Subject: Linux-Development Digest #717
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:     Fri, 13 May 94 21:13:08 EDT

Linux-Development Digest #717, Volume #1         Fri, 13 May 94 21:13:08 EDT

Contents:
  Re: wt news--version 0.02 - port to MSDOS  ("Nigel T. Barber")
  Re: Linux 1.1.12...compile problem (S. Joel Katz)
  Re: COMMODORE CALLS IT QUITS (Alexandra Griffin)
  Re: NFS Buffercache broken or not implemented? (Alexandra Griffin)
  Re: CD-ROM's do all read and write now? (Rob Janssen)
  Re: Compressed filesystem??? (Robert G. Smith)
  Re: COMMODORE CALLS IT QUITS (Grant Edwards)
  workaround for SB16 on stupid motherboard? (David Monro)
  Re: wt news--version 0.02 - port to MSDOS (Chuck Bentley)
  Xfree86/mouse/IRQ/config problem (Douglas Donahue)
  Strange story of a T4600C notebook (David Fox)
  Re: SLIP bitch (vir@45acp.slip.com)
  Re: Strange story of a T4600C notebook (Matt Cross)
  Re: BogoMips 7.?? on 386/40 (David Fox)
  Re: SIGHUP - Deep Kernal Guts question! (Matt Cross)

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

Crossposted-To: rec.games.programmer,comp.windows.x
From: hx5-ntb@gptcsta.demon.co.uk ("Nigel T. Barber")
Subject: Re: wt news--version 0.02 - port to MSDOS 
Reply-To: hx5-ntb@gptcsta.demon.co.uk
Date: Thu, 12 May 1994 12:56:47 +0000

I for one would love to see a Watcom port of wt, as I have V9.5 (hopefully
V10.0 soon).

Thanks for anything (source would be best :-) )

-- 
Nigel T. Barber

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

From: stimpson@panix.com (S. Joel Katz)
Subject: Re: Linux 1.1.12...compile problem
Date: 13 May 1994 08:04:14 -0400

In <CpqI2r.Ez6@tzv.fal.d400.de> heli@tzv.fal.d400.de (Helmut Lichtenberg(m)) writes:

>So I read all the posts in c.o.l.h and c.o.l.d, followed all the 
>"insert this ... in ..." and compiled the kernel although 
>I'm no expert in c and Makefiles but want to get use of
>kernel improvements. What's wrong with this?

        There is nothing intrinsically wrong with it, but the interim
releases are intended for testing and evaluation by experts. They are not
polished and may contain bugs and other problems.

        They are released to allow problems to be detected by people who
are competent to detect these problems and either fix them or report them
accurately. When the versions are polished, a new general release is issued.

        Using the interim releases exposes you to risks that you may not
be able to handle. If you are not doing serious work on your system, why
does it matter if it takes you an extra week or two to get the latest
features? If you are, why are you taking the risk of using dangerous code
when you will likely be unable to handle the problems it may throw you?

        SJK
        Stimpson@Panix.COM


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

From: acg@kzin.cen.ufl.edu (Alexandra Griffin)
Crossposted-To: comp.os.linux.misc
Subject: Re: COMMODORE CALLS IT QUITS
Date: 10 May 1994 22:51:59 GMT

In article <2q3q5a$8ht@clarknet.clark.net>, Ken Bass <kbass@clark.net> wrote:
>: I guess this means that the market for PCs will be focused around the
>: Intel, MIPS, Sparc and powerPC series of chips.
>
>  As opposed to what- the C64, Vic20, or Amiga chip?

The 64 & vic were 6502-based-- no danger of that architecture
re-emerging, I hope, even though they did manage to take it up to 32
bits at one point (68C832?).  The Amigas were MC68K machines.  Quite
ahead of their time in 1985 (I had an A1000), when CGA was still
considered high resolution, but it's a pity Commodore kind of let them
die out...

It would be interesting if HP purchased rights to the Commodore
product line... PA-RISC based Amigas would be quite nice!
-- 
______
\    / //////////////////////////////////////////////
 \  / / Alexandra Griffin /// acg@kzin.cen.ufl.edu /
  \/ //////////////////////////////////////////////

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

From: acg@kzin.cen.ufl.edu (Alexandra Griffin)
Subject: Re: NFS Buffercache broken or not implemented?
Date: 10 May 1994 23:03:34 GMT

Isn't the Andrew File System (AFS) supposed to yield better
performance than NFS?  Several supercomputing centers use AFS servers
for mounting filesystems onto Crays & the like...

What is the legal status of AFS?  Could a port to Linux be done
-- 
______
\    / //////////////////////////////////////////////
 \  / / Alexandra Griffin /// acg@kzin.cen.ufl.edu /
  \/ //////////////////////////////////////////////

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

Crossposted-To: comp.os.os2.programmer.misc,comp.os.minix,comp.os.mach,comp.periphs,comp.unix.bsd,comp.unix.pc-clone.32bit,comp.os.386bsd.development,comp.sys.ibm.pc.hardware.misc
From: rob@pe1chl.ampr.org (Rob Janssen)
Subject: Re: CD-ROM's do all read and write now?
Date: Fri, 13 May 1994 08:56:16 GMT
Reply-To: pe1chl@rabo.nl

In <2quq75$ft2@nic.smsu.edu> gdl297s@cnas.smsu.edu (Ledford George D) writes:

>Brian Fritz/ ADVISOR Russell (MAE) (rme81977@zach.fit.edu) wrote:
>: Can someone tell me if all CD drives read and write now?  Which ones
>: are the "top of the line"?
>: Thanks
>: -BRian

>In a word, no.  Only CD-ROM mastering drives can write, the rest still just
>read.  CD-ROM mastering drives still cost upwards of $2,500 for the drive
>alone

Indeed

>and require CD-ROM mastering software.

Doesn't all hardware require software?  A plain read-only CD-ROM drive
requires a driver and a filesystem as well.
CD-ROM mastering software is available for Linux for free, so I don't think
this is the problem...

Rob
-- 
=========================================================================
| Rob Janssen                | AMPRnet:   rob@pe1chl.ampr.org           |
| e-mail: pe1chl@rabo.nl     | AX.25 BBS: PE1CHL@PI8UTR.#UTR.NLD.EU     |
=========================================================================

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

From: rob@bip.anatomy.upenn.edu (Robert G. Smith)
Subject: Re: Compressed filesystem???
Date: 11 May 1994 01:50:21 GMT

Andrew R. Tefft (teffta@cs690-3.erie.ge.com) wrote:
: In article 16508@n5ial.mythical.com, jim@n5ial.mythical.com
: (Jim Graham) writes:
: >There is also a third option which, btw, is the one I use.  Some time ago,
: >there appeared a util called gexe.  I think it was posted to one of the
: >comp.sources.* groups, but I'm not sure.  Anyways, it allows you to

: It is gzexe, and it comes with gzip.

: I tried it for a while but it complained on every binary (they still
: worked, just complained) so I gave up. It wasn't worth the effort.
: I considered using it on seldom-used but big binaries but decided 
: to just delete them instead ;-)

Try "tcx", available on sunsite.  It "is a system designed for
the transparent decompression, execution, and recompression of
executibles under Unix. It allows ... timeouts between 
recompressions, and emergency directories in case a decompression
fails from shortage of disk space.  The system is designed with
a reasonable amount of robustness..."

It works, with less trouble than gzexe!

Rob Smith



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

Crossposted-To: comp.os.linux.misc
From: grante@reddwarf.rosemount.com (Grant Edwards)
Subject: Re: COMMODORE CALLS IT QUITS
Date: Wed, 11 May 1994 19:40:11 GMT

Maarten Boekhold (Who'd you expect??) (boekhold@morra.et.tudelft.nl) wrote:
: Alexandra Griffin (acg@kzin.cen.ufl.edu) wrote:

: > The 64 & vic were 6502-based-- no danger of that architecture
: > re-emerging, I hope, even though they did manage to take it up to 32
: > bits at one point (68C832?).  The Amigas were MC68K machines.  Quite

: The 68000 (A1000, A500, A2000) work internally with 32 bits. But it has 
: only a 24 bits adress-bus and a 16 bit data-bus. same for 68010. Since 
: teh 68020 the microprocessors were completely 32 bit. The thing you 
: mention is prbably the 68332, which is a microcontroller (notice the 
: _controller_). It incorporates extra interfaces (serial etc. timers 
: internally...) 

Nope, there was a dual-mode 8/32 bit wide 6502 compatible cpu (I don't
remember the number, but 68C832 doesn't sound quite right).  

--
Grant Edwards                                 |Yow!  An Italian is COMBING
Rosemount Inc.                                |his hair in suburban DES
                                              |MOINES!
grante@rosemount.com                          |

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

From: davem@extro.ucc.su.OZ.AU (David Monro)
Subject: workaround for SB16 on stupid motherboard?
Date: Wed, 11 May 1994 16:10:08 GMT

I am one of those poor bastards who has 1) a nice new Sound Blaster 16
and 2) a dumb motherboard which can't handle 16bit dma. Under MessyDOS
this can be worked around by using the same DMA channel for both
8 and 16 bit transfers (by default dma channel 1) which works perfecly
well. Unfortunately, this doesn't work under linux. Although the
configuration program will let you specify it, there is a note in the
readme to the effect that it can't be done yet.
In fact, I can't get the card to do anything at all under linux - not
8bit sound, FM synth or beeps come out. I can't record either.
Currently, all that works is the line in -> line out.
Even worse, if I boot dos after using linux, the card appears to be
in some sort of crashed state and I can't get any sound out of it. The
test program complains about failing to initialize the voice when it
gets to the FM test (which is the first test which actually produces
sound). Looks like something is getting badly messed up. This goes away
if I powercycle the machine. Can't remeber if the reset button fixes
it or not. Anybody got this to work at all, or has any pointers?
As far as I can tell from the Developer Kit I borrowed from a friend,
it looks to me as if this should be transparent - just unset the
appropriate bits in the config register and away it goes. As far as I
can see, this is exactly what the driver does. So why doesn't it work?
How does it put the card into some wierd state where nothing works?

        David Monro

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

Crossposted-To: rec.games.programmer,comp.windows.x
From: chuckb@hounix.org (Chuck Bentley)
Subject: Re: wt news--version 0.02 - port to MSDOS
Date: Fri, 13 May 1994 15:42:51 GMT

Please post these ports, or tell us where we can get them.
========
        Chuck......

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

From: odoncaoa@panix.com (Douglas Donahue)
Crossposted-To: comp.os.linux.admin,comp.windows.x.i386unix,comp.os.linux.misc,alt.uu.comp.os.linux.questions,aus.computers.linux,fr.comp.os.linux,maus.os.linux,de.comp.os.linux,fj.os.linux,no.linux
Subject: Xfree86/mouse/IRQ/config problem
Date: 13 May 1994 18:52:43 -0400

Greetings,

I am trying to determine and correct, what would appear to be a (soft)
configuration problem on my system. Previously, I had a similar
problem. The cause of which, apparently turned out to be a hard-
ware problem. Linux didn't like the other busmouse card. I'm
guessing that the I/O port address for it was not 23Ch-23Fh. In any
case, the problem I am having is still apparently mouse related.

X now seems to initialize ok. I get an X session with a workspace and
a virtual desktop. However, I have almost no control over the mouse. As
you may note, a warning has been generated by xinit reguarding the
mouse:

> Warning: unable to get status of mouse fd (Invalid argument)

The behaviour of the 'mouse to cursor' interaction resembles what I
would normally conclude to be an IRQ conflict. When trying to interact
via the mouse, the cursor keeps bouncing back to the [0,0] position.
As it is moved, the cursor jumps around to about 9 or so positions
with a degree of regularity. The positions are roughly [0,100],
[100,100], [100,0]; [0,200], [200,200], and [200,0], the corners of a
square. Occasionally, the window manager control pop-up menu apprears.
I am pretty confident that the problem is not an IRQ conflict
because I have double checked the jumper configurations for all of the
boards in the machine. The system works well under windoze/DOS. No
conflicts or strange behaviours are encountered (well, other than the
stock issue ones ;} ). I also pulled all but the video, adaptec scsi
disk controller and mouse cards and tried X again. The behaviour
remained.  Have you seen this before? Any suggestions? I have tried
using all of the other mouse drivers. i.e. signifying every mouse
keyword in the 'Xconfig' file and none of those worked (as expected).

The following is output of 'startx' on my machine. I have already
reconfigured and recompiled the kernel on my machine and, yes IRQ5 is
setup for the MS bus mouse that I am using. The mouse is found and works
correctly with MS windows.  It is jumpered correctly as well (IRQ5,I/O=
23Ch-23Fh=primary,AT slot). Oh yea, I've a Trident 9200cxr which I
configured as an 8900c graphics adapter.

Here are the entries in /dev:

lrwxrwxrwx   1 root     root            8 Mar 23 08:05 mouse -> bmousems
crw-rw-rw-   1 root     root      10,   2 Nov 30 14:30 bmousems

Here is the output of startx:

> XFree86 Version 2.0 / X Window System
> (protocol Version 11, revision 0, vendor release 5000)
> Operating System: Linux 
> Configured drivers:
>   VGA256: server for 8-bit colour SVGA (Patchlevel 0):
>       et4000, et3000, pvga1, wd90c00, wd90c10, wd90c30, wd90c31, gvga, ati,
>       tvga8800cs, tvga8900b, tvga8900c, tvga8900cl, tvga9000, clgd5420,
>       clgd5422, clgd5424, clgd5426, clgd5428, clgd6205, clgd6215, clgd6225,
>       clgd6235, ncr77c22, ncr77c22e, cpq_avga, oti067, oti077
> (using VT number 7)
> 
> Xconfig: /usr2/odoncaoa/Xconfig
> (**) stands for supplied, (--) stands for probed/default values
> (**) Mouse: type: Microsoft, device: /dev/mouse, baudrate: 9600
> (**) FontPath set to "/usr/lib/X11/fonts/75dpi/,/usr/lib/X11/fonts/Speedo/,
> /usr/lib/X11/fonts/misc/"
> (**) VGA256: chipset:  tvga8900c
> (**) VGA256: videoram: 1024k
> (**) VGA256: clocks:  25.20  28.30  44.90  36.00  50.35  40.35  65.00  75.00
> (--) VGA256: Maximum allowed dot-clock: 90MHz
> (**) VGA256: Mode "800x600": mode clock =  36.000, clock used =  36.000
> (**) VGA256: Mode "640x480": mode clock =  25.000, clock used =  25.200
> (**) VGA256: Virtual resolution set to 800x600
> (--) VGA256: SpeedUp code selection modified because virtualX != 1024
> Warning: unable to get status of mouse fd (Invalid argument)
>
> waiting for X server to shut down 
> 
> xinit:  Unknown error (errno 0):  Client error.

Cheers,

Doug

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

From: fox@graphics.cs.nyu.edu (David Fox)
Subject: Strange story of a T4600C notebook
Date: 11 May 1994 14:32:05 GMT

I just got a call from the repair shop about my Toshiba T4600C
notebook computer.  I brought it in because the floppy drive
was giving me troubles - it was unable to mount HD disks under
linux, formatting under DOS would fail when it tried to write
to the boot sector I think; whatever it does after it asks 
you for a label, it would get an INT 24 failure.  It usually
couldn't read HD disks in DOS, though some disks were ok.
They replaced the floppy drive and the motherboard with no
improvement.  They re-partitioned and re-formatted the hard
disk, still no improvement.  They just called up and asked
if I had any non-DOS partitions on the disk - of course, I
said yes.  They say that is the problem, and they're going
to replace the hard disk.  Any theories about what is going
on here, and how I can prevent it happening again?
--
David Fox                                                       xoF divaD
NYU Media Research Lab                             baL hcraeseR aideM UYN

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

From: vir@xmission.com (vir@45acp.slip.com)
Subject: Re: SLIP bitch
Date: 11 May 1994 02:29:23 GMT

Jim Wygralak (darus@MCS.COM) wrote:

: I guess I need to clarify things a bit. I made the original post in haste.
: First of all I should apologize for calling this a bitch, when it is
: really just a point for discussion.

        I bet, it was the title of your post that had guys on the net
        swarming all over you :-)

        I haven't used SLIP on Linux since I was down-sized by my former
        employer, but I'll try to clarify things.  I am happily using
        PPP now....no problems, no hassle.  connected the first time
        I set it up from Slackware 1.1.

: For everyone telling me to RTFM about dynamic addressing, I already have,
: and I understand it. Dynamic addressing is not my problem.
: *MY* IP address is static.

        Which IP address, the one you set up when you configured your
        Net package or the IP address shipped back by your HOST when
        you make a dynamic SLIP connection..... Read on...

: The part that is dynamic is the IP address of the HOST, depending which
: dialin line I call on. (He has about 100 dial in lines connected to the
: internet with a T1. He is not an amateur.)

        This is how most Host connected Internet accesses are normally
        set up.  The local Internet access here uses a terminal server
        connected to several modems and each port is assigned a different
        IP address.  So the IP address may be different each time you
        connect.  The only time the HOST here sends you a static IP address
        is if you requested it and you have a InterNIC registered IP
        address (normally set up by the Provider).


: His assertation is that MY end of the SLIP link shouldnt need to know
: what the IP address of the OTHER end is.

        Your end should know but should be transparent to you as a user
        of your system... is I think what your provider is trying to tell
        you.

: I have some software that runs under DOS (ka9q networking package)
: that works just that way. It couldnt care less what the IP address of the
: system at the other end of the phone line is.

        But ka9q package does care what the other end's IP address is, it is
        just transparent to you (look at the ka9q source).

: Dip on my linux system, on the other hand wont let me complete the setup
: of the slip connection until I tell it the IP address of the host system.

        I think others, who have answered your post were leading you to
        a version that supports bootp or rarp, which should automagically
        pick up the IP address when the terminal server SLIP connects
        and ships the IP address back to DIP.

        Even with the original DIP program that was supplied with the
        distribution you are using, you can manually pick up the IP
        address, once SLIP on the host side connects.  You just have
        to do a get $local on your end, when the TS sends it.  I was using
        an older dip-337.uri version before and I had to manually type the
        IP address once the TS put it up on screen.  Was a pain before,
        cuz you couldn't do auto dial dip scripts with it to a Host that
        uses dynamic IP addressing.  I think this has changed, though.

: Who's right here? How should this REALLY work?

        Get the new dip-337 version as someone suggested that supports
        bootp and start from there.


CUL,
vir
--
    _/_/_/  _/      _/  _/_/_/   Salt Lake Internet Project(tm)
   _/      _/      _/  _/    _/  y.a.c.c.  : Vir Lagua, Jr.
  _/_/_/  _/      _/  _/_/_/     voice     : 801.571.5602
     _/  _/      _/  _/          data / fax: 801.571.2020
_/_/_/  _/_/_/  _/  _/
email: vir@xmission.com : vir@world.std.com : vir@45acp.slip.com

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

From: mcross@sw.stratus.com (Matt Cross)
Subject: Re: Strange story of a T4600C notebook
Date: 13 May 1994 20:31:28 GMT

In article <FOX.94May11103205@first.cs.nyu.edu>, fox@graphics.cs.nyu.edu (David Fox) writes:
-> I just got a call from the repair shop about my Toshiba T4600C
-> notebook computer.  I brought it in because the floppy drive
-> was giving me troubles - it was unable to mount HD disks under
-> linux, formatting under DOS would fail when it tried to write
-> to the boot sector I think; whatever it does after it asks 
-> you for a label, it would get an INT 24 failure.  It usually
-> couldn't read HD disks in DOS, though some disks were ok.
-> They replaced the floppy drive and the motherboard with no
-> improvement.  They re-partitioned and re-formatted the hard
-> disk, still no improvement.  They just called up and asked
-> if I had any non-DOS partitions on the disk - of course, I
-> said yes.  They say that is the problem, and they're going
-> to replace the hard disk.  Any theories about what is going
-> on here, and how I can prevent it happening again?

My suggestion is to find a better repair shop.  There is no reason
that having a non-DOS partition would break your floppy drive.  There
is also no reason they should have to reformat your hard drive if your
floppy drive doesn't work.  It sounds like they're grabbing at straws.
I can't diagnose your problem from here, but my guess (if they did
actually replace the motherboard and the floppy drive) would be that
the cable is bad.

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

From: fox@graphics.cs.nyu.edu (David Fox)
Subject: Re: BogoMips 7.?? on 386/40
Date: 11 May 1994 16:33:20 GMT

In article <2qoni1$2s3@pandora.sdsu.edu> mgutierr@mentor.sdsu.edu (Mario Gutierrez) writes:

] On my overclocked 486/66 I get 39.94 BogoMips.  This is probably in the FAQ
] but what is a BogoMips?  Is 39.94 better or worse then 7.2?  (I assume
] it's something with instructions)

A BogoMips is a FAQ topic often mistaken for a benchmark.
--
David Fox                                                       xoF divaD
NYU Media Research Lab                             baL hcraeseR aideM UYN

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

From: mcross@sw.stratus.com (Matt Cross)
Subject: Re: SIGHUP - Deep Kernal Guts question!
Date: 13 May 1994 20:41:50 GMT

In article <1994May13.042516.27841@loreli.ftl.fl.us>, nate@loreli.ftl.fl.us (Sean Puckett) writes:
> The kernal apparently wants to send just one HUP/CONT pair
> only to the session leader, and none to any other processes
> on the lost tty.  The session leader (shell) is sucessfully
> killed on hangup, every time, but its children are not.

Take a look at something on job control and process groups.  The kernel
sends the signal to the whole process group, so unless the shell is
broken, the SIGHUP signal will be sent to all processes attached to the
terminal.  Also, make sure the processes themselves don't leave the
process group...

        -Matt

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


** 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
******************************
