Received: from SOUTH-STATION-ANNEX.MIT.EDU by po7.MIT.EDU (5.61/4.7) id AA13167; Sat, 9 Dec 95 23:20:14 EST Received: from SENATOR-BEDFELLOW.MIT.EDU by MIT.EDU with SMTP id AA03913; Sat, 9 Dec 95 23:20:19 EST Received: (from root@localhost) by senator-bedfellow.MIT.EDU (8.6.12/2.3JIK) id XAA03610 for Linux-Development-System-Dist@senator-bedfellow.mit.edu; Sat, 9 Dec 1995 23:17:02 -0500 Message-Id: <199512100417.XAA03610@senator-bedfellow.MIT.EDU> From: Digestifier To: Linux-Development-System@senator-bedfellow.MIT.EDU Reply-To: Linux-Development-System@senator-bedfellow.MIT.EDU Date: Sat, 9 Dec 95 23:17:00 EST Subject: Linux-Development-System Digest #101 Linux-Development-System Digest #101, Volume #2 Sat, 9 Dec 95 23:17:00 EST Contents: Re: Standardized Printing for Linux (Alex Butcher) Re: net balancing (root) Re: Library version number bugs (Mark Shadley) Re: tpqic02: LAST call for help (Unix Consulting) Library version number bugs (Patrick J. Volkerding) Re: Threads - a better model? ("Michael T. Peterson") Re: Weak Thread Support: Linux Weakness ? ("Michael T. Peterson") Re: How to make HD spin down? (Mitch Davis) What shared libs are loaded? (Cris J. Holdorph) Re: System does not reboot (Doug Sisk) Kernel Panics?? Help!!! (Donald) Programmer needed for ISDN Driver COMMERCIAL (Erik Petersen) Re: Where the offical place for libc? (Joe Buck) Re: Swap-space on diskless client (Rob Janssen) Re: Where to learn system-level C (Rob Janssen) Re: Motif development libraries for linux, correction :) (Arnt Gulbrandsen) Re: Worth upgrading to Slack 3.x?? (Christoph Lameter) ---------------------------------------------------------------------------- From: Alex.Butcher@bris.ac.uk (Alex Butcher) Subject: Re: Standardized Printing for Linux Date: Tue, 5 Dec 1995 09:28:30 GMT Hiya In article , albert@snowdon.ccs.neu.edu says... >OK, here is the request: print a man page on an inkjet printer. > >You could print it as ASCII, but you lose the bold, underline, etc. >(or whatever, I think troff has structure...) It looks poor. > >You could convert to PostScript, then convert to a bitmap image, >then s-l-o-w-l-y send the image to the printer. You may run out >of printer memory and crash the printer. While you do this, >you get to test the speed of your swap partition. > >You could use the nice formatting escape sequences provided by >the printer manufacturer for exactly this purpose. Unfortunately, >there is no printer API so all programs must assume ASCII or >PostScript only. You can't use this option with Linux. > Well, what the Amiga used to do was have two APIs for its' printer drivers - one was ANSI control codes + ASCII for text and the other was a colour/mono raster API. The drivers simply sucked in stuff in both of these formats and sent the appropriate printer control codes via the parallel.device or serial.device (the equivalents of /dev/lpx and /dev/ttySx respectively). It seems [Post|Ghost]Script is a better solution for graphics because at least if a printer supports PS already, that part of the equation can just be removed. PostScript is also far more flexible than rasters and easier for small apps to generate automagically. But what we could do with is some sort of filter (possibly part of MagicFilter) that takes the same sorts of codes as used on the console (yeah?) and converts them to printer-specific codes for bold/underline etc. Good Idea? >-- > >Albert Cahalan >albert@ccs.neu.edu Regards, Alex. -- Alex Butcher - Micro Support Technician Tel +44 (0)117 928 9000 x3038 Computing Service, University of Bristol, Tyndall Ave. Bristol BS8 1UD http://www.cse.bris.ac.uk/~ccajb/Welcome.html #include ------------------------------ From: root Subject: Re: net balancing Date: 7 Dec 1995 08:45:07 GMT I don't believe so. The kernel routingis consulted, and the first available route to the desination is used. Whichever interface is first in the table will get used. Does linux look for alternate routes upon network congestion? I would not think so, absent gated or some other config'able router. If I'm wrong, let me know. -Todd Graham Lewis (whois: tgl2) ------------------------------ From: Mark Shadley Subject: Re: Library version number bugs Date: Sat, 9 Dec 1995 12:32:40 -0800 On 8 Dec 1995, Patrick J. Volkerding wrote: > I've noticed what I think are serious problems in the selection of version > numbers for shared libraries under Linux. Allow me to provide a couple of > examples: > > 1. Linux libc-5.2.16 has been made public. The GNU gdbm and db libraries > are no longer part of this release, and we've been pointed to what are > supposed to be the new versions. However, both of these libraries have > lower version numbers (*.so.1 instead of *.so.2) than the ones in > libc-5.0.9. This effectively breaks all existing apps linked with the > versions from the last public release. I guess I haven't run anything yet that was linked with 5.0.9. > What happened? ELF was supposed to save us from these kinds of > problems, and would have here if only a library version number had been > correctly chosen. Note that the libraries now follow the version number > of the package e.g. gdbm-1.7.3 = libgdbm.so.1.7.3. Previously 2.0.0 > had been used as the so version number... probably a bad idea, but > IMHO it represented some kind of commitment to stick with the numbering > that had been decided upon. I think the reason is that ELF soname numbering hasn't matured into a standard yet. There appears to be harmony for the most part among developers, but perhaps not enough communication? > 2. An even crazier example: The first public release of shared ncurses > libraries was just made on tsx-11.mit.edu in /pub/linux/packages/GCC. > The links for the shared libraries have a totally different version > number than the libraries, e.g.: > > -rw-r--r-- 226650 /usr/lib/libncurses.so.1.9.7a > lrwxrwxrwx 0 /usr/lib/libncurses.so.2.1 -> libncurses.so.1.9.7a > > This is the expected name. If the link is changed to libncurses.so.1.9 > or libncurses.so.1 (the expected name, IMO) it won't work. Resulting > binaries will be expecting to find libncurses.so.2.1 and therefore won't > start. libncurses 1.9.8a was released yesterday, with a soname of 3.0. (This soname is expected to last for a while.) I don't know if this version is compatible with 1.9.7a (a query has been posted). If it is, it will be relatively easy to just make a symlink by hand and live with it until all our old apps have been recompiled. If it isn't (!), we will have to keep 2 libs around until all our old apps are recompiled! :( > In addition, if these links don't exist they are not correctly made by > ldconfig. Lack of communication? > Comments anyone? See above. Mark ------------------------------ From: consult@btoy1.rochester.ny.us (Unix Consulting) Crossposted-To: comp.unix.bsd.freebsd.misc Subject: Re: tpqic02: LAST call for help Date: 9 Dec 1995 18:39:57 -0500 Reply-To: consult@btoy1.rochester.ny.us Rui-Tao Dong ~{6-HpLN~} (rdong@mat.jhu.edu) wrote: : : QIC-02 doesn't seem to be actively supported in Linux. I am thinking about : booting FreeBSD to do my backup. (It is probably a good idea to backup : from a dead file system instead of a live one anyway.) I understand that : FreeBSD supports QIC-02 and it reads Linux Ext2 file system. It is just a : matter of getting the right boot image and some shell utilities. Linux supports QIC-02 very well. You have to rebuild the kernel to install the support and get the run-time config utility. It is at my site: ftp://btoy1-gw.roc.servtech.com/pub/unix/ELF/tpqic02-support-1.6b.tar.gz : ... : Regards, : : Rui-Tao Dong ~{6-HpLN~} 2760 Kelvin Ave., #3211 : rdong@printrak.com Irvine, CA 92714 : (714)752-0855(H) -- ======== Thomas Bullinger ======= Specializing in Linux systems ======== * Installation * Setup * Maintenance * Training * Software * Information * Email: consult@btoy1.rochester.ny.us FAX: +1-716-377-6732 * WWW: http://btoy1-gw.roc.servtech.com/users/consult ------------------------------ From: gonzo@magnet.mednet.net (Patrick J. Volkerding) Subject: Library version number bugs Date: 8 Dec 1995 01:27:58 GMT I've noticed what I think are serious problems in the selection of version numbers for shared libraries under Linux. Allow me to provide a couple of examples: 1. Linux libc-5.2.16 has been made public. The GNU gdbm and db libraries are no longer part of this release, and we've been pointed to what are supposed to be the new versions. However, both of these libraries have lower version numbers (*.so.1 instead of *.so.2) than the ones in libc-5.0.9. This effectively breaks all existing apps linked with the versions from the last public release. What happened? ELF was supposed to save us from these kinds of problems, and would have here if only a library version number had been correctly chosen. Note that the libraries now follow the version number of the package e.g. gdbm-1.7.3 = libgdbm.so.1.7.3. Previously 2.0.0 had been used as the so version number... probably a bad idea, but IMHO it represented some kind of commitment to stick with the numbering that had been decided upon. 2. An even crazier example: The first public release of shared ncurses libraries was just made on tsx-11.mit.edu in /pub/linux/packages/GCC. The links for the shared libraries have a totally different version number than the libraries, e.g.: -rw-r--r-- 226650 /usr/lib/libncurses.so.1.9.7a lrwxrwxrwx 0 /usr/lib/libncurses.so.2.1 -> libncurses.so.1.9.7a This is the expected name. If the link is changed to libncurses.so.1.9 or libncurses.so.1 (the expected name, IMO) it won't work. Resulting binaries will be expecting to find libncurses.so.2.1 and therefore won't start. In addition, if these links don't exist they are not correctly made by ldconfig. Comments anyone? Pat ------------------------------ From: "Michael T. Peterson" Subject: Re: Threads - a better model? Date: 9 Dec 1995 21:34:51 GMT davidsen@tmr.com (bill davidsen) wrote: >Regardless if what you think of the clone() model of doing >threading, I think it would be nice to use at least a name which >doesn't conflict with the standard AT&T implementation of streams, >where clone() duplicates a stream. > >Comments? Agreed! /mtp -- +-----------------------------------------------------------------------+ | Michael T. Peterson | For the latest status on DCE and POSIX threads | | mtp@big.aa.net | for Linux, see http://www.aa.net/~mtp/ | +-----------------------------------------------------------------------+ ------------------------------ From: "Michael T. Peterson" Subject: Re: Weak Thread Support: Linux Weakness ? Date: 9 Dec 1995 21:36:39 GMT In comp.os.linux.development.system, bass@silkroad.com wrote: >The next generation netstuff seems to be moving into the Java >world. Java requires a multithreaded environment. Linux is >very weak in the multithread arena. What's weak about it? /mtp -- +-----------------------------------------------------------------------+ | Michael T. Peterson | For the latest status on DCE and POSIX threads | | mtp@big.aa.net | for Linux, see http://www.aa.net/~mtp/ | +-----------------------------------------------------------------------+ ------------------------------ From: mjd@tpg.tpg.oz.au (Mitch Davis) Subject: Re: How to make HD spin down? Date: Sat, 2 Dec 95 09:32:23 GMT In article <49f2mo$99c@frigate.doc.ic.ac.uk>, Richard Jones wrote: >Carl Schott (cgschott@psu.edu) wrote: >: sp@questor.org wrote: >: : In article <497037$82v@arcadia.informatik.uni-muenchen.de>, >: : Thiemo Gehrke wrote: >: : >Hi Folks, >: : > >: : >I got a tool from a friend, thats makes my hard disc spin down after a >: : >given time (it's for D*S). So I watched out for a similar one for Linux, >: : >cause i didn't know before, how silent my comp can be if it wants to. :-) > >: "hdparm" will do this for IDE drives that support a spindown timeout. >: (Look for the "-S" switch). > >Works great except for my root disk. I _think_ syslogd likes to write to a >file every so often, so negating the effects of the spindown! Anyone know >how to cure this? On many computers, the BIOS can tell the hard disk to do this, as part of the power-on initialisation. I do such a trick on my laptop and it works a treat. The trick is to suspend the "update/bdflush" processes. In this way, old data sits in the RAM cache, and never gets flushed to disk. Hence, the disk never gets accessed (except of course when you read something from it that has never been read before.. :-) This more than doubles the life of my laptop battery. **** BIG FAT WARNING TIME **** Because new data is not written to the disk, if you lose battery power, any work you do won't have been written to the disk. It's not possible to trash your filesystem (since nothing ever gets half-written), but you will lose any work you've done. To do this, log on as root, and use "ps -aux" to find the "bdflush" and "update" processes and their process id's. Then, send them signal 18 (SIGSTOP) using kill -SIGSTOP x y, where x and y are the process IDs. When you want to save all your work to disk on a one-off basis, run the "sync" command (as any user). This will write the buffer cache to the hard disk. When you want to resume the write-every-30-secs-and-never-power-down-but it's-safer-this-way mode, send signal 19 (SIGCONT) to the processes, ie kill -SIGCONT x y. I dunno. This is just such an astoundingly useful trick for laptop owners that I'm surprised it hasn't appeared in the Linux TIPS Howto. Hmm. Might just send this off then..... Hope this has helped you. Mitch. ------------------------------ From: choldorp@piranha11 (Cris J. Holdorph) Subject: What shared libs are loaded? Date: 6 Dec 1995 20:55:42 GMT Is there any way to tell what shared libraries are currently loaded in the system? i.e. I want to know if the a.out librarie is being used by anything. So, I can try to track down and eliminate the a.out programs I still used on a regular basis (replace them with elf versions). ---- Cris J H -- Cris J. Holdorph choldorp@cig.mot.com Motorola Corporation Cellular Infrastructure Group ------------------------------ Crossposted-To: comp.os.linux.misc,comp.os.linux.hardware From: Doug Sisk Subject: Re: System does not reboot Date: Sat, 9 Dec 1995 16:22:47 GMT Shawn Ruttledge wrote: > > smurray@earthlight.co.nz wrote: > : On Sun, 3 Dec 1995, Tom Wheeley wrote: > : > Yes, my system has never reboted itself, and so far I have just pressed > : > the RESET button when it says 'Done' and the hard disk stops crackling. > : > Haven't lost / yet :^) > > : But why? I have my linux machine at (M$ banned zone) home and it hangs on > : reboot (IBM blue lighting 2/66) where as the linux box at work (Intel) is > : fine. Is this just because of the CPU? > > My AMD DX4-100 Award bios system does this too, whereas my old 386-33 > AMI bios system reboots fine. I would like to fix the problem. I had the problem with a micron p90. I turns out that linux will only reboot if the flash eeprom bios was not write-protected. I had to write-protect mine when some winbloze programs kept crashing and taking out my bios at the same time. Doug Sisk mailto:das@ade.att.com ------------------------------ From: donald@terra.igcom.net (Donald) Subject: Kernel Panics?? Help!!! Date: 8 Dec 1995 23:25:47 GMT I always seem to get kernel panics. After having linux up for a few hours, it always seems to freeze with a kernel panic. It says "Kernel Panic: Trying to free up swapper" and "In swapper: task not synching memory space". Before this crash, i usually get tons of General Protection errors, of some sort of command. Hope this helps. If anyone would like to see the whole message, mail me and i'll be please to show you. If anyone knows anything about this, please help me. i hate to reboot my machine every 4 hours. Other Info: I have Linu Kernel 1.2.8, Micron Pentium 100, 16MB RAM, EIDE drives...and mail me if you need more info. Thank you! PLEASE HELP ME SOMONE! If you post, mail the article to me also, as i may not be able to see it. Thank You again! ------------------------------ From: Erik Petersen Subject: Programmer needed for ISDN Driver COMMERCIAL Date: Thu, 07 Dec 1995 11:18:45 -0500 We are currently in need of a programmer to write a Linux driver and config utility for our ISDN BRI adapter. If you are capable and willing please drop me a note with particulars. This is a paying gig unless you want to do it for free :-) Cheers. -- Erik Petersen MediaGlobe Inc. mailto:erik@mediaglobe.com 416-361-7297 http://www.mediaglobe.com ------------------------------ From: jbuck@synopsys.com (Joe Buck) Subject: Re: Where the offical place for libc? Date: 6 Dec 1995 06:19:43 GMT barlow@xserver.sjc.ox.ac.uk (Daniel Barlow) writes: >>Now, what exactly is this? The GNU glibc is up to version 1.0.9, and libg++ >>is up to 2.7.1... Are these just specialized versions for Linux? > >These are specialised versions for linux, yes. The GNU libc shares some code >with the Linux libc, but it's hardly a drop-in replacement. They are >really quite different in some areas (although there is ongoing work to >merge the two threads). > >What relevance does libg++ have in this context? Linux's stdio library is based on the libio portion of libg++, but is modified, which means that FSF libg++ releases won't work on Linux, which is a shame. -- -- Joe Buck (not speaking for Synopsys, Inc) ------------------------------ From: rob@pe1chl.ampr.org (Rob Janssen) Subject: Re: Swap-space on diskless client Reply-To: pe1chl@wab-tis.rabobank.nl Date: Wed, 6 Dec 1995 08:51:59 GMT In <4a21pn$1jd@fbi-news.Informatik.Uni-Dortmund.DE> loebbing@turing.informatik.uni-dortmund.de (Martin Loebbing) writes: >Can anyone tell me how to set up a swap-space for a diskless client? >I'm using the kernel 1.3.45 on a PC served by a Sun workstation. >I tried the following without success. Linux can't swap over the network, or at least I don't think it can. You will need to use a local diskdrive with a swap partition, or install enough memory to run without swapping if you don't have a local disk. Rob -- +------------------------------------+--------------------------------------+ | Rob Janssen rob@knoware.nl | BBS: +31-302870036 (2300-0730 local) | | AMPRnet: rob@pe1chl.ampr.org | AX.25 BBS: PE1CHL@PI8WNO.#UTR.NLD.EU | +------------------------------------+--------------------------------------+ ------------------------------ From: rob@pe1chl.ampr.org (Rob Janssen) Subject: Re: Where to learn system-level C Reply-To: pe1chl@wab-tis.rabobank.nl Date: Wed, 6 Dec 1995 08:55:17 GMT In <4a22j5$8mi@gap.cco.caltech.edu> "Matthew A. Clapp" writes: >Hi, > I'm building a PC-card that when finished, I'd like to use linux to run. >Basically, it just has about 5 registers that I need to read and written to. > My question is, does anyone have suggestions as to the best source >to learn the lower-level system commands needed to either write a user program >to do this, or to actually write a device driver under linux. I already know >C enough to make normal programs that don't need to make system calls. A mini-howto on I/O port access is published regularly, and was published a few days ago. Look for it in the newsgroup. To write a driver you need some more experience and general systems programming knowledge, that you can't expect to find in Linux-specific documents. Look at other drivers that do things that are similar to what you want. But watch out for assumptions about interrupts, dma, memory layout, re-entrancy etc that may kill your system when you don't really know what you are doing... Rob -- +------------------------------------+--------------------------------------+ | Rob Janssen rob@knoware.nl | BBS: +31-302870036 (2300-0730 local) | | AMPRnet: rob@pe1chl.ampr.org | AX.25 BBS: PE1CHL@PI8WNO.#UTR.NLD.EU | +------------------------------------+--------------------------------------+ ------------------------------ From: Arnt Gulbrandsen Crossposted-To: comp.os.linux.development.apps,comp.os.linux.x Subject: Re: Motif development libraries for linux, correction :) Date: 09 Dec 1995 21:13:42 +0100 What I want to know is, why do you people all Motif? It's slow, buggy and bloated: What redeeming feature does it have? --Arnt ------------------------------ From: clameter@miriam.fuller.edu (Christoph Lameter) Subject: Re: Worth upgrading to Slack 3.x?? Date: 9 Dec 1995 13:34:39 -0800 Thanks for responding to my post Patrick. I am really very thankful for the work you have done in pioneering usable distributions of Linux for all of us. On Sat, 9 Dec 1995, Patrick J. Volkerding wrote: > In article <4a8chi$iaf@miriam.fuller.edu> you wrote: > : But by and large I am disappointed by 3.0. There are not many enhancements. I > : am thinking of using Debian in the future. I am always installing a lot of software > : that I get from sunsite or elsewhere to get the features I need. And then there are > : the bugs that are rarely fixed. > > Sorry I disappointed you, but you really have no idea how much work it was to > basically start over from scratch for the ELF format. It may look like much > of the software is the same version, but hundreds of patches were needed to > get it all to compile. I certainly can appreciate the work that it takes to get it there. But you are working in a one man team. With such a complex thing as nowadays distribution I doubt that you can on your own maintain a well functioning distribution. > Which bugs are rarely fixed? I mean, really. I get a little tired of > being the Linux whipping boy and would like some explanation, especially > as I work >12 hours a day on this sometimes. I only have a 14.4K link > to the internet, BTW, so I don't like to twiddle around with the archive > for no reason. But, I think it's well maintained. I just think of two bugs that I encountered last week: 1. crond is started before ypbind. If I now have a cron job of a user whose /etc/passwd info comes in via NIS then cron will stop working completely and try to contact NIS domain "(none)" 2. There is still a reference to "uuq" in the INN package although Slackware's Taylor UUCP does not include such a command. This made remote UUCP for me impossible. Then there is the minicom situation about which I emailed you ages ago. Still no xminicom and termcap contents. I wonder why you have only a 14.4K link? I thought Morse or some company employs you. I am certainly willing to offer you the resources of our Internet Setup (fuller.edu in Pasaden,CA) for any support I can give you. > Sunsite, on the other hand, causes me no end of troubles. They've had a > corrupted set of bootdisks up for FTP for quite a while, and haven't > mirrored my bugfixes in a couple of months. Yes. I know that is why we mirror slackware from ftp.cdrom.com. Debian has just overtaken you. They offer libc-5.2.x and all the most upgraded packages in Debian-1.0. I can upgrade that stuff easily. Bugs are fixed and there are scores of people working on maintaining it. When Debian insisted on staying on old stuff with their 0.93R6 then they could not compete with Slackware. But now... They are already organized better and have their organization running full speed. I doubt that anyone will be able to compete with that distribution once debian-1.0 reaches production stage. They are still lacking some packages but from what I see that is going away faster each day. On the other hand they do offer so many packages that Slackware does not, that I think I will not use Slackware again. I think you have done great work in the past and I think it would be a great benefit if you would join the Debian Project. It would make your life easier and would improve Debian trememdously. -- ============================================================================= Christoph Lameter FTS Box 466, Pasadena, CA 91182 Internet Administrator Who is like Jesus... who is like God...? ------------------------------ ** FOR YOUR REFERENCE ** The service address, to which questions about the list itself and requests to be added to or deleted from it should be directed, is: Internet: Linux-Development-System-Request@NEWS-DIGESTS.MIT.EDU You can send mail to the entire list (and comp.os.linux.development.system) via: Internet: Linux-Development-System@NEWS-DIGESTS.MIT.EDU Linux may be obtained via one of these FTP sites: nic.funet.fi pub/OS/Linux tsx-11.mit.edu pub/linux sunsite.unc.edu pub/Linux End of Linux-Development-System Digest ******************************