Subject: Linux-Misc Digest #606
From: Digestifier <Linux-Misc-Request@senator-bedfellow.MIT.EDU>
To: Linux-Misc@senator-bedfellow.MIT.EDU
Reply-To: Linux-Misc@senator-bedfellow.MIT.EDU
Date:     Thu, 27 Jan 94 23:13:23 EST

Linux-Misc Digest #606, Volume #1                Thu, 27 Jan 94 23:13:23 EST

Contents:
  DOSEMU 0.49pl3 and SS24X problems  (David Scholten)
  Diamond SS24X with FREQ program and GREEN Mag 17" monitor (David Scholten)
  How to format a new hard disk? (Wim van Dorst/Prof. Penninger)
  Re: Slackware needs a shadow package! (J.J. Paijmans)
  Re: PC-scancode terminals (Andries Brouwer)
  Re: Lisp anyone? How about CMU Lisp? Garnet? (Paul Wilson)
  Re: Archive of Torvalds/Tanenbaum discussion? (Kenneth L Mitchell)
  An idea for the "command line" DOS emulator. (was re: a new...) (Dan Newcombe)
  Re: IDE > 500MB? (Kevin J Butler)
  SLS and NE2000 (Brian Quandt)
  Re: BogoMips-is higher or lower better? (Damien Neil)
  term111 for BSDI anyone? (Jim Gifford)
  Re: PPP for linux (Blue)
  Re: Emacs-Problems (Johan Myreen)
  Another question Re: Sounds for Mosaic in Linux (Bill C. Riemers)
  Good Linux Package for C course ? (ta107109@menudo.uh.edu)
  Re: fsck -A -a (Ralf Keller)
  Re: Another kind of DOSEMU (David C. Niemi)
  Re: Intel Ethernet Express - need advice on checking it (Sebastian Rahtz)

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

Crossposted-To: comp.os.linux.help
From: scholten@esseye.si.com (David Scholten)
Subject: DOSEMU 0.49pl3 and SS24X problems 
Date: Wed, 26 Jan 1994 17:36:06 GMT

I compiled dosemu and everything works fine.  But when I tell it
to use graphics for the video setting, I always get a black screen when
I return to linux.  There is no way to get it back unless I reboot or
go back into dosemu and run 24xmode.com.  24xmode.com will fix it for the
dos emulator, but the problem returns when I exit the emulator.  I also
noticed that the screen will not scroll when I set graphics on.  (either
via the config file or the vgaon.com).  But when I run vgaoff.com, the screen
will start to scroll properly, but the cursor will not track.

Any ideas?
Thanks 
Dave Scholten


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

Crossposted-To: comp.os.linux.help
From: scholten@esseye.si.com (David Scholten)
Subject: Diamond SS24X with FREQ program and GREEN Mag 17" monitor
Date: Wed, 26 Jan 1994 17:47:45 GMT

I seem to be experiencing lots of problems trying to get my
SS24X card to work under Xfree.  At this point, it would have been
worth my money long ago to buy a supported card considering all of the
time and headaches I've gone through.

But yet, it seems that people are successful in getting it to work.
So now I have a few questions again:

My monitor is a multisync and has the capability to exceed the VESA
recommended settings.  So I decided to start out with the VESA settings.

I tried to use freq to setup clock 2 to 75  using the command
freq 75000 2

The first question that comes to mind is what clock on the SS24X is the
programmable one.  Am I using the wrong one?  There are 9 clocks on this card.

The next step is to startup X.  I run the 640X480 mode as default, since this
uses 25, the first clock on the card.  When I switch to 1024X768, the screen
loses sync and then the monitor powers down into the "green" state.  When
I switch back to 640X480, it repowers up.

Since I have to run freq to program the clock, I do this in a script as
follows:  freq 75000 2;openwin;freq 28000 2
I need to set the clock back for text mode.

This script changes the freq before starting X, so the screen loses sync
when in text mode for a very short period.  But this short period causes the
monitor to power down into the "green" state too.  Then when X starts up and
uses dot clock 25, it repowers up.

Is this powering down into the green state harmful for the monitor?  Especially
since they are short bursts.  Is there any way to not have the monitor
power down?  Basically de-green it (would this be called brown?)?

Last question, has anybody been able to get the SS24X to work with the
VESA recommended settings?

Thanks,
Dave Scholten


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

From: tgcpwd@rwb.urc.tue.nl (Wim van Dorst/Prof. Penninger)
Crossposted-To: comp.os.linux.admin,comp.os.linux.misc
Subject: How to format a new hard disk?
Date: 26 Jan 1994 21:42:04 +0100

Hello *,

How do I format my new 1GB SCSI disk? The only formatting program
that I can find is for floppies. The manual says I should do
'FORMAT C:/S' but I don't have/want dos.

Please?

Met vriendelijke groeten, Wim van Dorst
-- 
=====================================================================
Blue Baron = Wim van Dorst, Voice (+31) 074-443937, (+31) 02152-42319
(-: baron@clifton.hobby.nl  tgcpwd@urc.tue.nl  WvD@Akzo.400net.nl :-)
=====================================================================

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

From: paai@kub.nl (J.J. Paijmans)
Subject: Re: Slackware needs a shadow package!
Date: Thu, 27 Jan 94 18:02:50 GMT

In article <JHELBERG.94Jan27095019@nlsun8.oracle.nl> jhelberg@nlsun8.oracle.nl (Joost Helberg) writes:
>In article <14523@dirac.physics.purdue.edu> bcr@bohr.physics.purdue.edu (Bill C. Riemers) writes:
>   How about this:
>
>   1. Add an encryption feature into the file system.
>   2. Sell the CD at cost.
>   3. Encrypt any program on the CD that the the copyright allows.
>   4. Sell the encryption key needed to access the encrypted stuff.
>
>Are you serious?
>
>This is exactly one of the things the GPL wants to avoid: an
>artificial relationship between customer and distributor with all the
>power at the distributor's end.
>
>Ridiculous.
>--
>   Joost Helberg                              Rijnzathe 6

Why worry? Mr Riemers forgot to add point five:

    5. Wait for the host of small entrepreneurs who sell their encryptionkey
       to all other buyers of that CD...

;-)

Paai.

(Tot ziens in Tilburg...)




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

From: aeb@cwi.nl (Andries Brouwer)
Subject: Re: PC-scancode terminals
Date: Thu, 27 Jan 1994 17:20:31 GMT

frank@icce.rug.nl (Frank Brokken) writes:

>Would anybody have any information about termcap/terminfo definitions
>for PC-terminals, or: does anybody have information about how to access
>the PC's scancodes using Linux ? 

termcap/terminfo:
Look at /etc/termcap and /usr/lib/terminfo. Say man terminfo.
scancodes:
Put the console driver in RAW mode, or use showkey. Say man showkey.
[Showkey is in kbd-0.81.]


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

From: wilson@cs.utexas.edu (Paul Wilson)
Subject: Re: Lisp anyone? How about CMU Lisp? Garnet?
Date: 27 Jan 1994 13:52:46 -0600

In article <1994Jan10.170135.9329@wavehh.hanse.de>,
Martin Cracauer <cracauer@wavehh.hanse.de> wrote:
>xjam@fir.CS.Berkeley.EDU (Brian F. Dennis) writes:
>
>>The CMU guys have consistently claimed that porting CMU Common Lisp to a
>>CISC architecture would be a real drag and nobody's been motivated enough
>>to do it. The runtime is probably not that big a deal since Linux has (?) a
>>real mmap. However, getting Python to deal with the 486 might be a problem.
>
>One of the biggest problems is the lack of registers in Intel chips.
>They need it mostly for the garbage collector, I think.

I wouldn't think the GC would be the big problem.  The big problem
may be dynamic typing in the language itself, i.e., when you load a value
into a register, you often don't know whether it's a pointer or not.
If it is, you strip out the type tag to use it, often by using an addressing
mode with a literal offset that cancels out the tag.  If you have an
irregular instruction set and register file, you may have to move things
to address registers before using them as pointers.

>The Pentium does not change this situation (only 8 general-purpose
>integer registers). So much for Intel...

Isn't it worse than that, even?  I was under the impression there were
4 integer and 4 address registers, or something horrible like that, so
that if the compiler doesn't want to be limited to a tiny number of
registers for certain kinds of code, it has to know that it can do
adds but not multiplies on address registers, and multiplies and adds
but not loads and stores with integer registers.  These are problems
for C compilers, too, but dynamic typing makes it worse.

-- 
| Paul R. Wilson,   Computer Sciences Dept.,   University of Texas at Austin  |
| Taylor Hall 2.124,  Austin, TX 78712-1188       wilson@cs.utexas.edu        |
| (Recent papers on garbage collection, memory hierarchies, and persistence   |
| are available via anonymous ftp from cs.utexas.edu, in pub/garbage.)        |

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

From: klm10@DUTS.ccc.amdahl.com (Kenneth L Mitchell)
Subject: Re: Archive of Torvalds/Tanenbaum discussion?
Date: 25 Jan 94 22:17:24 GMT
Reply-To: klm10@DUTS.ccc.amdahl.com (Kenneth L Mitchell)

>
>|Get it from:  ftp.funet.fi
>|directory:    /pub/OS/Linux/doc/news
>|filename:     Linux_is_obsolete.Z
>

Can someone E-Mail me the IP address of ftp.funet.fi.

The DNS on our gateway says "Unknown Host".

Thanks,
Ken M.

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

From: NEWCOMBE@AA.csc.Peachnet.EDU (Dan Newcombe)
Subject: An idea for the "command line" DOS emulator. (was re: a new...)
Date: 27 Jan 1994 14:09:21 -0600

At least I think that's what the old subject was.  Anyway...

Here is an idea (untested) to be able to run a DOS command from the command 
line (or menu choice, etc...) without modifying the acutal emulator.

This makes a few assumptions:
    1) You have the DOS files/filesystem set up somewhere other than
        in the Dos Emulator hdimage file.
        (ie: the hdimage file does not contain your whole DOS partition.)
    2) You have the above dos file system mounted so that Linux can access.
Basically, what I'm trying to say is that you can access the DOS filesystem 
from Linux.

So suppose you wanted to run wp60.exe with the parameter wp60 d:\doc\paper.txt

You would do something like:
    dosrun wp60 d:\doc\paper.txt

dosrun would be a linux shell program that would 
    a) edit/modify/recreate the dos autoexec.bat from the above mentioned DOS
        filesystem.
            This would somehow keep all the stuff you'd normally want in
            the autoexec (mouse.com) and the last line would be
                wp60 d:\doc\paper.txt
    b) simply run dosemu:
            dos -C >/dev/null
            
On the dosemu side, beforehand, you'd have to:
    a) modify the config.sys (located in hdimage) so that it:
        1) used emufs to access the dos partition as D:
        2) COMPSEC=D:\
        3) shell=c:\command.com /p
            (I think that's the command in #2.  I don't have a DOS
             manual around.)
        
WHAT THE ABOVE IS SUPPOSED TO ACCOMPLISH:
    The idea is that for each time that you load the DOS emulator, you will
    recreate an autoexec.bat that is specific to that session.  What makes
    it specific is that the last line will execute the program you want.
    The modifications on the hdimage are to tell the emulator/DOS that you
    want to use (and effectivly) boot off of D:, which will be the actual
    DOS area/filesystem.
    
    If you do not use hdimage and access the DOS filesystem directly upon
    boot-up of DOSEMU, then this will work, and you don't have to go through
    the hdimage part of this all.
    
        -Dan            

--
Daniel A. Newcombe                                  Clayton State College
Computing Services                                  Morrow, GA 30260
E-Mail Address: newcombe@aa.csc.peachnet.edu        (404)-961-3421
-=-=-=- I can handle MIME mail, so don't be afraid to send me some -=-=-=-
- So if you want my address it's number one at the end of the bar.
                                  (Marillion, Sugar Mice)


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

From: butler@bert.cs.byu.edu (Kevin J Butler)
Crossposted-To: comp.sys.ibm.pc.hardware.storage,comp.periphs.scsi
Subject: Re: IDE > 500MB?
Date: 26 Jan 94 20:47:00 GMT

mlord@bnr.ca (Mark Lord) writes:
(Re:  Limitation of 500MB on IDE drives, naturally...)

>The limitation applies to any PC software that uses the standard BIOS 
>disk I/O routines.  The parameters passed to those routines do not have 
>enough bits for addressing cyls greater than (0..1023).

>In addition to MS-DOS, the linux loader LILO is also affected by this limit,
>as it boots via the BIOS.  Once booted, LINUX itself takes over and talks 
>directly to the bare metal, using 16-bit cylinder numbers for MFM/IDE/RLL drives.
Same applies to OS/2 Boot Manager--if you have a partition with one (or
more :-) ) endpoints > cyl 1023, you can't make it "bootable"...

kb
-- 
Kevin Butler            butler@bert.cs.byu.edu                  8-)
This is a sample .sig file!

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

From: quandt@cs.umr.edu (Brian Quandt)
Subject: SLS and NE2000
Date: Thu, 27 Jan 1994 19:48:23 GMT


Looking for an A1 disk on the SLS distribution that does not
probe the IO addresses (and is already configured for NE2000 support).
Does anyone have such a beast?  I'm tired of having to yank the 
NE2000 card, install LINUX, then rebuild the kernel and then pu the
card back in.  

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

From: damien@b63519.student.cwru.edu (Damien Neil)
Subject: Re: BogoMips-is higher or lower better?
Date: 27 Jan 1994 17:24:38 GMT

In article <MAILQUEUE-101.940127093520.288@aa.csc.peachnet.edu>,
Dan Newcombe <NEWCOMBE@AA.csc.Peachnet.EDU> wrote:
>Last night I was playing around with my CMOS settings and I noticed that
>my BogoMips delay level had changed.
>
>My question then is:  Is it better to have a higher or lower BogoMips number?
>(Is it better to have loved and lost than not to have loved at all?)

Lower is definately better. You see, your BogoMips rating is a measure of
how many bogons are being produced by your machine. If too many bogons are
emitted, you computer will crash. Generally, the more "advanced" chips in
the Intel 80x86 line produce more bogons than the older ones. (You know that
thing on the Pentium that Intel says is a heat sink? Don't believe them --
that's a *bogon* sink, used to help shield it from all the bogosity created
by Intel's marketing department.) The result of all this is that a 486/66
will usually crash the system in a matter of minutes.

In order to counteract this problem, Linux temporarily slows down the CPU
whenever the system bogon count begins to get too high. This of course
makes your compiles take longer to complete, but at least the system stays
up and running.

So you see, you really want to keep your BogoMips as low as possible. A
good start is to remove any bogon sources that may be located near your
computer. Make certain that there are no copies of Windows magazine lying
around. If there is a television set in the room, it must be shielded.
Merely turning it off is not sufficient, since it will have absorbed
large quantities of bogons from news programs, Rush Limbaugh, and the
like.

If, after freeing the area of dangerous bogon sources, your BogoMips
rating remains high (above 1.5 or so), you may wish to install protective
shielding around your CPU to protect the rest of the system from
excessive bogon exposure. Coating your CPU with a dense, bogon-absorbant
substance can radically lower your BogoMips and increase your system's
lifetime. Personally, I like to use peanut butter. (Chunky is best.)
Other people swear by paste, grout, or plaster of paris. Any of these
should be effective.

Remember -- bogons are a danger to your system. Take steps to protect
your data now, before it is too late.

Damien "Everything I say is true" Neil
-- 
Damien Neil  [MIME OK]   CMPS/EEAP  "Until somebody debugs reality, the best
damien@b63519.student.cwru.edu       I can do is a quick patch here and there."
  dpn2@po.cwru.edu  Case Western Reserve University         - Erik Green

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

From: jgifford@thor.fcs.uga.edu (Jim Gifford)
Subject: term111 for BSDI anyone?
Date: 27 Jan 1994 03:27:18 GMT

The sun I have an acct on (moe.coe.uga.edu) likes to 'lose' my
connection frequently.  The only other unix box I have an acct. on is
a BSDI box(thor.fcs.uga.edu).  Since I have started useing it, I have
grown to both love and hate BSDI.  It still has gcc 1.40!  (Linux I
just love, not hate.  ;-)

Anyway, I thought I would get term111 and compile it on thor, but it
keeps breaking at various levels.  I had similar problems with 1.07
and 1.08.  I figure there is no need to wrestle with this if someone
has already ported it to BSDI.  If so, would you mind either mailing
me the binaries or placeing them for ftp somewhere?

Thanks in advance,
Jim

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

From: jleitess@midget.towson.edu (Blue)
Subject: Re: PPP for linux
Date: 27 Jan 1994 23:35:14 GMT

--
xxxxxxxxxxxxxxxxxxxxxxx

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

From: jem@snakemail.hut.fi (Johan Myreen)
Subject: Re: Emacs-Problems
Date: 26 Jan 1994 21:07:19 GMT

In article <JHELBERG.94Jan26154824@nlsun8.oracle.nl> jhelberg@nlsun8.oracle.nl (Joost Helberg) writes:

>Why are there 2 different packages? Emacs will support ASCII as 
>well as X within the same binary.

>This might confuse people.

Sigh. You can't please all people all the time, can you? Once upon a
time there was only one Emacs binary with both X and ASCII support.
This resulted in a ton of messages from people without X asking "why
does Emacs dump core/say that it can't find the shared X libs".

-- 
Johan Myrien
jem@cs.hut.fi
60 11'55" N, 24 53'30" E

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

From: bcr@bohr.physics.purdue.edu (Bill C. Riemers)
Subject: Another question Re: Sounds for Mosaic in Linux
Date: 27 Jan 94 03:35:05 GMT

>(If you have a sound card installed it works perfectly..)

How about for Adlib compatable cards (i.e. /dev/audio is 
not supported.)

                       Bill

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

From: ta107109@menudo.uh.edu
Subject: Good Linux Package for C course ?
Date: 27 Jan 1994 17:16:07 -0600


Hi, I am teaching C programming course this semester (on unix platform), 
and when I talked about Linux in the class, many students were interested.
However, my section is for non-cs majors who do not have any computer
backgrounds.   I myself am using Slackware and happy with it, but
I guess that the package is too big.

So, I am looking for a very small minimum Linux package for C programming
( gcc, make, ...etc. + tcsh + kermit : hopefully within 4 or 5  diskettes 
alltogether), yet very easy to install (for non-experienced users).

Could anyone suggest me such package if it exists ?

Thanks.

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

From: keller@aristoteles.informatik.uni-mannheim.de (Ralf Keller)
Subject: Re: fsck -A -a
Date: 27 Jan 1994 07:29:07 GMT

In article <1994Jan17.225626.683@devnull.adsp.sub.org> froh@devnull.adsp.sub.org (Frohwalt Egerer) writes:

   Newsgroups: comp.os.linux.misc
   Path: darum.uni-mannheim.de!rz.uni-karlsruhe.de!news.uni-ulm.de!news.belwue.de!surz03.hrz.Uni-Marburg.DE!news.th-darmstadt.de!fauern!castle.franken.de!devnull.adsp.sub.org!froh
   From: froh@devnull.adsp.sub.org (Frohwalt Egerer)
   References: <1994Jan9.190430.14161@jupiter.sun.csd.unb.ca> <2gqpgq$236@charm.magnus.acs.ohio-state.edu>
   Organization: none
   Date: Mon, 17 Jan 1994 22:56:26 GMT
   Lines: 45

   cemeier@magnus.acs.ohio-state.edu (Charles E Meier) writes:

   >In article <1994Jan9.190430.14161@jupiter.sun.csd.unb.ca>,
   >Adam Karpowicz <akarpowicz@mta.ca> wrote:
   >>
   >>I have a line in my rc "fsck -A -a" and fsck checks and repairs the fs
   >>every time I boot Linux. It actually came like that with Slackware 1.1.0.
   >>What worries me is that fsck always finds something to do on the drive
   >>giving from a few to two screenfuls lines, such as:
   >>Inode xxxx not used with links_count not null. Repair? Yes.
   >>
   >>Does that indicate a problem with hardware? partition? I do not want
   >>to disable this, but it makes the booting time so much longer.
   >>
   >>Thanks for advice
   >>
   >>Adam Karpowicz   akarpowicz@mta.ca
   >>

   >Never trust a file checker to "fix" your file systems automatically.
   >The -A option says to check all file systems, but the -a says go ahead
   >and repair them if there is a problem.  Change the -a to -r for interactive
   >repair.  The -a option is used in an example in the bootutils package.
   >It should be changed to a -r there as well.

   That won't solve his problem - fsck always find something to repair :-)

   I have this behavior when I boot a kernel which contains an old e2fs. The
   e2fs version of the kernel and the version of the e2fs-tools (as e2fsck) 
   should match. The version numbers are displayed when e2fsck is starting and
   when an e2fs partion is mounted. e2fsck 0.4 will repair partitions written by
   e2fs 0.3 since some structures on the disk changed since then.

   On the other hand, if the version numbers are the same, you might have some
   serios problem. Are any errors displayed when Linux is running? Does MS-DOG
   work fine? (for small amounts of 'work fine' ;-))

I run into the same problem after compiling a new kernel and installing
it using lilo (Slack 1.1.1). Yesterday late (very very late) in the
evening I discovered how to fix this. You has to set the ReadOnly flag
in your kernel. The kernel which comes with the distribution has set
this flag and the new compiled not (this night I want to figure out
why ...). But nevertheless, set this flag using 'rdev -R <your kernel>
flag'. I forgot if flag has to be '1' or '0' (it was really late
yesterday and now I am sitting in my office and not at home in front
of my LinuX box) so please read the man-page.


Ralf
--
=====
Ralf Keller                       E-mail: keller@pi4.informatik.uni-mannheim.de
University of Mannheim            Phone: +49-621-292-1407
Praktische Informatik IV          Fax:   +49-621-292-5745
68131 Mannheim, Germany


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

From: niemidc@YP.lab (David C. Niemi)
Subject: Re: Another kind of DOSEMU
Date: 27 Jan 1994 17:35:08 GMT
Reply-To: niemidc@oasis.gtegsc.com

It definitely would be nice to be able to completely invoke a DOS program
via dosemu without intervention in between.  But it seems to me there is
another capability that could be implemented:

Has anyone looked into the idea of porting 4DOS or one of the other free-
with-source DOS-alikes to Linux directly, and integrating it with DOSEMU?
It seems to me that this would result in a much, much faster way to run
DOS apps and to appear to be in a DOS shell.

On another front, I think it would be nice to have a DOS-style shell available
(as opposed to nice UN*X shells like bash and tcsh) to make DOS refugees
more comfortable.  I can think of quite a few things that would help them:
like built-in "mem", "dir", "path" type commands, perhaps some leeway on
uppercase vs. lowercase (i.e. if you type one case it looks for both; but
if you create a file it will be the usual lower case).

I'd also like to see "mtools" behave a bit more like their DOS counterparts
(e.g. "mcopy a:").

But, unfortunately, I am quite busy with other stuff these days.
---
David C. Niemi  David.Niemi@oasis.gtegsc.com
======================================================
Now I must sit here and ponder the yonder
Herbivores ate well 'cause their food didn't never run



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

From: rahtz@dxcern.cern.ch (Sebastian Rahtz)
Subject: Re: Intel Ethernet Express - need advice on checking it
Date: Thu, 27 Jan 1994 22:26:17 GMT

In <2i67gl$if@aurora.ncdc.noaa.gov> jfauerba@aurora.ncdc.noaa.gov (John Fauerbach) writes:

>If anyone gets linux to work on this card, let me know.  We have several
>PC with this card that we can't use for linux.


theres working and theres `working'. with the help of several kind
people, i eventually got mine alive. but its totally unreliable - the network
connection can last for hours, or minutes, and either just die, or lock
the machine *rigid*. have i compiled the kernel (up to pl14v) and alpha
version of the driver in every conceivable combination a thousand times?
yes, in a word. how many times has ittold me a hundred times a second
it is rebooting the card? many, in a word.

on the other hand, someone else here at CERN has much better performance
with the same card on the same machine. he blames my COM ports.


Draw what conclusion you like...

sebastian rahtz

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


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

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

    Internet: Linux-Misc@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-Misc Digest
******************************
