Subject: Linux-Development Digest #14
From: Digestifier <Linux-Development-Request@senator-bedfellow.MIT.EDU>
To: Linux-Development@senator-bedfellow.MIT.EDU
Reply-To: Linux-Development@senator-bedfellow.MIT.EDU
Date:     Thu, 11 Aug 94 12:13:09 EDT

Linux-Development Digest #14, Volume #2          Thu, 11 Aug 94 12:13:09 EDT

Contents:
  Re: IRQs attached to what? (Donald Becker)
  Re: Suggestion: Lets have a standard for numeric uid/gids (Matthias Rabe)
  Re: DOSEMU 0.53: Developers and testers needed! (Jonathan Noel Tombs)
  Re: Linux/GPIB interfaces (Bernd Mielke)
  Re: IRQs attached to what? (Anselm Lingnau)
  Taylor-UUCP and LinuX, severe Problems!! (Fred Meyer)
  Any success with FuzzyCLIPS? (Naresh Sharma)
  ISDN unter Linux / ISDN with Linux (Egbert Hinzen)
  Re: Suggestion: Lets have a standard for numeric uid/gids (Harald T. Alvestrand)
  kernel 1.1.41 breaks lilo on floppy (Andre Schroeter)
  Re: Program that demonstrates a problem with getwd (H.Boehme U.Kunitz)
  Setting up SVGALib with ET4000 (Paul M Sargent)
  Re: Database for Linux Summary: Expires: References: <328qef$i27@crl4.crl.com> Sender: Followup-To: Distribution: Organization: WorkGroup Solutions (FlagShip) & GCS, Inc.  Keywords: Cc:  (Mark Bolzern)
  Typo in kernel 1.1.43 (Richard Lindner)
  Linux/GPIB interfaces
  Re: Is SONY CDU31A driver in kernel 1.1.4x robust?? (Allan Clearwaters)
  FS-recovery, searching for superblocks
  IPX on Linux (Robert Herrmann)
  More nfs problems (Peter Choynowski)

----------------------------------------------------------------------------

From: becker@cesdis.gsfc.nasa.gov (Donald Becker)
Subject: Re: IRQs attached to what?
Date: 10 Aug 1994 14:20:04 -0400

In article <328j1b$6tj@smurf.noris.de>,
Matthias Urlichs <urlichs@smurf.noris.de> wrote:
>In comp.os.linux.development, article <326aob$f5r@cesdis1.gsfc.nasa.gov>,
>  becker@cesdis.gsfc.nasa.gov (Donald Becker) writes:
>> 
>>    if (irqaction(irq,...) ... 
>>        irq2dev_map[irq] = dev; 
>>        irq2name_map[irq] = "Frobbitz SuperHyper 6000"; 
>
>Hmm. Better to have a diferent call, along the lines of register_XXXdev.

Well, just writing a table takes less code.  That's important since this
feature is "bloat".

>> prompt% cat /proc/net/irq
>> IRQ  Count   Flags   Device
>>  0      79879            0   Timer
>
>/proc/irq, please, and the title line doesn't belong in the kernel.

Ooops, yes.  My fingers can't type "/proc" without changing it to
"/proc/net".

Why don't you think the title line belongs there?
(That's a real question, not just flamebait.)


-- 
Donald Becker                                     becker@cesdis.gsfc.nasa.gov
USRA-CESDIS, Center of Excellence in Space Data and Information Sciences.
Code 930.5, Goddard Space Flight Center,  Greenbelt, MD.  20771
301-286-0882         http://cesdis.gsfc.nasa.gov/pub/people/becker/whoiam.html

------------------------------

From: rabe@mathematik.uni-bielefeld.de (Matthias Rabe)
Subject: Re: Suggestion: Lets have a standard for numeric uid/gids
Date: Wed, 10 Aug 1994 07:02:41 GMT

In article <root.776460959@olivia.ping.dk>,
Torben Fjerdingstad <tfj@olivia.ping.dk> wrote:
[...]
>I wish there were standards for root,bin,sys,wheel,mail,daemon,
>sync,news,uucp,adm, (or admin?)... and so on.
>
>Earlier i lost my hair trying to make the mail programs share
>/usr/spool/mail. It was also because some of the mail programs
>from a binary linux distribution had a built-in check for the
>numerid id's. They expected the group mail to have id 12, while
>they were 6 or 7 on my system.

Such things cost me hours when I upgraded to MCC 1.0+.
I think, such a standard is really needed.

Matthias

-- 
rabe@mathematik.uni-bielefeld.de                          Matthias Rabe
Universit"at Bielefeld                            Privat: Avenwedder Str. 494
U5-133                                                    D 33335 G"utersloh
Tel.: (0521) 106-3871                                     Tel.: (05209) 6673

------------------------------

From: jon@obelix.cica.es (Jonathan Noel Tombs)
Subject: Re: DOSEMU 0.53: Developers and testers needed!
Date: 10 Aug 1994 17:09:32 +0200

In article <7529@raven.ukc.ac.uk>, Greg Harewood <gjh@ukc.ac.uk> wrote:
>In article <Ctz1Jp.2q9@undergrad.math.uwaterloo.ca>,

>
>>Some of these tasks require very tricky stuff like instruction
>>emulation, kernel modification (to get access to the CPU for protected
>>instructions), and more.  A person named Lutz Mologedy is doing some
>
>Why do you need kernel modifications? I'm no great expert but I thought the
>idea of emulation was that you never needed the real thing - you emulate
>the hardware and the firmware and surely protected mode too.
>
>Isn't it just a case of intercepting the signals for memory violations and
>illegal instructions and then emulating them?

yes, but if you just let the kernel send a SIGILL or SEGV to the dosemu code
then let dosemu examin the instruction to work out why the fault happened
everything goes to slow. This is what happened in dosemu <0.52. when the
handling of a few functions got moved to the kernel things have speeded up
2-4 fold, it is quite likely to get reasonable performance out of dpmi the
same will be true. You could do it in user space, but it is easier and 
much quicker to handle it in the kernel (probably using a loadable dosemu
module). 



Jon.

------------------------------

From: mielke@omega.physik.fu-berlin.de (Bernd Mielke)
Subject: Re: Linux/GPIB interfaces
Date: Thu, 11 Aug 1994 07:50:34 GMT

ceilbeck@dra.hmg.gb () writes:


>Has anyone done any work on a GPIB driver for Linux?  I've been thinking
>about using Linux to control some lab instruments but no-one seems to
>want to write a driver for their GPIB card to work with Linux.  Help!

>Chris
I think Claus Schroeter (clausi@laue.chemie.fu-berlin.de) has written
the whole driver, ask him.

Bernd

mielke@omega.physik.fu-berlin.de

------------------------------

From: lingnau@tm.informatik.uni-frankfurt.de (Anselm Lingnau)
Subject: Re: IRQs attached to what?
Date: 11 Aug 1994 07:54:49 GMT

In article <1994Aug10.215841.545@kf8nh.wariat.org>, bsa@kf8nh.wariat.org
(Brandon S. Allbery) writes:

> Anything beyond just the raw information --- that includes both title lines
> and formatting --- is just bloat.

Random crud like title lines is a nuisance to people trying to parse the
output, too.
-- 
Anselm Lingnau ......................... lingnau@tm.informatik.uni-frankfurt.de
Good judgement comes from experience. Experience comes from bad judgement.
                                                                --- Jim Horning

------------------------------

From: frett@downtown.oche.de (Fred Meyer)
Crossposted-To: comp.mail.uucp
Subject: Taylor-UUCP and LinuX, severe Problems!!
Date: 10 Aug 1994 00:22:16 +0100


Hi!

...i'm running a Usenet/Internet-site, providing mail&news via UUCP.
My system is running under Linux (Kernel 1.09) with Taylor-UUCP (V.104 +
V1.05) installed (taylor-config + HDB-log).

Here comes the description of a strange and ugly problem, i recently
discovered:

  - assuming that there are several elements in the send/recieve-queue, when
    a call arrives, one can recognise the following:

     - the tranfer-rate of the data itself (as reported in the HDB
       "xferstats"-log) is really execellent.
       (i.e.: 14.4K -> 1700CPS, 19.2k -> 2300CPS, ISDN-64k -> 7700CPS,
              TCP on ISDN-64k -> 7550CPS, raw data without online compression)

     - the overall performance is, sometimes, really bad.
       (i.e: 19.2k, 123k transfered in 60 (data-)files -> 420CPS)
       This is because the uucico program performs a sort of "refresh"
       or "pause" between the files...the resulting effect is a bad
       overall CPS-rate, especially when there are lots of (short) files 
       to transfer.

...talking to other people using UUCP-installations (taylor and others)
on different platforms, leads me to the idea, that this is a Linux specific
problem...i've already tested different protocols with packet/window sizes
without a clue.

Taylor V1.04/V1.05 compiles "out_of_the_box" without problems under Linux,
but maybe there is a specific patch i missed to put into the installation?

Any help VERY welcome! Please reply via eMail!

TIA,
   Fred

-- 
Fred Meyer                | Phone: + 49-241-4097551
Kapuzinergraben 8         | eMail: frett@downtown.oche.de
52062 Aachen, Germany     |        fred@milenka.ltt.rwth-aachen.de 

------------------------------

From: nash@dutllu4.gmd.de (Naresh Sharma)
Subject: Any success with FuzzyCLIPS?
Reply-To: Naresh.Sharma@LR.TUDelft.NL
Date: Wed, 10 Aug 1994 07:55:41 GMT

Hi,

Has anyone successfully compiled fuzzyCLIPS under Linux?

Naresh

--
_______________________________________________________________________________
Naresh Sharma [N.Sharma@LR.TUDelft.NL]  Herenpad 28            __|__
Faculty of Aerospace Engineering        2628 AG Delft   \_______(_)_______/
T U Delft               Optimists designed the aeroplane,     !  !  !  
Ph(Work) (+31)15-783992 pessimists designed the parachute!
Ph(Home) (+31)15-569636 Plan:Design Airplanes on Linux the best OS on Earth!
==============================PGP=KEY=AVAILABLE================================

------------------------------

Crossposted-To: dinoco.general
From: egbert@garfy.dinoco.de (Egbert Hinzen)
Subject: ISDN unter Linux / ISDN with Linux
Date: Wed, 10 Aug 1994 19:46:51 GMT

Kennt jemand von Euch ein ISDN Package fuer Linux?
Am liebsten eins, dass die Teles-(Passiv)-Karte unterstuetzt?
...
Is there any ISDN Package for Linux?
Best choice would be a package for non-active-cards.

TIA

--
Egbert Hinzen ...speaking for himself... 
Email: egbert@garfy.dinoco.de (preferred) *** 100113.1011@compuserve.com

------------------------------

From: hta@uninett.no (Harald T. Alvestrand)
Subject: Re: Suggestion: Lets have a standard for numeric uid/gids
Date: 10 Aug 1994 08:53:37 GMT

In article <root.776460959@olivia.ping.dk>, tfj@olivia.ping.dk (Torben Fjerdingstad) writes:
|> Today I compiled at-2.7a on my fastest machine.
|> Then I nfs'et to it from another machine and did an 'make install'.
|> 
.....
|> On the fast machine, the group daemon has gid 2
|> On the other one, it is gid 12.

I think this is a shortcoming of your local environment.
If you are doing NFS mounts without some means of synchronizing your
pw/group files, you are in BIG TROUBLE.

(Just think of the innocent user "sacrifice(42)" on one machine who discovers
that all his work has been destroyed by "evildoer(42)" on another machine
via an NFS mount.....)

-- 
                   Harald Tveit Alvestrand
                Harald.T.Alvestrand@uninett.no
      G=Harald;I=T;S=Alvestrand;O=uninett;P=uninett;C=no
                      +47 73 59 70 94
My son's name is Torbjxrn. The letter between "j" and "r" is o with a slash.

------------------------------

From: andre@borneo.dkfz-heidelberg.de (Andre Schroeter)
Subject: kernel 1.1.41 breaks lilo on floppy
Date: Wed, 10 Aug 94 09:02:36 CET

as with kernel version 1.1.41 lilo can't write the bootsector af a mounted
floppy with a minix filesystem.

  andre schroeter 


-- 
================================================================================
Andre Schroeter                         German Cancer Research Center
tel.: (+49) 6221 - 42 2382              Dept. Medical and Biological Informatics
fax.: (+49) 6221 - 42 2345              Im Neuenheimer Feld 280
e-mail: A.Schroeter@DKFZ-Heidelberg.de  D-69120 Heidelberg, Germany
================================================================================

------------------------------

From: news@informatik.hu-berlin.de (H.Boehme U.Kunitz)
Subject: Re: Program that demonstrates a problem with getwd
Date: 11 Aug 1994 09:55:11 GMT

In article <FOX.94Aug9130908@first.cs.nyu.edu>,
David Fox <fox@graphics.cs.nyu.edu> wrote:
>#include <stdlib.h>
>
>#define MAX_PATH_LENGTH         256
>
Here you assume the Maximum of path length is 256, that is not correct.
in /usr/include/linux/limits.h 
#define PATH_MAX        1024


>char buf[MAX_PATH_LENGTH];
>char after[1024];
>
>main(int ac, char *av[])
>{
>  int i;
>  for (i = 0; i < 1024; i++) after[i] = 0;
>  getwd(&buf[0]);
getwd() calls internal getcwd(), witch need a parameter length.
The programmers of libc assumed, that the buffer has at least 1k size.
getwd(&buf[0]) ==> getcwd(&buf[0],1024);
>
>  for (i = 0; i < 1024; i++)
>    if (after[i] != 0) printf("after[%d] = %d\n", i, after[i]);
>}
getcwd() now think it can use all of the 1k buffer;

-- 
 >>>># ># >>>>>>>       Harald Boehme -- News-Administration 
 <<<#### <# <# <<       Humboldt Universitaet Berlin, Fachbereich Informatik
 >># ># ># ># >>>       Voice: +49 30 20181 320
 <<<<<<<#### <<<<       SMTP : news@informatik.hu-berlin.de

------------------------------

From: Paul M Sargent <ee92pms@brunel.ac.uk>
Subject: Setting up SVGALib with ET4000
Date: Thu, 11 Aug 94 10:00:38 BST


Hi There,

I'm having a couple of problems setting up my Diamond Stealth 32 with SVGALIB, 
(Strangly enough this is causing more hassle than X). I've used the Tseng3.exe 
program under DOS and so got all te register maps for all the modes, but:

a) Pageflipping has one corrupt screen and one good one. (3d is a good example) 
This also means when you leave the program the text font is completly corrupt.

b) Hi-Resolution modes have clock frequencies completely outside anybodys 
monior spec. (i.e. 800x600 at H-Freq of 29Khz & V-Freq of 45Hz and lower)

c) Hi-Colour Modes have completely wrong colours.

I put b) down to the dot clocks being set incorrectly, Can I set these up 
either in the libvga.et4000 file or by the clock program I use for X?
I think c) is down the the type of DAC on the card, I've been through all the 
possable selections for DAC's but no luck. Any Ideas.

Thanks in advance

Paul

P.S. I do not have the source for the SVGALIB library. Is it available to hack 
around with if necessary? If so where?


------------------------------

From: mark@gcs.com (Mark Bolzern)
Subject: Re: Database for Linux Summary: Expires: References: <328qef$i27@crl4.crl.com> Sender: Followup-To: Distribution: Organization: WorkGroup Solutions (FlagShip) & GCS, Inc.  Keywords: Cc: 
Date: Thu, 11 Aug 1994 06:10:26 GMT

In article <328qef$i27@crl4.crl.com>, Dennis Heltzel <dheltzel@crl.com> wrote:
>If anyone is looking for a dBase style program for Linux, check out the 
>demo version of FlagShip. Ftp fsdlx.gz from ftp.wgs.com:/pub2/wgs. This 
>is a crippled demo, the full program costs $199 single user, $499 
>unlimited users. I have used the demo to sucessfully compile an existing 
>Clipper app that my company uses, and it works great. The demo is full 
>featured enough that you can thoroughly test the compiler, but crippled 
>enough that you won't be able to use the executables for real work (10 
>day expiration on executables, 15 minute runtime per execution).

Dennis, Thanks for posting this!!!!

Please be sure to point people at the 
ftp://ftp.wgs.com/pub2/Filelist though..... It will always reflect the
current state of the archive....    the other filenames are fluid
and change constantly.......  the name of the linux demo will change this 
week sometime as a new version is put online.....

-- 
Mark Bolzern :  mark@gcs.com    USA Tel: (303) 699-7470  Fax: (303) 699-2793 
WorkGroup Solutions, Inc.    The FlagShip "CA-Clipper and XBase on Unix" People
  FlagShip is a 4GL Database Development System & XBase Porting Tool for Unix
No Runtime Fees   Info at ftp.wgs.com : /pub2/wgs/Filelist OR mail: info@wgs.com

------------------------------

From: rjl@spectre.apana.org.au (Richard Lindner)
Subject: Typo in kernel 1.1.43
Date: Thu, 11 Aug 1994 12:37:18 GMT

the 1.1.43 patches include a typo - in tpqic02.c, SA_INTERRUP should be
SA_INTERRUPT.

-- 
Richard Lindner - System Manager        Upper Murray Public Access Unix
rjl@spectre.apana.org.au                PO Box 1555, Wodonga, Vic. 3689, Oz
Data: 060 208773   Fax: 060 562105      Voice: 060 562072 (bh) 060 208813 (ah)

------------------------------

From: ceilbeck@dra.hmg.gb ()
Subject: Linux/GPIB interfaces
Date: 10 Aug 1994 19:12:33 GMT
Reply-To: ceilbeck@dra.hmg.gb ()


Has anyone done any work on a GPIB driver for Linux?  I've been thinking
about using Linux to control some lab instruments but no-one seems to
want to write a driver for their GPIB card to work with Linux.  Help!

Chris

------------------------------

From: allan@mct.uucp (Allan Clearwaters)
Date: 11 Aug 1994 15:36:17 GMT
Subject: Re: Is SONY CDU31A driver in kernel 1.1.4x robust??

In article <1994Aug10.055755.12042@umr.edu> jlu@cs.umr.edu (Eric Jui-Lin Lu) writes:

   Hi *,

   Is the subject statement true?  I have a SONY 33A attached to
   a SB16 card.  It works flawlessly under 1.1.21.  Since I 
   upgraded to 1.1.4x, I keep getting

   cdu31a: Too many consecutive attention: 11

   from dmesg.  I recalled I saw something similar in the net
   but couldn't find it.  This error results in failure
   of both mounting CD and workman (or any audio CD players).  Is 
   there a way to get around this?  Thanks!!


     --Eric

   -- 
   ***************************************---       Grad. student          ---*
   * Obviousness is always the enemy of  *   \     Jui-Lin Lu (Eric)      /   *
   * correctness.  -- Bertrand Russell   *   /      jlu@cs.umr.edu        \   *
   ***************************************---   Univ. of Missouri-Rolla    ---*

I noticed this about a week ago when I posted the message to which you
refer - I found it when I upgraded from v1.1.38 to v1.1.41.  I've
played with this a bit more since then and it turns out the problem
rears its ugly head in 1.1.41 - I am currently running 1.1.40 and it
happily mounts CDROMs from my Sony CDU31a drive.  I have also tried
.42 and (today) .43 - they still have the same problem.

Both you and I (and I think a number of others) would wish to have
some insight into this problem since from what I can see, patch41 had
little or nothing to do with cdroms or the iso filesystem.  Could it
be that these is some interaction between the floppy disk changes in
.41 and some assumptions about cdrom drives??

Any help would be greatly appreciated.

Thanx,
Al
--

=======================================================================
= Allan Clearwaters             Phone:  44+(684)569292                =
= MCT Ltd.                      Fax:    44+(684)561153                =
= 62 Albert Rd. North           Email:  allan%mct.uucp@britain.eu.net =
= Malvern, Worcs.                       (allan@mct.uucp)              =
= England  WR14 2TL                     The first address guarantees  =
=                                         delivery                    =
=======================================================================

------------------------------

From: gold@wpfd24 ()
Subject: FS-recovery, searching for superblocks
Date: 10 Aug 1994 13:11:44 GMT

Hello folks,

As Linus told in Heidelberg (as far as I can remember):
".... most email I receive is quite positive. It often starts with: 
"Dear God,
Your operating system is really wonderful, but I have this (little) problem
.....
It ate my hard disk ! ............"

..... Oh, yes it did; and it did a very good job with that. :-(

Well, I have two IDE disks in my System. One is for /home and /dos (mmh!)
And the second one for / , /usr, and /usr/local

The System war updated a few weeks ago to "slackware 2.0"

Yesterday I compiled a new kernel. After "make zlilo" I tried to reboot.
I had a second root-session and shutdown told it couldn't unmount one fs.

After rebooting the partition table of my first hd was totally corrupted.
The rc-script ended with a fatal fsck error and I got a login prompt.
(Hey Patrick, root-logins are not allowed on that terminal !?!)
The system could be started from floppy because /dev/hdb was not affected.
I knew /dev/hda1 was /dos and /dev/hda2 was the /-partition of the 
formerly running SLS. /dev/hda4 was the exetnded-part, /dev/hda5 was swap
and /dev/hda6 was former /usr and now /home with ~50MB of user-data.

With lde I found the "LILO" characters in Block 0 of hda2.
(Very special thanks to whoever wrote that and the documentation)
Block 1 covers the superblock of /oldroot. With help of the 
documentation of "lde" I was able to recover hda1 and hda2.

But I failed in recovering the very important user data on hda6.
The size of the swap-partition was something about 8-12 MB.

This was the second accident I had with a corrupted partition-table in 
4 Weeks. The other one was on a SCSI System (not on the same Machine).
After my first crash Remy told me that copies of the superblock can be 
found every 8k. (Thanks for the information!)

Here are my recovery trials:
============================
First of all I made dumps of every block #1 after cylinder-boundaries
and compared the first 16 byte with the first 16 bytes of Block #8193.
This worked on every ext2-fs I have except the one I want.
The documentation of the superblock mentioned a "magic signature".
On every superblock I found on my valid fs's this was 0xEF53 
(is that right ? ).
I ran through the entire hd and searched for this magic number on that
certain position. Negative!
Is there another signature for superblocks on logical volumes?
(All the other ext2 partitions I have are on primary partitions)
Are there other locations for the searched data on logical volumes?

This system is in the institute and not at home so it is neccesary to
run an automatic fsck at boot-time.
Was the partition corrupted from fsck ?
Does fsck really do anyhing when it is unable to read the superblock?
Hello Ted, if you read this: What about a flag, that fsck checks for
valid parition table before any repair ?

I didn't find any documentation about a swapspace signature.
If I could locate the swapspace there might be a chance to locate the
ext2 partition.

I post that to c.o.l.d because I want some fs-hackers to read that. I 
already read all the documentation I could find and spent the whole 
night for that problem.

Werner

PS: Yes I have a backup, but it is 14 days old and some work was done
since then !
--
==============================================================================
Werner Gold                             email: gold@physik.uni-wuerzburg.de
Universitaet Wuerzburg                  voice: ++49/931/888-5124
Experimentelle Physik III/MBE           fax:   ++49/931/888-5142
Am Hubland
97074 Wuerzburg
                  This article was created on a machine
================= that runs the incredible LinuX system ======================


------------------------------

From: bherrman@netcom.com (Robert Herrmann)
Subject: IPX on Linux
Date: Thu, 11 Aug 1994 14:54:49 GMT

I am looking for software to route Novell IPX packets on a Linux Box.
IPX kernal based sockets would also be nice.

any pointers appreciated
bherrman@netcom.com



------------------------------

From: pkc@scs.carleton.ca (Peter Choynowski)
Subject: More nfs problems
Date: Thu, 11 Aug 1994 15:15:59 GMT

I  noticed a  number of  NFS  problems being  reported. I  too have  a
problem, which  might or might not  be related to those  reported. The
problem shows up  on most kernels I have tried  above 1.1.18 ( .24,.36
). I  have narrowed  it down  to a  combination of  kernel mountd+nfsd
daemons. The symptom  is that reading files mounted over  NFS gives an
IO error.  Programs like gcc or  tar fail to read  some files. Killing
off nfsd and mounted, and restarting  them on the server, fixes things
for a  period of time (  about 2 days of  low NFS activity ).  Both my
server and clients  are based around Slackware 2.0. I  am setting up a
lab of about 14 Linux systems, so NFS is very important to me. If I go
back to 1.1.18 things work fine, but the performance is just not there
for a network of client and server systems.


Peter

------------------------------


** 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-Request@NEWS-DIGESTS.MIT.EDU

You can send mail to the entire list (and comp.os.linux.development) via:

    Internet: Linux-Development@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 Digest
******************************
