Subject: Linux-Misc Digest #197
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:     Mon, 11 Oct 93 19:13:17 EDT

Linux-Misc Digest #197, Volume #1                Mon, 11 Oct 93 19:13:17 EDT

Contents:
  Re: Don't need BogoMips (Stephen Tweedie)
  [ANSWER] Re: Why only compressed kernels? (Stephen Tweedie)
  Re: keep linux mag on paper... (Alan Cox)
  Re: Don't use SLS (Re: Which linux should I install...or which is best?) (Andrej Bauer)
  Re: SCSI adapter for linux? (robjan@rabo.nl)
  Re: Linux counter passes 2000 entries (Julian Francis Day)
  Re: Bogomip (Julian Francis Day)
  Re: Why only compressed kernels? (Kristian Ejvind)
  Re: Term and more Term (Julian Francis Day)
  Would Linux run on PCI / Pentium based systems? (Chung Ng)
  Re: Linux Magazine... (James H. Haynes)
  Re: keep linux mag on paper... (Andrew R. Tefft)
  Re: [Announce] New util* packages uploaded (Zeyd M. Ben-Halim)
  Re: Linux Magazine... (Mark Buckaway)
  Personal info about Linus (fro)
  >Re: Linux magazine now accepting articl (Gareth Bult)
  Re: Linux user count passes 2000 users (Philippe Bonal)
  Re: Kernel .99.13 and sig 11's (Gregory Gulik)

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

From: sct@dcs.ed.ac.uk (Stephen Tweedie)
Subject: Re: Don't need BogoMips
Date: Mon, 11 Oct 1993 17:02:51 GMT


In article <CEqDDJ.L84@informatik.uni-rostock.de>,
klein@informatik.uni-rostock.de (Marko Klein) writes:

> I have installed SLS 1.03 now and found the kernel computing
> BogoMips.  I was not so happy about seeing that a benchmark is
> calculatet every time I boot linux because it's a waste of time (not
> too much but time is money).  I have commentet out the parts of
> source defining this and everything works great. That sounds well
> and somebody could ask "Why is he telling us this ?"  Well, I wanted
> to say, that I'd liked it if it would be possible to disable
> BogoMips via 'make config' in future versions.

It's not quite so simple.  The bogomips calculation is *not* there as
a benchmark; it is there as a necessary feature which is used by some
parts of the kernel.

Certain hardware devices (in particular the floppy-interface-driven
tape drives) require the kernel to generate signals at a precise, high
frequency.  To get accurate timing, it is necessary to use delay
loops.  The bogomips calculation is performed to calibrate the delay
loops.

So, beware, it is not in general safe to remove the BogoMips
calculation, although in principle it should be safe enough as long as
you are not using any device drivers which employ delay loops.

Cheers,
 Stephen.
---
Stephen Tweedie <sct@dcs.ed.ac.uk>   (JANET: sct@uk.ac.ed.dcs)
Department of Computer Science, Edinburgh University, Scotland.

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

From: sct@dcs.ed.ac.uk (Stephen Tweedie)
Subject: [ANSWER] Re: Why only compressed kernels?
Date: Mon, 11 Oct 1993 17:08:17 GMT

In article <CEqGEI.D1L@dcs.ed.ac.uk>,
st@epcc.ed.ac.uk (Scott Telford) writes:

> I've just upgraded from 0.99.10 to 0.99.13 and I notice that
> uncompressed kernels are no longer supported in the kernel Makefile. I
> can see compressed kernels are very handy for boot floppies, but why
> have uncompressed kernels been dropped altogether, especially as it
> prevents you from doing things like "strings /vmlinux" etc.? Is the
> kernel size getting too close to the 640k limit? Can somebody
> enlighten me?

You got it in one.  The kernel boot image, plus LILO and the initial
kernel page tables, have all got to fit into this 640K limit.  If you
compile a kernel with networking, SCSI and a few filesystems, you will
exceed this limit by quite a large margin.  Ever since 0.99pl10 or so
(when the new networking went into the kernel), it has been very easy
to generate kernels so large that they wouldn't boot unless
compressed, so Linus (quite rightly) decided that it would be easier
just to make all kernels compressed.  It saves a lot of grief.

Cheers,
 Stephen.
---
Stephen Tweedie <sct@dcs.ed.ac.uk>   (JANET: sct@uk.ac.ed.dcs)
Department of Computer Science, Edinburgh University, Scotland.

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

From: iiitac@swan.pyr (Alan Cox)
Subject: Re: keep linux mag on paper...
Date: Mon, 11 Oct 1993 15:42:09 GMT

In article <1993Oct9.131457.2224@ivax> trwills@indyvax.iupui.edu writes:
>While the idea of using a floppy for distribution sounds neat,
>it would be nice to keep it in paper form to allow non Linux users
>to get at the data, and hopfully convert over to the "the promised
>land" with the rest of us......

Paper urghhhhh... Its expensive to do small runs on paper (try it), why not
send it on a 1.44Mb disk, then its highly recyclable and linux goodies
can fill the rest of the disk.

Alan
iiitac@pyr.swan.ac.uk


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

From: Andrej.Bauer@ijs.si (Andrej Bauer)
Subject: Re: Don't use SLS (Re: Which linux should I install...or which is best?)
Date: 11 Oct 93 18:33:26 +0100

> : I'm afraid I've yet to see any serious problem with SLS.  After installing
> : SLS 1.03 I found that yes, some permissions weren't set quite right, but
> : after a little fiddling the whole thing works wonderfully. 

I've encountered aleph_0 bugs in SLS 1.01 (you tell me how much 1.03
changed). Here's a short list of what comes to my mind first:

1) Installation of SLS via network (NFS) did not work. I had to hack
   the sysinstall file and mount NFS manually.

2) File permisions all over the filesystem are wrong. If I remember
   correctly, those were 'world writable' files which shouldn't really
   be writable.

3) Duplicated binaries in /bin and /usr/bin (e.g. ls, mv, cp, cat, ...)
   I had a hard time figuring which ones were *the right* ones.
   Also, mount and umount were duplicated in /bin and /etc.

4) Missing /etc/brc. As a result my filesystem did not unmount properly
   at shutdown. At boot-time Linux barked at 'unchecked file system'.
   This problem was solved after I got bootutils and efsck 0.3.

5) They did not bother to tell the users that there is such a thing
   as 'selection' which can be compiled and installed. After that
   Linux has cut & paste on VC.

6) broken /usr/lib/makewhatis. If you run it, nothing is left of
   your whatis files (not that they were useful in the first place).

7) Some of the non-nroffed man pages are in /usr/cat? directories
   instead of /usr/man? directories.

8) Emacs 18.59 which came with SLS 1.01 had incompatible executable
   and DOC files, which meant you coouldn't get any Help in Emacs.

9) Couldn't they inlclude a 9-pin printer support for TeX?

10) There are .tfm files for Euler fonts, but there are no .mf files.
    Also, MakeTeXPK is duplicated and is not a very nicely written
    script.

11) xdm without shadow password support. Running xdm is a particularly
    nice way of breaking the system.

12) Coming back to 'sysinstall' (see ad. 1), why doesn't sysinstall
    say it has a '-series' option in the 'usage' text?

13) UUCP related files and directories had UID which was different
    from the UID that belonged to 'uucp' according to /etc/passwd.

14) Place you favourite error here: ...............................


That's all for now. If anyone has a solution to feature no. 6, please
tell me.

Andrej

===================
Andrej.Bauer@IJS.si
===================

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

From: robjan@rabo.nl
Subject: Re: SCSI adapter for linux?
Date: 11 Oct 93 19:22:08 GMT

Newsgroups: comp.os.linux.misc
Path: pe1chl!rob
From: rob@pe1chl.ampr.org (Rob Janssen)
Subject: Re: SCSI adapter for linux?
Date: Mon, 11 Oct 1993 15:52:04 GMT

>>Basically, the system will hang, with the computer's hard-disk light on
>>solid (which indicates to me that the SCSI bus is hung).  I'm using SCSI
>>only, so there's no other controller which affects this light.

>I am having the exact same problem here- I added a Fujitsu M2511A
>128Meg MO drive to my Ultrastor 14f+Texel 3024 (v.1.10) combination,
>and all of the sudden I am having tons of SCSI problems with Linux.

A while ago I took my system (including Adaptec 1542B and WD disk) to work
and attached a Sony SMO with the intent to copy the system over to an MO
disk.  This must have been woth 0.98-pl6 or such.
At that time I got the SCSI timeouts as well, and believed they
were caused by this slow SMO on the bus, maybe it wasn't disconnecting
during a write and the harddisk timed out on its next pending read command.

I successfully copied the disk by dumping it to streamer (QIC-02) and
then restoring that onto the SMO.  So, this problem must be related to
using the two devices (SMO and harddisk) alternately.
Since then I have added a CD-ROM on the SCSI bus without any problem.

In Adaptec's technical docs I found something about requesting the controller
to disconnect after every command, and included that in the device driver
startup.  it did not seem to make difference.  Anyway, I would believe that
Drew knows what he is doing and decided that this command wasn't necessary
in the Linux driver.

So, it must be something else...  but I didn't see it anymore without the
SMO so I did not really check it.

Rob Janssen
pe1chl@rabo.nl

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

From: jfd0@aber.ac.uk (Julian Francis Day)
Subject: Re: Linux counter passes 2000 entries
Date: Mon, 11 Oct 1993 19:22:58 GMT

Mark A. Davis (mark@taylor.uucp) wrote:
: ....  For example:  a Unix version of WordPerfect
: will support X-Windows, ASCII terminals, multiuser applications, many devices,
: many spooling systems, has system administration tools, knows about file
: permissions, etc, etc, etc.   The MS-"DOS" version of WordPerfect, even running
: under Unix, will be very much a resource hog, not multiuser knowledgable,
: and run into countless other problems making it all but useless except in a
: single-user, console only environment.

I won't argue with what you say about WP for DOS, but if WP for unix is as
bad as WP for VMS *anything* is better. It is a processor hog, X support is
naff, sysadmin tools are naff where they exist is extremely buggy (I found
(and reported at least 20)).

-- 
Julian Day
=====
No sig.                 (jfd0@aber.ac.uk)

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

From: jfd0@aber.ac.uk (Julian Francis Day)
Subject: Re: Bogomip
Date: Mon, 11 Oct 1993 19:14:58 GMT

Joost Helberg (jhelberg@nlsun8.oracle.nl) wrote:
: In article <CEKz9D.MC@scrum.greenie.muc.de> root@scrum.greenie.muc.de writes:
:    choltje@ux1.cso.uiuc.edu (Christian Holtje) writes:

:    >>>>>>Two more datapoints:
:    >>>>>>486DX-33 -----> 16.5 bogomips
:    >>>>>>486DLC-33 -----> 11.2 bogomips   (Cyrix cross between 386 and 486)
:    >>>>>386DX-25  --> 3.91 BogoMips(tm).
:    >>>>386DX-33/387DX-33 --> 6.03 BogoMips(tm)
:    >>> 486DX66/2 --> 34.06 BogoMips (tm)
:    >>386DX40   ---> 6.99 BogoMips (tm)
:    >486DX50/2  ---> 25.0 BogoMips (tm)
:    486SX25  ---> 12.24 BogoMips (tm)
: 386-20 --> 1.67 BogoMips (tm)
386DX40 (128K Cache) --> 7.23 BogoMips (tm)
-- 
Julian Day
=====
No sig.                 (jfd0@aber.ac.uk)

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

From: krej@electrum.kth.se (Kristian Ejvind)
Subject: Re: Why only compressed kernels?
Date: Mon, 11 Oct 1993 19:22:52 GMT

In article <CEqGEI.D1L@dcs.ed.ac.uk> st@epcc.ed.ac.uk (Scott Telford) writes:
>    I've just upgraded from 0.99.10 to 0.99.13 and I notice that
>    uncompressed kernels are no longer supported in the kernel Makefile. I
>    can see compressed kernels are very handy for boot floppies, but why
>    have uncompressed kernels been dropped altogether, especially as it
>    prevents you from doing things like "strings /vmlinux" etc.? Is the
>    kernel size getting too close to the 640k limit? Can somebody
>    enlighten me?

  The problem is with LILO, it can't load boot images larger than
512k. Thus, if your boot image is smaller than 0.5M uncompressed, you
can go on running it uncompressed.

/Kristian
--
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Kristian Ejvind - krej@it.kth.se         I   The best way of accelerating
The Royal Institute of Technology, KTH   I   a macintosh is by 9.82 m/(s*s)
Stockholm, Sweden                        I
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

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

From: jfd0@aber.ac.uk (Julian Francis Day)
Subject: Re: Term and more Term
Date: Mon, 11 Oct 1993 19:30:36 GMT

Michael K Patterson (mikep@iastate.edu) wrote:
: Anytime I go into xwindows, I can't use my modem. It 
: shows the sure signs of an IRQ conflict (you can dial,
: but you can't send or recieve characters). I can set
: the IRQ for my modem, but I can't find where in Linux to
: change the IRQ Linux uses... Any suggestions?
It probably is an IRQ conflict. At a guess, your modem is on COM3 and your
mouse is on COM1. Try moving your mouse to COM2 and changing the link for
/dev/mouse to /dev/cua1. (both COM1 and COM3 use the same IRQ as do 2 and
4)

-- 
Julian Day
=====
No sig.                 (jfd0@aber.ac.uk)

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

From: chungn@liege.ICS.UCI.EDU (Chung Ng)
Subject: Would Linux run on PCI / Pentium based systems?
Crossposted-To: comp.os.linux.help,comp.os.linux
Date: 11 Oct 93 20:18:36 GMT

Hello,

Would Linux run on PCI / Pentium systems?  Specifically the
Gateway 2000 P5-60 with the PCI based ATI Graphics Ultra Pro?

Any comment regarding this machine is welcomed.

Thanks for the information in advance.

Chung Ng

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

From: haynes@cats.ucsc.edu (James H. Haynes)
Subject: Re: Linux Magazine...
Date: 11 Oct 1993 21:07:32 GMT


In article <1993Oct8.033117.4035@fylz.com> linux@fylz.com (Linux Journal)
   writes:
>The answer is yes.  The name is Linux Journal.  We were moving along
>quite well and then a set of fairly amazing disasters struck (including
>the death of one of the people involved).  I send out a "it looks like

Why wasn't this cross-posted to alt.conspiracy?    (Not to make light of
your disasters, but, sorry, it just hit my funny bone.)
-- 
haynes@cats.ucsc.edu
haynes@cats.bitnet

"Ya can talk all ya wanna, but it's dif'rent than it was!"
"No it aint!  But ya gotta know the territory!"
        Meredith Willson: "The Music Man"


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

From: teffta@cs690-3.erie.ge.com (Andrew R. Tefft)
Subject: Re: keep linux mag on paper...
Reply-To: teffta@cs690-3.erie.ge.com
Date: Mon, 11 Oct 1993 20:50:23 GMT

I have a great source of Linux information online, so an online
magazine is kind of redundant. An actual publication however is
much different (and much more convenient to browse; about the only
things that online documentation is better at are searching/referencing
and archival!). Besides, a linux-accessible online magazine is hardly
useful if one does not have linux available (don't have a laptop
to take with me to work, for example, and not everyone interested in
Linux will be running it). And not all of us will really want to
or be able to *print* it either!

I have not yet jumped onto the online-documentation bandwagon. I much
prefer to have things nicely printed out than to fight with them
on the screen. And even though postscript prints great, it's hard
to get published-quality documents, double-sided and nicely bound,
by printing them yourself.

---

Andy Tefft               - new, expanded .sig -     teffta@cs690-3.erie.ge.com



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

From: zmbenhal@netcom.com (Zeyd M. Ben-Halim)
Subject: Re: [Announce] New util* packages uploaded
Date: Mon, 11 Oct 1993 20:51:01 GMT

In article <29c0ad$ddq@samba.oit.unc.edu> Rik Faith <faith@cs.unc.edu> writes:
>The following packages are available from:
>
>util-etc-2.0.src.tar.gz
>util-etc-2.0.bin.tar.gz
>
>     This is a distribution of all those (Linux-specific) programs that
>     "belong" in /etc:
>
>     agetty clock ctrlaltdel dmesg doshell fastboot fasthalt fdformat fdisk
>     frag fsck fsck.minix halt ipcrm ipcs kbdrate makehole mkfs mkswap
>     mount ramsize rdev reboot rootflags setfdprm setserial shutdown
>     simpleinit swapdev swapon sync syslogd update vidmode

Aren't most of these supposed to move to new digs at /sbin?


-- 
Zeyd M. Ben-Halim       zmbenhal@netcom.com
10479 1/4 Santa Monica Blvd, LA, CA, 90025 (310) 470-0281

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

From: mark@datasoft.com (Mark Buckaway)
Subject: Re: Linux Magazine...
Date: Sun, 10 Oct 1993 14:44:44 GMT

Mark J. Mollere (mmollere@whale.st.usm.edu) wrote:

: : ...
: : If there's enough interest, I *might* be able to make a proper commercial 
: : venture out of it and put it in a "glossy" format for folks to buy. (BTW, 
: : the previous magazine I put together was sold at virtually no profit to 
: : myself, - the idea was that the cover price would pay for the production, 
: : mailing and sundry costs and hence be self financing. I'm thinking along 
: : the same lines for a LINUX mag. [ though I wouldn't *discount* profit... 8-) ] )
: : What does the community think?

: Do it, dude!

I'd be interested.

Mark
--
==============================================================================
Mark Buckaway           | root@datasoft.north.net   |  DataSoft Communications
DataSoft Communications | uunorth!datasoft!root     |  62 Rock Fernway
System Administrator    | Voice: +1 416 756 4497    |  Willowdale, ON M2J 4N5
==============================================================================
    "UNIX and OS/2 are operating systems. Windows is a shell, and
                         DOS is an boot virus"
==============================================================================

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

From: fro@hacktic.nl (fro)
Subject: Personal info about Linus
Date: 11 Oct 1993 22:36:30 +0100

Hi Linux Lovers!  

I'm just a newbie, attracted to this nice, free Unix version, and I'd like
to explore it to increase my knowledge of Unix (currently NULL) - inside
and outside. 

I wonder if anybody has some personal data of Linus: What is his
profession, what did he study, is he a professor... I'd really like to
know a bit more about the man that created this OS. 

Thanks in advance!  

Frank
fro@hacktic.nl






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

From: gareth@gblinux.demon.co.uk (Gareth Bult)
Subject: >Re: Linux magazine now accepting articl
Date: Mon, 11 Oct 1993 13:17:51 GMT

On 11 Oct 93 03:14:51 GMT;                                                  
----Bill C. Riemers (bcr@bohr.physics.purdue.edu) said:                     
>                                                                           
>I've never heard of AmiPro... What is it?                                  
>                                                                           

You've never header of AMI Pro !                                            
AMI Pro is Lotus's MS Windows wordprocessor.
It is VERY popular, and a damn sight better than things like MS Word and    
Wordperfect.

.....                                                                      

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

From: philb@blips.frmug.fr.net (Philippe Bonal)
Subject: Re: Linux user count passes 2000 users
Date: Mon, 11 Oct 93 22:48:31 +0100
Reply-To: philb@frmug.fr.net

Hello,

hta@uninett.no (Harald T. Alvestrand) writes:
[..]
> BY DOMAIN
[..]
>   38 net       38 gov       37 at        35 dk        27 org
>   26 fr        20 ch        19 unknown   15 za        15 uucp

In France the net is very badly distributed.
Only some school and universities have the .fr  .

Many others users use *.fdn.org or *.fr.net or *.gna.org

In the 3rd case many users post on fr.comp.os.linux from a french bbs net 
but they have not an email adress. The email adress you can see 
(a_user@a_bbs.frmug.fr.net for example) is in reality only the email
adress of the bbs which make the connection beetween usenet world and bbs
world.

I think that the good thing would be to scan each country.comp.os.linux.


Philippe

+---------------------------------------------------------------------------+
 philb@frmug.fr.net  Philippe Bonal (France)  Please,Not email > 25k without
 philb@blips.frmug.fr.net  bonal_p@epita.fr          inform me, No NeXT mail
+---------------------------------------------------------------------------+

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

From: greg@serveme.chi.il.us (Gregory Gulik)
Subject: Re: Kernel .99.13 and sig 11's
Date: Mon, 11 Oct 1993 20:22:04 GMT

In article <29bo1f$fcc@vixen.cso.uiuc.edu> btf57346@uxa.cso.uiuc.edu (Byron Faber) writes:
>Does anybody else here get sig11s more often with certain kernels?
>
>I upgraded from pl10 to pl13 and in the process noticed that the number
>of signal 11's I get has gone down alot.  
>
>Does anybody else get these, or do I have a bad memory chip somewhere?

Check your cache settings and make sure there are enough wait
states for your memory.  I had the same problem and it only started
after I played around with my cache, and went away when I added
a wait state for my slow memory. 8^(

-greg

-- 
Gregory A. Gulik                                 Call Gagme, a public access
       greg@serveme.chi.il.us                    UNIX system at 312-282-8606
   ||  gulik@rtsg.mot.com                        For information, drop a note
                                                 to info@gagme.chi.il.us

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


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