Subject: Linux-Activists Digest #183
From: Digestifier <Linux-Activists-Request@senator-bedfellow.MIT.EDU>
To: Linux-Activists@senator-bedfellow.MIT.EDU
Reply-To: Linux-Activists@senator-bedfellow.MIT.EDU
Date:     Sun, 29 Aug 93 18:13:07 EDT

Linux-Activists Digest #183, Volume #6           Sun, 29 Aug 93 18:13:07 EDT

Contents:
  Re: Need a Good DERIVE clone (Michael Plate)
  Re: Bug in atof? (Brandon S. Allbery)
  Re: linux slow > 16mb RAM? (Brandon S. Allbery)
  Emacs Problems with 0.99pl11 (news@triode.apana.org.au)
  Re: DOS partition (Wolfgang Schelongowski)
  Problem installing gcc-2.4.5 (Harry C. Johnson)
  Stacker-like Compression?
  Re: 486DX50 or 486DX266 (Peter C. Norton)
  Tcsh 6.04 and Linux SLS 1.03pl12 - bug! (Michael Cederberg)
  Re: Driver yet for Adaptec 1522? (Drew Eckhardt)
  Re: SCSI Performance (Yet Again) (Andreas Klemm)
  Re: 99pl12 hard crash? (Andreas Klemm)
  Re: 99pl12 hard crash? (Andreas Klemm)
  Re: 32 bit compiler for Linux?
  Re: linux slow > 16mb RAM? (Drew Eckhardt)
  mouse woes (Dante)
  Source for /usr/bin progs? (Sreekar Shastry)
  Re: help with system to run unix (Drew Eckhardt)
  Re: 99pl12 hard crash? (Bill Henning)

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

From: plate@uni-kassel.de (Michael Plate)
Subject: Re: Need a Good DERIVE clone
Date: 28 Aug 1993 09:01:43 GMT

Hallo,
Just use maxima. It runs under clisp, and is just easily to compile.
Don't ask where to find, use archie ans search for
                maxima.tar.Z
or something like that. I don't know how much (if..) it costs for
students, but I know somebody running it here (they also buyed the
commercial one, macsysma). The only prob is to use it... sometimes
not so easy, read the readme's.


Michael Plate

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

From: bsa@kf8nh.wariat.org (Brandon S. Allbery)
Subject: Re: Bug in atof?
Date: Sun, 29 Aug 1993 13:35:00 GMT

In article <930828.220210.4B1.rusnews.w165w@mulvey.com> rich@mulvey.com (Rich Mulvey) writes:
>quantum@stein1.u.washington.edu (Tsung-lung Li) writes:
>>     x = atof( st );
>>     printf( "%s %f\n", st, x );
>> }
>> 
>> Execution results:
>> 1E+4 1.000000
>   You need to include stdlib.h so that you get the correct prototype for
>atof.  As it is now, atof is implicitly defined to return an int.

I think the problem here is that atof() is traditionally declared in math.h.

++Brandon
-- 
Brandon S. Allbery         kf8nh@kf8nh.ampr.org          bsa@kf8nh.wariat.org
"MSDOS didn't get as bad as it is overnight -- it took over ten years
of careful development."  ---dmeggins@aix1.uottawa.ca

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

From: bsa@kf8nh.wariat.org (Brandon S. Allbery)
Subject: Re: linux slow > 16mb RAM?
Date: Sun, 29 Aug 1993 13:39:46 GMT

In article <CCICsp.2Ls@world.std.com> entropy@world.std.com (Lawrence Foard) writes:
>Maybe someone should make a patch so that Linux can be made to use
>>16Megs for buffer cache only, on defective machines?

That's the *last* thing you want to patch it to do.  Except on EISA machines
--- which will most likely get the cache right --- you can't do DMA above
16MB.  And the disk driver likes to do DMA to the buffers....

++Brandon
-- 
Brandon S. Allbery         kf8nh@kf8nh.ampr.org          bsa@kf8nh.wariat.org
"MSDOS didn't get as bad as it is overnight -- it took over ten years
of careful development."  ---dmeggins@aix1.uottawa.ca

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

From: news@triode.apana.org.au
Subject: Emacs Problems with 0.99pl11
Date: Sat, 28 Aug 1993 11:00:02 GMT


Every since I upgraded from 0.99pl9 to 0.99pl10 and then 0.99pl11 I have had a problem
with my Emacs cursor control keys. When I hit say the up arrow key I get odd characters
written into my active buffer. Same with the other cursor control keys they no longer work.
I can use emacs commands such as ^N and ^P ^F and ^B to move the cursor around. 

When running under Vi all of the cursor control keys still work correctly. Can anyone
tell me where to look to sort out this problem. Thanks... Paul


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

From: ws@xivic.bo.open.de (Wolfgang Schelongowski)
Subject: Re: DOS partition
Date: Sun, 29 Aug 93 12:53:46 MEST

teffta@cs690-3.erie.ge.com (Andrew R. Tefft) writes:

> In article <CCEI0y.FF3@world.std.com> swift@world.std.com (Peter Lu) writes:
> >
> >     I inadvertently did a "dd of=/dev/hda if=..." and trashed the
> >partition table on my C disk.  Luckily, I knew its layout and just
> >manually recovered the partition set-up.  However, the first partition I
>
> Hm I think I will use dd to back up the important part of my
> drive in case I ever do something like this.
>
> Question -- how many bytes should I save? I know next to nothing
> about the layout of these drives.

Well, dd of=/dev/hda if=/dev/null   DESTROYS ALL   of your /dev/hda.
You must save your _complete_ disk to recover from _that_.

To save your bootsector (includes partition table),
 dd if=/dev/hda of=<your save file> bs=512 count=1
and save that file on a floppy. You restore by
 dd of=/dev/hda if=<your save file> bs=512 count=1

--
Wolfgang Schelongowski  ws@xivic.bo.open.de
Getting people to give vast amounts of money when there's no firm
idea what that money will do is like throwing maidens down a well.
  -- P. J. O'Rourke, Give War a Chance

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

From: hcj@gull.unccsun.uncc.edu (Harry C. Johnson)
Subject: Problem installing gcc-2.4.5
Date: Sun, 29 Aug 1993 15:16:08 GMT

I have the SLS 1.02 distribution of a 386DX40 with 4M ram and 12M swap
space.  I recently decided to upgrade to pl11, which of course needs
gcc-2.4.5 which needs libc-4.4.1....  Anyway I retrieved the new
compilers, libraries, and kernel source, but now I can't compile the
new libraries.  When I try to compile anything, or just run gcc with
no arguments I get the following error:

Warning: gcc: libc.so.4: incompatible minor version numbers (DLL jump 4.4pl1).

I obtained the compilers from sunsite, and they were precompiled
binaries.  I'm sure that I am overlooking something, but I don't know
what.  If anyone has any ideas, I would appreciate hearing them.

Thanks,
-Harry
--
=============================================================================
                             Harry C. Johnson IV

                             Phone: (704)547-4339

                         E-Mail: hcj@gull.uncc.edu

                            Department of Chemistry
                   University of North Carolina at Charlotte
                        Charlotte, North Carolina 28223

"Seems to me, its all just chemistry..."
                                        -Neil Peart
=============================================================================



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

Crossposted-To: comp.os.linux.misc
Subject: Stacker-like Compression?
From: <LENBC@CUNYVM.BITNET>
Date: Sunday, 29 Aug 1993 11:16:03 EDT

Hi folks,

Is there a stacker-like on the fly disk compression software
for Linux?  I'm running out of disk space real quick.



                                                       Edward Galarza/2 2.1
                                                           Brooklyn College
                                                City University of New York
                                                      Lenbc@cunyvm.cuny.edu
                                                        Lenbc@cunyvm.bitnet


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

From: pn002b@uhura.cc.rochester.edu (Peter C. Norton)
Subject: Re: 486DX50 or 486DX266
Date: Sun, 29 Aug 93 15:23:55 GMT

Well, alll I can add is that if you want to run anything useful, then
avoid getting a DIamond Card.  You can't use Xfree86 without a hell of
a lot of trouble.  And even after fiddling with it for ever, you can't
always get it to work perfectly.

My advice: Don't bother getting a super-quick video card unless you
plan on playing games under DOS.  Spend the money on a larger moniter
(Ahh... for a 5FG!).  There aren't a whole lot of applications that
come to mind that require lightning fast video refreshes, but if you
need it, well then go for it.

---Peter

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

From: c878109@id.dth.dk (Michael Cederberg)
Subject: Tcsh 6.04 and Linux SLS 1.03pl12 - bug!
Date: Sun, 29 Aug 1993 16:52:02 GMT

Hi there!

I'm having trouble running tcsh 6.04 under Linux SLS 1.03 pl12. Tcsh
starts up allright, but after I haved runned a few programs (<5), it hangs.
ps reports that the child-shell is a Zombie, and that the parent shell
is sleeping in sys_sigsuspend.

After recompiling tcsh 6.04 and the kernel several times, I gave up and
fetched tcsh 6.03 again and it gives me no trouble ... Wether it's a bug in
Linux or tcsh I don't know, but it's a bug!

Michael Cederberg

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

From: drew@juliet.cs.colorado.edu (Drew Eckhardt)
Subject: Re: Driver yet for Adaptec 1522?
Date: Sun, 29 Aug 1993 19:41:29 GMT

In article <25megj$4m0@vixen.cso.uiuc.edu> mikeh@ux1.cso.uiuc.edu (Mike Hollyman) writes:
>That's my big question.  Is there a driver out yet for the 1522?

Not yet.  One is being written.

-- 
Boycott USL/Novell for their absurd anti-BSDI lawsuit. | 
Condemn Colorado for Amendment Two.                    | Drew Eckhardt
Use Linux, the fast, flexible, and free 386 unix       | drew@cs.Colorado.EDU 
Will administer Unix for food                          |

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

From: andreas@knobel.knirsch.de (Andreas Klemm)
Subject: Re: SCSI Performance (Yet Again)
Date: Sun, 29 Aug 1993 17:42:08 GMT

root@hip-hop.suvl.ca.us (Remco Treffkorn) writes:

>soup@SonOSam.execnet.com (John R. Campbell) writes:
>: I keep seeing bitching and moaning about "poor SCSI performance" during
>: reads with all kinds of hand waving and inconsistent numbers.

>: Linux's performance is EXTREMELY respectable against $CO.  I'm not even
>: counting Bang for the Buck, either.

To *what* SCO release running with *which* hardware ...
Which benchmark, which results ....

Please, facts ;-)
-- 
/-\       Andreas Klemm   <andreas@knobel.knirsch.de>      +-----------------+
|@|########################################################-@ "pay for it !" |
\-/   41469 Neuss     Germany     phone +49/ 2137 12609    +-----------------+

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

From: andreas@knobel.knirsch.de (Andreas Klemm)
Subject: Re: 99pl12 hard crash?
Date: Sun, 29 Aug 1993 17:46:14 GMT

john.will@satalink.com (John Will) writes:

>I've seen someone else mention a problem with the release 99pl12 kernel,
>and I believe I've stumbled over a bug in it as well, semi serious, because
>it causes a hard crash.  The easiest way to make it happen is to run 
>X-windows and start opening Xterms, at about 9 or 10, a hard crash occurs.
>I don't have this problem with the 99pl12 Aug 8 ALPHA, or 99pl11, so it
>seems a recent introduction.  Since it's a hard crash, I don't have much
>data on what happens, but I've also been having trouble with tar 1.11.2
>doing a similar thing, now I'm not so sure it's tar's problem!  Anyone
>else seen hard crashes in the 99pl12 release version?

Last week I gave 0.99pl12 a try. I used that file:
-r--r--r--   1 andreas  root       862003 Aug 24 23:01 lx099p12.tgz
-r--r--r--   1 andreas  root        55413 Aug 24 23:01 lxpatch.pl12.tgz

After having crashes (a hanging system without any error message)
- running X11 or not is meaningless - I switched again to an older
release level. Now - again - everything runs fine.

-r--r--r--   1 andreas  root       793633 Aug 11 15:40 linux-0.99.11.tar.gz

My hardware: ECS 80486/33 Eisa, 16 Mb, AHA 1542B.
-- 
/-\       Andreas Klemm   <andreas@knobel.knirsch.de>      +-----------------+
|@|########################################################-@ "pay for it !" |
\-/   41469 Neuss     Germany     phone +49/ 2137 12609    +-----------------+

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

From: andreas@knobel.knirsch.de (Andreas Klemm)
Subject: Re: 99pl12 hard crash?
Date: Sun, 29 Aug 1993 18:19:01 GMT

john.will@satalink.com (John Will) writes:

>JW>I've seen someone else mention a problem with the release 99pl12 kernel,
>JW>and I believe I've stumbled over a bug in it as well, semi serious, because
>JW>it causes a hard crash.  The easiest way to make it happen is to run 
>JW>X-windows and start opening Xterms, at about 9 or 10, a hard crash occurs.
>JW>I don't have this problem with the 99pl12 Aug 8 ALPHA, or 99pl11, so it
>JW>seems a recent introduction.  Since it's a hard crash, I don't have much
>JW>data on what happens, but I've also been having trouble with tar 1.11.2
>JW>doing a similar thing, now I'm not so sure it's tar's problem!  Anyone
>JW>else seen hard crashes in the 99pl12 release version?

>I know it's bad form to reply to your own messages, but I fixed this with
>the new patch that was issued for the release pl12, so anyone running with
>this version should CERTAINLY apply this patch!  The patch appeared in
>c.o.l.a a couple of days ago, but the description didn't seem to be urgent,

Which c.o.l.a ? There are two.
Under what subject ... I didn't find it yet ....
Please reply through email.
-- 
/-\       Andreas Klemm   <andreas@knobel.knirsch.de>      +-----------------+
|@|########################################################-@ "pay for it !" |
\-/   41469 Neuss     Germany     phone +49/ 2137 12609    +-----------------+

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

Date: Sun, 29 Aug 1993 22:24:48 CDT
From: <K111114@ALIJKU11.BITNET>
Subject: Re: 32 bit compiler for Linux?

Wrong question.
You should ask, is there a 16bit compiler for Linux?
The answer is: no.

/Herp

PS: does as86 count as 16bit compiler?


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

From: drew@juliet.cs.colorado.edu (Drew Eckhardt)
Subject: Re: linux slow > 16mb RAM?
Date: Sun, 29 Aug 1993 20:15:08 GMT

In article <1993Aug29.133946.12206@kf8nh.wariat.org> bsa@kf8nh.wariat.org (Brandon S. Allbery) writes:
>In article <CCICsp.2Ls@world.std.com> entropy@world.std.com (Lawrence Foard) writes:
>>Maybe someone should make a patch so that Linux can be made to use
>>>16Megs for buffer cache only, on defective machines?
>
>That's the *last* thing you want to patch it to do.  Except on EISA machines
>--- which will most likely get the cache right --- you can't do DMA above
>16MB.  And the disk driver likes to do DMA to the buffers....

No.  

- The normal disk driver doesn't do DMA period, because standard ST-506
        and many IDE controllers don't support DMA.

- The SCSI system only uses DMA above 16M when the low level driver supports 
        DMA (some such as the Seagate and Future Domain drivers don't),
        and when the low level driver explictly requests that destination
        addresses above 16M be allowed.  If the low level driver says the 
        hardware can't handle it, transfers that ultimately end up above
        16M are done directly to a buffer in low memory (allocated at boot
        time, based on the types of SCSI controllers in the system and 
        number of devices) and on completion copied by the CPU to their 
        ultimate destination.

        Usually, there aren't any problems with this, except for when 
        the field in the Ultrastor 14F control structure got botched
        and incorreclty claimed that the board could handle DMA > 16M.

In any case, this would get a little messy, since Linux uses a fully 
unified buffer cache and user memory pool.  When you run a program,
the VM points to buffer cache pages and says "That's me, over there".
So, you'd end up running the program out of buffer cache and things could
be slower than if you restricted memory usage to the low 16M.  Having
a separate section of buffer cache only buffer cache could get messy.
-- 
Boycott USL/Novell for their absurd anti-BSDI lawsuit. | 
Condemn Colorado for Amendment Two.                    | Drew Eckhardt
Use Linux, the fast, flexible, and free 386 unix       | drew@cs.Colorado.EDU 
Will administer Unix for food                          |

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

From: gt2584a@prism.gatech.EDU (Dante)
Subject: mouse woes
Date: 29 Aug 93 20:28:19 GMT


    
   Well, I finally got MGR to compile after I
laid aside the swap space correctly.  It will run, but
it will not let me move the mouse or use any of the buttons.
right now, /dev/mouse is linked to bmouselogitec, because
i have a logitech bus mouse.  i have tried linking it
to various other things, but none seem to work.
When I boot the machine, the BIOS recognizes the mouse,
and when Linux starts booting, it says it detected a 
PS/2 type mouse.  i tried linking /dev/mouse to
/dev/ps2aux,but to no avail.  Does anyone know how
to make it work ?

 
-- 
DISCLAIMER:  The opinions contained herein are not mine, but 
belong to my cat who has been reprogramming my brain while
I sleep.


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

From: sreekar@panix.com (Sreekar Shastry)
Crossposted-To: comp.os.linux.help,comp.os.linux.misc
Subject: Source for /usr/bin progs?
Date: 29 Aug 1993 17:13:54 -0400


Does anyone know where I can get the source for some of the programs in /bin
and /usr/bin like ls, whereis, df and the like?

Thanks.
--

Sreekar Shastry
sreekar@panix.com
-- 

Sreekar Shastry
sreekar@panix.com

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

Crossposted-To: comp.unix.pc-clone.32bit,comp.os.minix,alt.amateur-comp
From: drew@juliet.cs.colorado.edu (Drew Eckhardt)
Subject: Re: help with system to run unix
Date: Sun, 29 Aug 1993 21:08:13 GMT

In article <25opra$ohe@news.ysu.edu> ag173@yfn.ysu.edu (Jay Hauben) writes:
> 
>She has been quoted a price for a system at the upper limit of 
>what she can afford. The system consists of the following 
>components. Does anyone see in this system anything that would 
>prevent her needs from being met? Might any of the components be 
>too weak for what she wants to do? The forms of UNIX she might 
>have available are Minix (which appears not to have uucp) 
>and off the net Linux. She has a 2400 baud modem now and hopes to upgrade 
>it and maybe buy a CD-ROM player and controller some time in the 
>future when she is in a position to afford them.
> 
>The system she has the quote for consists of:
> 
> 2) Mid height tower case with 230 watt power supply. (Will this    
>    give enough power and enough VENTILATION?)

You shouldn't have any problems.

> 3) 245 MB hard drive by Seagate or Maxton. (Is either 
>    preferable?)

I've had nothing but problems with small Seagate drives - the old 
251's suffered from stiction, my 296N was cheap at the time but 
sounds like a jet airplane taking off, my ST1144A developed bad sectors 
that the drive locks up when accessing (as in you have to power 
cycle the system before using the drive again, a controller reset
or hard reset won't clear things up).  I haven't seen any problems 
on Maxtor drives, or large CDC (bought by Seagate) drives.

> 4) 128 kB cache (Might she in fact need a 256 kB cache?)

You'll see a bigger performance increase spending the money on
real memory.

> 5) 4 MB RAM expandable to 64 MB. (Should she insist on SIMM    
>    chips?)

Every modern system uses some form of SIMMs, 9 or 36 bits wide.

If she wants to run X, 8M is recommended, if she turns off all the 
drivers she isn't using to reduce kernel bloat and doesn't run X,
4M will be useable.

> 8) Pentium Overdrive with Intel (?) chips

Modern chips are limited by memory bandwidth.  To address this, chip
makers put a wider bus interface on - ie, the Pentium has a 64 bit 
interface to main memory, MIPS R4000SC's have a 128 bit interface to
cache, etc.  You'd be strangling a pentium if you ran it in a 32 bit 
overdrive socket (most benchmarks fit in cache, but you'd see problems
in the real world).  Since you'd have to replace the motherboard to get
full pentium performance, pentium upgradability is a non-issue.

> 9) 14" Acer 34T (SVGA NI) monitor.

Monitors vary greatly, even those of the same make and model.  If the 
shop is local, she should insist on seeing the monitor she's purchasing.

>11a) Trident 1 MB SVGA video card with a DC 009 IDE controller     
>    card

Trident boards are dog slow.  A cheap alternative (I've seen new for
for $60) would be a TSENG et4000 based board.

>11b) Genoa IMD VESA local bus graphics card with a VLB IDE     
>    controller card (for about $150 more). (Is there a significant
>    advantage gained for the extra money even though she doesn't     
>    anticipate doing much graphics?)

If she runs X (I do because it's a great way to get six command lines 
on the screen at one time), she should get a 10 fold improvement in 
scrolling, etc. over the trident.

A cheaper but not as fast alternative would be an ET4000 based ISA
board.

>Is there anything in the above system that might prevent her from 
>successfully running UNIX or handling her DOS files? Is there 
>anything she should change in order to better insure that her 
>system might last for a long time?

If she wants to run X, I'd suggest going with 8M of memory, trading 
the 486DX-33 for something slower like a FPU less Cyrix 486DLC or even
a 40Mhz AMD 386 as price concerns dictate.  The overall performance 
will be better when things run in real memory, not swap.

-- 
Boycott USL/Novell for their absurd anti-BSDI lawsuit. | 
Condemn Colorado for Amendment Two.                    | Drew Eckhardt
Use Linux, the fast, flexible, and free 386 unix       | drew@cs.Colorado.EDU 
Will administer Unix for food                          |

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

From: bhenning@vanbc.wimsey.com (Bill Henning)
Subject: Re: 99pl12 hard crash?
Date: 29 Aug 1993 14:52:58 -0700

In article <1993Aug29.174614.2501@knobel.knirsch.de> andreas@knobel.knirsch.de (Andreas Klemm) writes:
>john.will@satalink.com (John Will) writes:
>>I've seen someone else mention a problem with the release 99pl12 kernel,
>>and I believe I've stumbled over a bug in it as well, semi serious, because
>>it causes a hard crash.  The easiest way to make it happen is to run 
>
>After having crashes (a hanging system without any error message)
>- running X11 or not is meaningless - I switched again to an older
>release level. Now - again - everything runs fine.

Ah!

So I am not alone in experiencing mysterious crashes.

I suspect the block device hard disk drivers, or the caching code, as
the crashes *seem* to occur when I start up a lot of disk activity.

For example, running xdvi has hung my system twice.

Once after a hang my IDE controller/hard drives were left in a pretty
confused state - it was not enough to hard reset, I had to power off
and turn the machine back on to get it to read the disk - I kept getting
bad sector reads etc.

Bill

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


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

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

    Internet: Linux-Activists@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
    tupac-amaru.informatik.rwth-aachen.de	pub/msdos/replace

The current version of Linux is 0.99pl9 released on April 23, 1993

End of Linux-Activists Digest
******************************
