-mode card and socket services available on the machine, so my SCSI adapter (and thus the SyJet and CD-ROM drive) are invisible to DOS (at least to pure DOS; I can access them in a DOS-box under Win95 but this is obviously no good). Do any of the Linux installation utilies by any chance have built-in access to the PCMCIA cards? Is there any alternative? OUCH! My guess is you may not have gotten any answers because this one is a hard one. Problem number one: booting. In order to boot, you will have to be able to see some of Linux at boot time. Not having any experience with a PCMCIA SCSI card (I'm supposed to be getting a New Media Bus Toaster in a couple of days), I don't know if you'll be able to boot off that drive or not. The BIOS has to be able to see the drive in order to be able to boot off it. Since the PCMCIA card hasn't been initialized yet, it can't be recognized by the BIOS as a valid bootable device. I suspect you'll need something to boot from the internal hard drive and then bootstrap up to the external drive after the PCMCIA drivers are loaded. Perhaps a boot floppy with a kernel and the PCMCIA modules on it might do the trick, but I'm in *way* over my head here. I have no idea how to go about such a thing. Even if you boot from the internal hard drive, I think the kernel wants to mount the swap partition and any other partitions you have *before* the PCMCIA drivers are loaded :-( Another major problem is that installation will be tricky. In short, the PCMCIA drivers have to be running *before* the installation program copies anything to the disk; I've installed Debian on 4 laptops in the past 6 months and I've always had to add the PCMCIA support in myself. This could be fixed by adding PCMCIA support and the appropriate SCSI driver to the rescue disk, which I think amounts to solving the same problem as the general booting problem with a floppy disk. Again, I'm in over my head here. Finally, on shutdown, the shutdown stuff wants to unload the PCMCIA drivers before unmounting the drives; this means that every time you boot up, you will at a minimum have to go through a fsck of the whole Syquest drive. The major difficulty is that the PCMCIA drivers have to be loaded before you can get anything off the Syquest drive, and the PCMCIA drivers are only available as kernel modules, and so aren't loaded until too late in the boot process (and correspondingly get unloaded too soon). If you could spare 10-15 megabytes for a bare install on the internal hard drive and could figure out how to a) load the PCMCIA drivers early or b) delay mounting the swap, /home, /usr and /var partitions until later in the boot process, you could probably get away with it. (On my system, I have just a little over 10 megabytes in /boot, /bin, /sbin, /lib, /dev and /etc combined. I have 22 megabytes in /root, which I probably shouldn't have there, and some miscellaneous stuff in /tmp (maybe this should get a partition on the Syquest drive, too?); then there is 1 gigabyte in the /usr partition and 1.5 gigabytes in /home. /var needs some space because that is where new packages get downloaded to before they get installed. Correspondingly, I have a small partition mounted as the root partition, and two very large partitions mounted under /home and /usr, and a medium sized partition mounted under /var. I'm sorry that this comes out so negative sounding; what you've asked is (I think) fairly difficult, and is going to involve changing the boot sequence for anything to work. Also, on an unrelated topic, I've been monitoring this list for a few days now (watching for a response to my initial request, and also attempting to
Re: X11 HELP!
On 21 May, Doug Thistlethwaite wrote: Hello help! I hope someone can help me with this problem that is getting worse and worse... I tired to install fvwm95 and all I got was a plain login screen that would log into a plain window with a single xterm. I tried to remvoe the window manager and install another using dserver... Now I CAN NOT LOGIN to the computer at all from the console. When I log in I get a colorful flash and another plain login screen. The only way I can connect to the computer is via a telnet. How do I disable the X11 so I can get my console back? Also any tips on getting a WM working would be greatly appriciated! I should have a pretty much generic installation, so I'm somewhat confused to why I am having all of this trouble. Are you running XDM (the graphical login screen)? If so, you can press Control-R to kill it and go back to a text console login screen; alternatively, you can login and press F1 after the password instead of Enter to go into safe mode, which doesn't start a window manager, just an xterm so you can fix things; or, you can press Ctrl-Alt-F1 to switch to a text console and login from there (this last one leaves XDM running; you can switch back to it with (usually, if you haven't modified the number of virtual consoles available) Alt-F7). I don't know how to restart XDM after the Control-R bit except (as root) to run /etc/init.d/xdm stop and then /etc/init.d/xdm start - the person I got the tip from didn't say if there was some way to get the XDM screen back or not. If you haven't changed it, the default is to start the first window manager listed in /etc/X11/window-managers. You could try to edit that file to see if it has been corrupted or if the first window manager listed actually exists. You may also want to check to see if there is a .xsession in your home directory, and if so, what is in it. The window manager should be the last line in it if it does exist. Incidentally, the safe mode login is a good way to test new window managers, because you can start a new window manager from the prompt and kill it if it doesn't work out. i.e. you could login in safe mode, type fvwm95 to start the fvwm95 window manager, and either quit it or kill it if you start having problems. HTH, -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Toshiba and HP's GNU/Linux compatibility
On 17 May, Stefan Baums wrote: Hi all! I'll buy a notebook computer shortly, and one of my prime concerns is that GNU/Linux will not only run on it somehow, but actually use the available hardware features. At the moment, I'm considering the Toshiba Satellite 300 CDT and the HP Omnibook 2000 (+ internal CDROM for its accessory bay). To judge from the Linux on Laptops Page (http://www.cs.utexas.edu/users/kharker/linux-laptop/), Toshiba in general seems to be quite compatible with Linux, and HP far less so (though, or maybe because, it has the more interesting hardware). But the exact models I'm interested in have not been reviewed there. So, is anyone out there running GNU/Linux on one of these machines and can tell me about his experiences? Many thanks in advance, Stefan My guess is that the exact models you're interested in are not reviewed there because someone (like me) sat down to try to write a web page about how to use Linux on one of them, and had a great deal of difficulty doing so. Web pages that say Follow the installation instructions. It works. aren't too helpful. Just as an example, I have an Omnibook 5500CT, and installation instructions consist of: Follow Debian install directions. Install the SVGA X server. Recompile the kernel for APM support and Crystal Sound audio support. Beg, borrow, or steal the XF86Config file (I got mine from David Puryear - thanks again David!). Turn off a couple of daemons that like to hit the disk fairly often. Absolutely nothing that doesn't apply to pretty much any laptop. As a laptop-usage data point, there are 9 people in the department here running Linux on a laptop (at least 7 are using Debian, and it's entirely possible that all 9 do); I think that there are 3 Toshiba's, 3 HP's and 3 other. As far as I know, we all have X, sound and power management support. Buy on features / price / durability / service / whatever else meets your fancy. Just make sure you don't get stuck with one of those evil Neomagic chipsets and you should be fine. In case you do go for the Omnibook, I have a (close to useless) web page up at http://www.math.dartmouth.edu/~sryan/omnibook.html which has the XF86Config file available for download; as the OB 2000 (I think) uses the same graphics chipset as the OB5500, it might work. I know it worked on the OB 5000CTS, which is similar but not exactly the same. If you want any other configuration files, feel free to write and ask for them. Sorry I don't have similar information on the Toshiba, but as you remarked, Toshiba does seem to be quite compatible with Linux. HTH, -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Building a network at home
On 21 Apr, Adalberto da Silva wrote: Hello! I wish my question is only a little bit ou of scope... My computer is a dual boot MMX230 with 120 MBytes RAM and three IDE hard disks (2 Megs for Win95 and 11 Megs for Linux - this is an indication for my choices. But I do use Netscape under W95...). My girlfriend is using an old 486DX with W95 and I have an 'spare' P133 (almost complete, it lacks only a video monitor). I'm thinking about put all those guys to talk but I don't know where I may start. Well, I put an eye on Lars' System Administrator.. and Olaf's Network Administrators... but I think my problems are at an ealier stage. Is this possible to put my P133 to make some work for me without a monitor? Is there any major difficulties to put a W95 'slave' machine on a Linux powered network? Is there a good book or manual I can start reading before begin the surgery inside my machines? Is there any suggestion about good and not so expensive network cards? Yes, you can use the P133 without a monitor - though you may want to borrow the monitor from one of the other machines to get it set up the first time. After it is running, you can do most things through telnet. You shouldn't have any trouble using a Win95 'slave' machine; I would probably use a private TCP/IP network (I am using the 192.168.1.x addresses for my home network) and Samba for Windows file sharing. You can also use telnet with your Win95 setups to the headless P133. If you turn on IP Masquerading in the Linux kernel on the P133, you can have that machine be your connection to the Internet for the other two, and have a single connection work for all of them. Best of all, you can do all this with the very inexpensive NE2000 clone network cards. I'm using an inexpensive NE2000 clone card, an IBM Home Away (in my laptop) and an old SMC something-or-other that I got for free in a box of old junk, hooked together with a cheap hub that I got from ONSALE.com. I would suggest looking through the HOWTOs. There is a wealth of information in there (Without actually knowing the contents right now, Firewall-HOWTO, Intranet-Server-HOWTO, SMB-HOWTO, mini/IP-Alias and mini/IP-Masquerade all look like they might have some information that would help you. -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Major Failure re-installing system!!!
On 10 Apr, John Maheu wrote: There is a good signal 11 reference on the web somewhere - unfortunatly I can't remember the address. http://www.bitwizard.nl/sig11/ -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: clean shutdown
On 8 Apr, Kenneth L. Summers wrote in reply to David B. Wilson Is there a way to do a clean shutdown without becoming root? (Preferably still restricting access to those who can push the power button.) The method I tell folks around here to use is to go to a virtual console, press Ctrl-Alt-Delete and watch for the bios screen. Then shut it off. If there is a better way, I'd be interested in hearing it (mostly so I can make sure nobody does it :) ObDisclaimers: I have no idea what, if anything, I just broke. I have modified /etc/inittab and /etc/kbd/default.map.gz. In /etc/kbd/default.map.gz, I have Control-Alt-Insert mapped to KeyboardSignal, and in /etc/inittab, I have added a line (okay, I uncommented one that was already there and modified it) that runs shutdown -h now when it receives a kbsignal. This way, when I press Control-Alt-Insert, it does something almost like Ctrl-Alt-Del, but halts instead of rebooting. On my laptop, or on some machines which support power management, this turns off the power as well. I suppose you could modify the Ctrl-Alt-Del sequence to halt instead of rebooting, but I like the option to choose rebooting or halting. I can provide copies of the relevant parts of those files on request - I have three machines running Debian right now, and I've made this modification on two of them. Guess which of the three has net access :-) -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: 1FA:
On 22 Feb, Hamish Moffatt wrote: On Fri, Feb 20, 1998 at 11:33:03PM -0500, Stephen P. Ryan wrote: On 20 Feb, Bruce Dobrin wrote: 1FA: is what I get when I try to boot from My hard drive. I had this problem once before and the problem mysteriously went away. It hasn't gone away yet on this box. this disturbs me. This is a new machine with a 9gig cheeta drive, 1st (system) partition is 500meg. Any Ideas??? Thank you That is a prompt from your BIOS, wanting to know which partition / drive to boot from. You might get this from pressing a key too soon in the boot sequence, or from not having a proper boot sector installed. If you are using LILO, check to make sure that it is installed on the boot sector rather than in the partition. Actually it's a prompt from the package mbr, not your BIOS. The effect is the same though. hamish I stand corrected. I thought it was from the BIOS because I also thought I got that message from one of my machines before I installed Linux on it. I just checked my motherboard manual and found no trace of this; I just checked /usr/doc/mbr and read all about it. Learning something new everyday, -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: 1FA:
On 20 Feb, Bruce Dobrin wrote: 1FA: is what I get when I try to boot from My hard drive. I had this problem once before and the problem mysteriously went away. It hasn't gone away yet on this box. this disturbs me. This is a new machine with a 9gig cheeta drive, 1st (system) partition is 500meg. Any Ideas??? Thank you That is a prompt from your BIOS, wanting to know which partition / drive to boot from. You might get this from pressing a key too soon in the boot sequence, or from not having a proper boot sector installed. If you are using LILO, check to make sure that it is installed on the boot sector rather than in the partition. -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: [OFF-TOPIC] Pentium II performance?
On 16 Feb, Marcelo E. Magallon wrote: Hi, I know this is off topic, but I don't have access to cola (and newsgroups in general) and I feel more confortable asking here, because I want Linux specific answers. Ten days ago a professor here bought a Pentium II/233 system. He promptly installed Debian on it, and let me use it for my (thesis) work. First thing I did was to benchmark the thing using a program of my own. This program says a Pentium MMX/166 (my old pc) gives about 24 Mflop/s. A Pentium/100 is about 16 Mflop/s. If the numbers are accurate or not, is not in dispute now. What's important is the relative speed, and I find the numbers quoted to be reasonable. (For those curious, it's 3 sums, 3 multiplications, 1 division) The PII says 39 Mflop/s. Over the weekend, I bargained a PMMX/233, which says 33 Mflops/s. I don't find this this reasonable at all! Taking as a reference the performance leap from a 486DX4 - Pentium (same clock speed) I was kinda hoping something near 80 Mflop/s for the PII (yes, I know, it's silly to take that as a reference, but one can only hope) I know I'm not playing fair comparing the systems this way (different kernels, memory, chipset, ...) but I was hoping somebody could give better statistics on this. I'd really appreciate if somebody can help me on this one. We are planing to build a Debian-based compute farm, and the cost difference between PII's and plain Pentium's could translate into a big difference in the number of hosts installed. Side note: K5/133 = 9; K6/200 = 31; 486/66 = 4; RISC 9000 = 18; VAX 3000... oops, forgot about it, but it was surprisingly low. TIA, Which compiler were you using? Programs not specifically optimized for the PPro/PII don't get nearly the performance gain that they could. My experience is something between a 30 and a 50% performance gain possible from using code compiled and optimized specifically for a PPro. I haven't had time to try it yet, but gcc 2.8.0 compiles for PPro/PII, as does egcs (and I think, therefore, pgcc, as that is based on egcs(?) - but its been a while since I checked either of these). -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Hello
On 16 Feb, Ben Pfaff wrote in response to someone else who wrote: I would like to know how debian handles in a machine with two pentiums. = Are two cpus quicker then one in any situation. I mean does every = program run quicker. Generally any particular process will only run on one processor at a time, so particular programs won't run faster. However, two programs can run simultaneously each with a whole processor to themselves, in theory anyway. Anyone with any experience in this want to comment? There are ups and downs to the SMP support in Linux; I have even heard of situations where running two large memory-intensive applications at the same time ran slower than one-at-a-time on a single processor machine. That said, though, I find that the X server runs on one processor, and the application runs on another, leading to a *very* responsive system. You can have an application computing at full speed on one processor, while the screen is being updated and you are movinng windows around etc. on the other processor. Multi-threaded applications run faster, of course, because they get more CPU time; several single-threaded applications running simultaneously depends on the applications and whether they are compute-bound or I/O bound. If the applications are I/O bound, then no amount of extra CPU power will speed them up. If your applications are CPU-bound, then you should get nearly twice the performance from having two CPUs installed. If you are running only one application at a time then the only gain you will see comes from having the background daemons and the X server running on a different CPU from the one running your application, and I think you'd be better off just getting a faster video card or more RAM or a faster drive or just about anything else. Be warned - the extra speed is addictive. The only reason I'm giving up my dual PPro 200 is to get a faster multi-cpu system. -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: [OFF-TOPIC] Pentium II performance?
On 17 Feb, Marcelo E. Magallon wrote: On 16 Feb 1998, Ben Pfaff wrote: Can you point me to the source code for the benchmark? I can run it on my PII/233 for comparison if you want. and Alex Yukhimets: Same here, only with PII/300. You can find the source code here: http://www.efis.ucr.ac.cr/~mmagallo/flops_p.c This is extracted from the flops program written by Al Burto ([EMAIL PROTECTED]). This is just module 1, using 7 sums, 6 mults, 1 div. I think I got this from http://www.hensa.ac.uk/parallel/ or netlib. Compiling with gcc -O2 gives a bit more than 39 on the PII. Using egcc (1.0.1-0.3) with -mpentiumpro gives almost 40.9 Many thanks in advance. Marcelo All right, if I was your professor, I'd ask for my money back. I have a dual PPro 200, and I get roughly 67, from gcc 2.7.2.3 and gcc 2.8.0. gcc 2.8 didn't help nearly as much as I was hoping it would. I also have access to a PII/266, and get roughly 81 from that. Both of these machines were running the distributed.net client in the background, in case that matters. For the person who was asking about SGI's, we have an SGI Origin here (one of the little, 4x180 Mhz processor ones), and get 72 from that, using just -O2 for optimizations. All of these numbers are per processor, as I've only tested one process at a time. I've spent $2800 on the dual PPro, $3500 if you count the peripherals I brought over from my old 486. I have no idea what kind of obscene amounts of money the SGI Origin cost. You may draw your own conclusions about cost effectiveness :-) -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: dual mode?
On 6 Feb, [EMAIL PROTECTED] wrote: I've been eyeing some multi-processor motherboards. After checking several web sites, I've come across a factiod that only Intel supports dual mode. So, what is dual mode? And does one need dual mode to run Debian under SMP kernel option? Or can one use non-Intel processors? I think this refers to the fact that these motherboards will work with either one or two CPUs installed, but only when using Intel CPUs. When using non-Intel CPUs, only one will work. This is due to the fact that Intel has some proprietary interest in the multiprocessor specification and refuses to license it to competitors; my guess is that something in the spec is patented, because the competitors haven't bothered to reverse engineer it, instead coming up with their own specification for multi-CPU support. Unfortunately, I have yet to hear of a motherboard (or chipset!) that supports the other spec. Bottom line: if you want to run multiple CPUs, you need to use Intel. I do not know if Linux supports the other multiple CPU hardware spec, but since there are no products available, the support would be purely academic anyway. As for what actually does work with Linux, anything that says Intel MultiProcessor Spec v1.1 or v1.4 should work. There is a web page at http://www.uk.linux.org/SMP/title.html which explains what does work. -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Installing Debian on Laptop
On 29 Jan, David B. Miles wrote: New Debian User! I have a Compaq Elite 4/75 CX laptop. I have an IC ethernet card (network card). I installed Debian on the laptop using disks. I would like to make a network connection through the network card and then install the rest of the debian files through the network. Ok. You also need to install (still using disks) the pcmcia packages. You will need pcmcia-cs and pcmcia-modules to match your installed kernel version. Once you have installed these, it should then recognize your network card and let you connect to the network. Since we have dhcp service here, I also installed dhcpcd when I did this not too long ago. This should give you enough network service to be able to use dselect's ftp method to install the rest (or the NFS method, but I didn't have an NFS server with Debian available when I did it, so I just used the ftp method.) After you install everything, you should compile a custom kernel (as usual); install the pcmcia-source package and compile a custom version of the pcmcia modules to match your new kernel. You should also check out the Linux laptop home page at http://www.cs.utexas.edu/users/kharker/linux-laptop/ which has more information about running Linux on a laptop than you wanted to know. It appears as though there is in fact a link to a page dedicated to the Compaq Elite series there. I found the Linux Laptop Home Page to be invaluable in getting my laptop started and tuned under Linux. HTH, -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Debian max file size is 1GB instead of 2GB?
On 24 Jan, Steve Hsieh wrote: Sorry if my previous post on this made it out as well. It seems that on Debian, the maximum single file size on ext2fs is 1GB and not 2GB. Can someone confirm this, and suggest how to fix the problem, if possible? I no longer have the space to test this, but I did have a 2GB file in a 2.5GB partition on a Debian 1.3 system several months ago. -- Stephen Ryan Debian GNU/Linux Mathematics graduate student, Dartmouth College -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
* Formal call for the retention of Bruce Perens *
--- dc wrote: Enough is enough. I agree. I'm sick and tired of debian-user being hijacked by whining dissidents. It is time for Bruce Perens to step down form the leadership position, and an interim leader take his place until a new leader can be chosen at the formal election. whining about Bruce deleted yawn Not this again. Stop pickin' on Bruce, willya? Last time I checked, Debian was the product of a huge number of volunteers; larger, in fact, than even the Debian group. Debian is a product of all the Debian volunteers, plus all of the people who wrote the Linux kernel, plus all the people who wrote the software programs packaged by the Debian volunteers, plus the people at the Free Software Foundation who wrote the C compiler and emacs and all the GNU utilities, plus the people from The XFree86 Project, Inc., plus the X Consortium, plus Donald Knuth who wrote TeX and Leslie Lamport who wrote LaTeX, plus ad infininausem. All of these people have allowed *you* to use their work, in freedom. Go read the GPL. Read it very carefully. On second thought, go get someone literate to read it to you. Find out what free means. Hint: free does NOT mean no charge. It means freedom. Think about that for a while. Then go quit bellyaching and start your own distribution. You can even base it on Debian if you like. Bruce can't stop you. That's what free means for you. On the other hand, you aren't allowed to go spoil the party for the hard-working developers. That's what free means for them. Much (private) discussion has gone on about forming a new distribution based on how the old Debian used to be. This is my last attempt at my inclusion and 'fixing' the Debian distribution, and if it does not succeed I (and others with discontent) will turn my efforts to that. sarcasm Please do. I think it would be a really good idea if you created your own distribution. You could even give it whatever version number you like! And just so you don't feel left out, I'll complain about *your* version number for you, right away. I think it should be i, the imaginary number, to represent your own contribution to Debian. Oh, and I think you should base it on Debian 1.0, too. That one's not changing too fast for you, is it? Maybe you should try 0.96r3 then. /sarcasm Look, if you want to express your discontent with a piece of free software, the correct way to do it is in code, not in calling for others to stop. Personally, I'd like to thank Bruce and the rest of the Debian team for an excellent product. We just installed the seventh Debian machine in the department today, right here in the middle of Mac-mania. Stephen Ryan Debian GNU/Linux 1.3 Mathematics graduate student, Dartmouth College who is installing a mail filter next week so he doesn't have to hear any more of this nonsense -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: hamm mirror
--- Timothy Phan wrote: I've been mirroring the ftp.debian.org on the bo/ and bo-updates/ for months now. The activities on this version seems to slow down quite a bit. Now, I'm thinking to mirror the hamm/ as well. Would someone tell me approximately how much more disk space do I need for the hamm/ mirror (giving that I've mirror the whole bo/ and bo-updates!) Thanks in advance! --- end of quote --- I've been trying to run a mirror of bo/, bo-updates/ and hamm/ for people in the department here. (We have 5 people running Debian in the math department now, right in the middle of the second-largest Mac network in the world! One more will start soon and I'm trying to convert another.) I split off a 1Gb partition on my hard drive for it, but it is nearly full now, and hamm has been growing. Figure on at least 1/2 Gb for hamm/ if you strip out all the non-x86 binaries and the source. I don't know how much for a full mirror, because my PC is out of disk space. Stephen Ryan Debian GNU/Linux 1.3 Mathematics graduate student, Dartmouth College -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: netatalk question...
--- Jason Costomiris wrote: Macs (using netatalk, of course) can file just fine, but can't print. What am I missing? --- end of quote --- I haven't tried this, as I'm sitting in the middle of more Macs than I care to think about, and I need to print to AppleTalk printers (the reverse problem, which I have successfully done). There is, however, a sort of a HOWTO on this, which I found at http://artoo.hitchcock.org/~flowerpt/projects/linux-netatalk/ghostscript.html The directions there are for using Ghostscript with netatalk, but I believe the directions there should work with any printer defined in your /etc/printcap. HTH, Stephen Ryan Debian GNU/Linux 1.3 Mathematics graduate student, Dartmouth College -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Multiple network cards. ARGH! Double ARGH! ;-)
--- Alair Pereira do Lago wrote: Perhaps, the driver must be compiled inside the kernel, since some drivers have their parameters analised in boot time but not in loading module time. If there is some PNP card, you may have problems. --- end of quote --- I think you can put the parameters in the /etc/modules file. I'm guessing, though, since I have no parameters in there and it works ok. fx: sound of large piece of wood being hammered, hard I did have to compile NE2000 support as a module, though; I never did get it to work compiled into the kernel. Stephen Ryan Debian GNU/Linux 1.3 Mathematics graduate student, Dartmouth College -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: 2 CPU machines
--- Jesse Goldman wrote: I've just put Debian 2.0.30 with version 6 of the deadlock patch on a new 2 CPU machine. The messages file tells me that the two CPUs were initialized OK and that the machine is running roughly at 400 MHz. If I run bogomips, however, I still get out 198 MHz. I'm guessing that bogomips is not designed for 2 CPU machines. Is that right and, if so, how would I find out on the fly how fast the machine really is going? Thanks --- end of quote --- You are getting the correct results. BogoMIPS is the result of a single timing loop, which will only execute on one processor. The other number is actually the sum of the BogoMIPS numbers for each processor. Try cat /proc/cpuinfo to see info on both CPU's. There is also a kernel patch at http://www-isia.cma.fr/~forissie/smp_kernel_patch/ which will allow patched versions of top and xosview (and maybe other programs, if they support this) to show you what each CPU is doing, and how fast. This will also provide additional information on each CPU in the /proc hierarchy. However, the 2.0.x kernels have a known bug which makes it very difficult to find out which CPU is doing what (the second CPU is reported as all system time, and no user time). You can also use top, which will report the sum of the two cpu's efforts. See the recent thread on top doesn't understand multiple processors for more info. Stephen Ryan Debian GNU/Linux 1.3 Mathematics graduate student, Dartmouth College -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: top doesn't understand multiple processors
--- You wrote: When running top on my 2 cpu machine, I get some interesting states: 12:12pm up 18:03h, 5 users, load average: 1.52, 0.76, 0.49 65 processes: 61 sleeping, 4 running, 0 zombie, 0 stopped CPU states: 90.2% user, 107.5% system, 0.9% nice, 0.0% idle Mem: 128044K av, 110520K used, 17524K free, 37088K shrd, 42576K buff Swap: 129020K av, 0K used, 129020K free 30244K cached The total is approximately 200%, which makes sense. Is there anything to be done about this? --- end of quote --- http://www-isia.cma.fr/~forissie/smp_kernel_patch/ has a kernel patch and some patched utilities (top and xosview) to make use of the new information. top will then show separate lines for each cpu. I was using this before, but I've upgraded to 2.0.31-pre-9 haven't repatched my kernel source. It did work the last time I tried it. Otherwise, my top display looks remarkably similar to yours. As far as I can tell, it doesn't really cause any problems, except that it looks weird. The fact that one CPU is reported as system time only, and not as user time, is a known bug in the 2.0.x kernels, fixed in 2.1.x. As far as your earlier message about what each CPU is being used for, I have no idea how to track that down, though possibly you may be able to figure something out with the kernel patch from the above web site. HTH Stephen Ryan Debian GNU/Linux 1.3 Mathematics graduate student, Dartmouth College -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: graphics accelarator cards ?
--- Marc Fleureck wrote: Does somebody know what cards are compatible and give good performance ? Is e.g. Matrox a good choice (say Matrox Mystique or Matrox Millenium) ? What's the difference between Matrox Mystique and Matrox Mystique ii ? --- end of quote --- I think Matrox is a good choice, though you do have to be careful. Current support is excellent for the Millenium and the Mystique, alpha / early-beta level for the Millenium II, and this is the first I've heard of the Mystique II. You need XFree86 3.3 for the Millenium or the Mystique and XFree86 3.3.1 for the beta driver for the Millenium II. The Matrox / XFree86 home page at http://www.bf.rmit.edu.au/~ajv/xf86-matrox.html leads me to believe that you would be better served by a Millenium or Mystique if you need this now, as the support for the newer cards will be a while in coming. That also indicates that newer Millenium I's have problems (revision 3), so caveat emptor there, too. The Millenium is a faster card, with better expandability, if all you are going to do is Linux and X Windows. If you also want to do dos games or Win95 games, get the Mystique instead, as it has better 3D support. http://www.matrox.com/mgaweb/home.htm is the Matrox web page, and they say that they just cut prices on their whole product line, which should be good news. I don't have a Mystique, but I do know that the Millenium driver is the current speed king in X servers (1 million x-stones on xbench!) I've been very happy with my Millenium and recommend it strongly. Stephen Ryan Debian GNU/Linux 1.3 Mathematics graduate student, Dartmouth College -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Proposal and questions about chemistry Debian packages
--- Frits Daalmans wrote: ...have recently obtained a new program for molecular modelling. It was produced under the GPL and (very well) written in C (IMHO). ... - If I contact the author of this 'moldy' program, Keith Refson, and ask his permission to compile and package it for Linux (only i386 for now), in which Debian mailinglists must I discuss this proposal to get it accepted? --- end of quote --- It appears that others have taken up your other question(s), but not this one, so I'll take a crack at it. Probably the best source of information on this subject is at http://www.debian.org/doc/FAQ/debian-faq-14.html which answers the question, How can I become a Debian software developer? Since Debian is a volunteer effort, the packages included in Debian are mostly the ones that somebody wanted badly enough to go ahead and package them. If this 'moldy' program is the one you were referring to as being under GPL, there will be no problem - your package will probably go into the main distribution. Otherwise, the program may have to go into either contrib or non-free, depending on how it has been released. In any case, the official policy is that development of Debian is open to all, and new users with the right skills and/or willingness to learn are needed..., so I would encourage you to go ahead and get started on making a package. I'm not in chemistry myself, so I don't know if this program will do it, but it might be just the right thing to convince my father to try Debian - I know he's been trying out molecular modeling programs for teaching purposes for a while, and if the software was free, that would help a lot towards getting the tightwads in charge of his university's budget to approve it :-) Stephen Ryan Debian GNU/Linux 1.3.1 Mathematics graduate student, Dartmouth College -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Video card
--- Karsten Bolding wrote: Hi Does anybody know if this card woks with XFree. Matrox MGA PowerDocEd 2MB PCI grafikkort --- end of quote --- Yes, it should. That's a Matrox Millenium PowerDoc Edition, which is supported by the SVGA server in XFree86 3.3 Stephen Ryan Debian GNU/Linux 1.3.1 Mathematics graduate student, Dartmouth College -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: I really need some help
--- [EMAIL PROTECTED] wrote: The following files are corupt debmake.list mgetty-docs.list mgetty-fax.list mgetty.list mount.list xmix.list These files reside in the /var/lib/dpkg/info directory could someone or a few poeple even compress those for me and send them this way ;-) If they can't be sent, could someone point me in the direction of repairing these files dpkg simply will not work if it can't read the data in files under info --- end of quote --- Sorry, I missed your message before; then again, I probably wouldn't have known what to do about it until late yesterday anyway. I'm not going to send you copies of the files, mostly since I think I only have one of the files you are missing installed on my system. However, I ran into the same sort of problem trying to remove the DES-Solnet package. I had removed some of the package by hand previously and dpkg complained that the install was broken and wouldn't remove the rest of it. I eventually just deleted the .list file for that package (and the .prerm and .postrm files, too, I think), then reinstalled the offending package. dpkg fussed about the lack of a .list file, but did go ahead and install the replacement package, which I could then remove easily. I think you could probably just delete (or move) the .list files you mentioned, along with the .prerm and .postrm files, then reinstall the packages in question. This would put fresh copies of those files back into /var/lib/dpkg/info (Fingers crossed, knocking on wood). This may or may not work through dselect, depending on whether or not the version number of the package you are installing is newer than the one it thinks is installed. In any case, the .list file is just a plain text file with a list of filenames belonging to that package, and so should be editable with a plain text editor (vi, emacs, etc.), but it sounds like the corruption is so bad that just getting a new copy would be easier. HTH and good luck. Stephen Ryan Debian GNU/Linux 1.3! Mathematics graduate student, Dartmouth College -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Debian 1.3 install enters a reboot cycle
--- Peter S Galbraith wrote: However, I don't think your 2940U is the problem. The no_reset parameter is to prevent the HBA from resetting the SCSI bus, which it does by default to allow all the SCSI devices on the bus to initialise. Then I don't need the `linux' label anyway. ... The last line I see during kernel bootup refers to IDE-1 or IDE-2 and then it reboots. The PC is a Compaq Deskpro 6000: PPro 200, 64 MB Adaptec 2940U SCSI (IRQ 11) w/ one 4.3 GB disk attached. ATAPI PD-CD (The CD shows up as /dev/hdc on Slackware) Matrox Millenium graphics card, IRQ11 SMC EtherPower 10/100 (Tulip-based) Ethernet card. PS/2 mouse. Any help greatly appreciated. I've been using Slackware since 1994 (and SLS before that), and was looking forward to trying out Debian. --- end of quote --- I had this sort of thing happen to me with a now-defunct PC. It had an Adaptec 2840 SCSI controller in it, and the stock Debian install disks never worked on that machine. The Adaptec controller doesn't appreciate being probed by all of the SCSI drivers in the standard kernel, AFAIK. I was using mine for a CDROM, so I just pulled it, installed without it, and put it back in after I compiled a custom kernel. It did work fine once I had installed a custom kernel. Try the suggestion to build a custom kernel and put that on the install disk. I no longer have a working machine that can use that Adaptec controller, so I've since misplaced the instructions on how to replace the kernel on the install disk, but I think such instructions are available on the Debian web page. An install disk kernel needs to have ram disk support and a couple of other things compiled in, so you probably want to check those before building a custom kernel. There also used to be alternate boot disks (at least in Debian 1.1), one of which was for Adaptec users, but I haven't looked for or needed one in a while. HTH Stephen Ryan Debian GNU/Linux 1.3 Mathematics graduate student, Darmtouth College -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Boot linux from two linux partition
--- [EMAIL PROTECTED] wrote: Sorry for this silly question. I am new to both Debian and Linux. My disk has following partions: /dev/hda1DOS /dev/hda2extended partion /dev/hda3swap /dev/hda4Red Hat linux /dev/hda5Debian 1.3 By installing debian, my computer starts from /dev/hda2 and boot /dev/hda5. Can someone tell me how to setup lilo so that I can also /dev/hda4 ? I know there are doc about it, but I can not find them. :( Here is /etc/lilo.conf - boot=/dev/hda2 root=/dev/hda5 compact install=/boot/boot.b map=/boot/map vga=normal image=/vmlinuz label=Debian Linux read-only other=/dev/hda1 label=Windows 95 --- end of quote --- Hmmm. I'm not exactly new to Linux, though I feel like it sometimes, and I had trouble with almost this exact problem. I was trying to do Debian 1.2 and Debian 1.3 rather than Red Hat and Debian 1.3, and I was using separate physical drives, but I think the problem is the same. The only solution I've found (I would be *very* happy to hear from someone who had a better solution) was to install LILO twice. The first copy of LILO is almost exactly the same as your setup above, with the addition of other=/dev/hda4 label=Red Hat Linux loader=/boot/chain.b The second copy of LILO is installed in the boot block of /dev/hda4 (NOT in the MBR), and just boots by default into whatever setup that copy of Linux uses. Since I experiment with new kernel versions from time to time (part of the hazards of owning a dual processor machine), I also like to keep two kernel versions available in /etc/lilo.conf, so that there is always an old, working kernel available, no matter what sort of disaster^H^H^H^H^H^H^H^H jdevelopment kernel I try. That complicates the scenario, but not by much (just an additional entry in each lilo.conf). email if you'd like copies of my two lilo.conf files. Stephen Ryan Debian GNU/Linux 1.3 Mathematics graduate student, Dartmouth College -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Has anyone used multi- cpu's on debian
--- [EMAIL PROTECTED] wrote: with the cost of multiple cpu motherboards going down i was wondering if anyone has created a super debian system? allan ps what apps are you runnning? --- end of quoted material --- Yes, I have. 2xPPro 200/256k on a Tyan S1662D, 128M RAM. I pulled the hard drives from my old 486 and plugged them into the new system - presto! instant power! I did have a little trouble because I put the two hard drives on separate busses, so my Linux drive moved from hdb to hdc, but that was a detail. You do need to recompile the kernel after uncommenting the SMP=1 line at the top of Makefile in /usr/src/linux/ (use make-kpkg for this) and reboot with the new kernel, and you need to remove all references to setserial from /etc/init.d. Otherwise, no more trouble than installing Debian on a single processor box, which is to say, easy. The usual caveats with quality hardware go double with a multiple CPU setup, but that's important for a trouble-free setup under any circumstances. One thing I have noticed is that the X server will run on 1 CPU, while your application will run on the other. This helps everything. Programs which I know aren't multithreaded still go over 100% CPU usage when the X Server is involved. For two CPU bound tasks (the RSA datasecurity challenge program is a prime example), top reports 99.9% CPU usage for one process and 97.9% CPU usage for the other. I'm not going to complain about that :-). For two memory intensive programs the efficiency drops quite a bit, but is still faster than a 2 processor Sun Sparc Ultra (also at 200Mhz) with the same program. My understanding is that Pentium based multi-CPU boards lose even more due to the shared cache vs. the on chip cache of the PPro. Bottom line: efficiency varies, and how good it works out for you depends a lot on what you want to do with it. Try before you buy, with your applications, if you can. Personally, I just got this box a couple of weeks ago and haven't really done much with it yet. I've been busy teaching this term and haven't had time to play. However, I do have some research problems to work on, which is why I bought this box in the first place; I'll be using custom-written software for that. Oh, and BTW, it is *real* nice to be able to run TeX at light speed. I haven't really found any Debian-specific issues with a multi-CPU setup; they seem to be Linux issues, and the linux-smp mailing list (send mail to [EMAIL PROTECTED] with subscribe linux-smp, without quotes, in the body of the message to subscribe) is a good place to get many of those questions answered. Good luck, and feel free to ask for more help if you do decide to go SMP. Stephen Ryan Team-OS/2 Debian GNU/Linux Mathematics graduate student, Dartmouth College
