Subject: Linux-Misc Digest #877
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:     Sat, 26 Mar 94 23:13:08 EST

Linux-Misc Digest #877, Volume #1                Sat, 26 Mar 94 23:13:08 EST

Contents:
  Re: STRAW POLL RESULT: Linux groups automonitoring (Pierre Uszynski)
  Re: Wine status March 11, 1994 (Ken Hornstein)
  Re: 360k floppies not supported? (Thomas Quinot)
  CD ROM Drives (Paul Quinn)
  Re: STRAW POLL RESULT: Linux groups automonitoring (Richard Gill)
  Linux in FedCompWeek (David "Fuzzy" Wells)
  Re: STRAW POLL RESULT: Linux groups automonitoring (Andries Brouwer)
  STRAW POLL RESULT: Linux groups auto-intimidation (dan@oea.hacktic.nl)
  Slackware 1.2.0/Linux 1.0 Problems (Superuser)
  Re: Pentium Optimized GCC (Chris Smith)
  Re: Impressions: FreeBSD vs Linux (Shawn M Carey)
  Re: Maximum serial port speed (REPOST due to non-propogation) (Nemosoft Unv.)
  Re: Maximum serial port speed (REPOST due to non-propogation) (Nemosoft Unv.)
  Re: How do I remove LILO after ptting DOS back on my machine (J.S. van Oosten)

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

Crossposted-To: news.groups
From: pierre@shell.portal.com (Pierre Uszynski)
Subject: Re: STRAW POLL RESULT: Linux groups automonitoring
Date: Sat, 26 Mar 1994 20:19:39 GMT

>In article <2mpm9r$et5@bmerha64.bnr.ca> mlord@bnr.ca (Mark Lord) writes:

>> It is going to waste incredible email bandwidth, [...]

I wish at least experienced users would think before talking about
bandwidth. Please consider that a netnews post is transmitted  from
a feed to a host for each host it shows up on. So, each post is
transmitted once per netnews host (compression saves a factor 2-4, say).
Some sites even use feeds that are a few internet hops away, we'll
neglect that.

The email messages we are talking about are send at most one per post,
from Ian (or one of the co-monitors), to the poster. Granted, there are
a few hops between Ian and the poster.

So, that's at most as many email messages as there are posts for the
monitoring, versus (posts * hosts) post transmissions for netnews itself.
The bandwidths we are talking about are in a ratio O(1/hosts). Now, do
you wish you hadn't open your mouth about bandwidth?

Your only excuse is that one is netnews and does not show to you, and
the other is email and you are more aware of it. But this is not an
excuse to post garbage to the newsgroups, rather than in email.

[for the nit-picking among us, my estimate was even further conservative
as Ian will not send one email per post, and because netnews also spends
quite some bandwidth on various control issues]

Pierre USzynski.
pierre@shell.portal.com

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

From: kenh@wrl.epi.com (Ken Hornstein)
Crossposted-To: comp.windows.x.i386unix,comp.os.386bsd.apps
Subject: Re: Wine status March 11, 1994
Date: 26 Mar 1994 15:10:39 -0500
Reply-To: kenh@wrl.epi.com

In article <JRG.94Mar25034109@blodwen.demon.co.uk>,
james r grinter <jrg@blodwen.demon.co.uk> wrote:
>In article <newcombe.22.000D9C0F@aa.csc.peachnet.edu> newcombe@aa.csc.peachnet.edu (Dan Newcombe) writes:
>   WordPerfect has a version for X, I'm surprised that MicroSloth hasn't
>   followed suit, as it's seems like it'd be a good place to compete.  Or
>
>There is, however, the programs from Astrix (is that right?) that look
>remarkably like Word and Excel. I believe that they're only currently
>available on Suns - but they do look good. Does anyone have any more
>info or experience of them?

We have Asterix here at work; I wouldn't say it looks "remarkably" like Word
and Excel.  It looks like your standard commercial word processor/spreadsheet.
The one big thing I don't like about Asterix is that it has a truely bizarre
macro language.  If it used Tcl for a macro language, it would be great :-)

--Ken

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

From: thomas@melchior.frmug.fr.net (Thomas Quinot)
Subject: Re: 360k floppies not supported?
Date: 26 Mar 1994 00:12:59 +0100

Lloyd Miller (lloyd@lfmcal.cuc.ab.ca) wrote:
: Alex Ramos (ramos@engr.latech.edu) wrote:

: > I've been unable to mount or to use mtools on /dev/fd1, an
: > old 360k drive. Is there anything I'm overlooking?

: I am also having trouble using an old 360/pc floppy drive with linux.
Double Density disks (5"1/4 360 Kb as well as 3"1/2 720 Kb) are *not*
supported by Linux. And it *must* be in the FAQ, since I know it...

Thomas.

-- 
ThoThoThoThoTho
        Totolitoto !

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

From: p_quinn@ECE.Concordia.CA (Paul Quinn)
Subject: CD ROM Drives
Date: Sat, 26 Mar 1994 21:31:49 GMT



I would some questions concerning CD ROM Drives.

I might but a Sony 33A.  Since I'm interested in buying Linux on CD I'd
like to know if this drive is supported by Linux?

        Are there any other drive that you would suggest?

Thanks,
--
+-----------------------------------------------+
|  Paul Quinn                   |               |
|  p_quinn@ece.concordia.ca     | System        |
|  p_quinn@cs.concordia.ca      |  Architecture |
|GCS/E -d+ p++ c++ l++ u++ e+ m+ s++/+ !n h f+  |
|!g w t(+) r- y+                                | 
=================================================

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

Crossposted-To: news.groups
From: dick@gp.com (Richard Gill)
Subject: Re: STRAW POLL RESULT: Linux groups automonitoring
Date: Sat, 26 Mar 1994 20:52:21 GMT

Any chance that this discussion can be moved out of news.groups?
As far as I can tell it has nothing to do with forming a new group and
therefor does not belong here. news.groups is busy enough without this
extraneous discussion adding 25% to the volume.

Thanks.


-- 
Dick Gill                                              dick@gp.com
Gill & Piette/Capital Systems                          uunet!gandp!dick
1568 Spring Hill Road, McLean, VA 22102                 (703)761-1163
You might not get what you pay for, but you never get what you don't pay for.

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

From: wdwells@nyx10.cs.du.edu (David "Fuzzy" Wells)
Subject: Linux in FedCompWeek
Date: Sat, 26 Mar 94 21:07:08 GMT

FYI:  For all those US Government types out there, the latest issue of
Federal Computer Week has a full page review of Linux with a "to be
continued..." message at the end. 

It was quite reassuring to know that I wasn't the only person in the
government to know about Linux.  

If interest (and time) permits, I shall try and post the article
(violating tons of copyrights....oh well...).  BTW, credit for this
project was thrown in various directions with special emphasis on Linus.


Fuzzy

IS Director
Space Warfare Center

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

Crossposted-To: news.groups
From: aeb@cwi.nl (Andries Brouwer)
Subject: Re: STRAW POLL RESULT: Linux groups automonitoring
Date: Sat, 26 Mar 1994 23:04:52 GMT

mdw@cs.cornell.edu (Matt Welsh) writes:

>Maybe it's time to ditch the whole plan.

Yes.
I promised you that it would do no good at all, and increase the noise
on these groups by 5%. As it turns out my estimate was very conservative.

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

From: dan@oea.hacktic.nl
Subject: STRAW POLL RESULT: Linux groups auto-intimidation
Date: Sat, 26 Mar 1994 14:15:21 GMT

Warner Losh (imp@boulder.parcplace.com) wrote:

: Well, since he is in the UK, doesn't he have to pay for his email?
: Wouldn't an effective deterant to him sending you email be to bounce
: the email?  After all, it wasn't requested mail, so it should go back
: to him, right?  Return to sender and all that.

He is using an acount on nyx.cs.du.edu to post his most unpopular
contributions. My guess is that he does that to avoid flames coming
to his main acount on cam.ac.uk. (Maybe he even has his mail on nyx
forwarded to /dev/null). I think that is an abuse of the privilige
accorded to him by the good people at cs.du.edu.

We'll see if he uses the same MO again. If he does, you know where
to direct your flames.

-- 
|< Dan Naas        dan@oea.hacktic.nl >|
+--------------------------------------+

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

Crossposted-To: comp.os.linux.help
From: root@fusion.cuc.ab.ca (Superuser)
Subject: Slackware 1.2.0/Linux 1.0 Problems
Date: Sat, 26 Mar 1994 19:00:04 GMT

Slackware 1.2.0 and Linux 1.0.  Good concept, poor execution-

Several points:

1- Is it just me or is emacs 19.22 buggy as hell in addition to being about
   2 times slower than version 18.59 that I'm using right now?  For example,
   I've had emacs 19.22 crash with a segmentation violation, and then it
   would refuse to work again, instead, on subsequent invocation just eating
   memory until virtual memory is exhausted.  The only way I've found to
   "fix" this is to reboot.  (this problem may be related to #3 , but I'm
   not certain).  Also, under X (both fvwm and olvwm), emacs 19.22 opens
   a window that's twice as wide as is needed, and everything you type is
   initially displayed by emacs as s o m e t h i n g  l i k e  t h i s.

2- Where's the link kit objects for the SVGA server?  The only thing that's
   in the drivers directory is the stuff for the s3 server..  I'd like to
   relink the SVGA server to remove all the drivers for video cards I don't
   have.  (yes, I know they don't take up much memory when running because
   they're not loaded in, but it would be nice just to save the couple hundred
   kbytes of disk space).

3- Some wierd buffer cache interaction problems- (this may be an emacs
   problem, #1)- once in a while, if a change is made to a file in emacs,
   when you save and exit, the change is not visible in the file afterwards
   (ie, more/less shows the old version, running emacs again brings up
   the old version).  Typing 'sync' to flush the buffer caches makes the
   changes register.  This is reproducible on my system, but not consistently.

4- The authors of pkgtool should check how the remove function works-
   try something other than grep to search for files existing in other
   packages- for example the files doctor.el (in elisp1.tgz) and
   doctor.elc (in elispc.tgz) are *NOT* the same, and the .el file may
   be safely removed.

5- The color 'ls' is just too slow for me, even if you turn off the color..
   first order of business was to replace the 'ls' binary with the old
   monochrome-only 'ls'.  Besides, "real" Unix types hate color.


Oh well, unless someone can give me a hand (particularly with items #1 and #3)
I'm going to have to back to 0.99pl9..

c4
-- 
Christopher Lau- "Mr. Unix"    |     /       Fusion: Playing With Fire!
StarBright Research            |    / /      H + H -> He + 24 MeV
            --                 |   /_/_/_    "Bring back Trudeau!"
root,lauc@fusion.cuc.ab.ca     |____________ "This space for rent"

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

From: csmith@convex.com (Chris Smith)
Subject: Re: Pentium Optimized GCC
Date: 26 Mar 1994 23:46:30 GMT

   Newsgroups: comp.os.linux.misc
   From: jwest@jwest.ecen.okstate.edu ()

   I have read several places that a side-effect of Pentium optimized 
   compilers under DOS/Windows is that they produce code that also runs 
   significantly faster on a 486. Does anyone know if this is also true of
   the Pentium-optimized GCC? Or did the GCC developers already know better 
   than the DOS type compiler-writers how to take advantage of the 486 in
   the first place?

The better-486-code rumor is not really true.  They [Watcom] had a better
compiler at the time of the Pentium benchmarks, which was also a better
486 compiler.  But not because of Pentium optimizations, it was just a
better compiler.

The Intel gcc compiler is highly tuned to produce great Specmark numbers.
Which has good effects on other code too, and good effects on 386/486
Specmark numbers.  But for questions like 'what optimizations shall
we do here', 'what order does this need to happen in', etc, etc, there
was a very simple (too simple) criterion.

GCC isn't a terrific x86 compiler, because the architecture is too
strange.  The really good x86 compilers all do only that.  It's sort of a
tribute that GCC can do a reasonable job at all, but it's never going to
beat first class one-target compilers.

A lot of Intel's work on gcc was in the machine-independent part, and
a lot of it was done in a one-target way.  That's why it hasn't shown
up in large quantities in gcc releases, it doesn't necessarily do
reasonable things for non-x86 targets.

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

Crossposted-To: comp.os.386bsd.misc
From: smcarey@mothra.syr.edu (Shawn M Carey)
Subject: Re: Impressions: FreeBSD vs Linux
Date: Sat, 26 Mar 1994 23:28:02 GMT

In article <2n1l3n$821@clarknet.clark.net> rob-n@clark.net (Rob Newberry) writes:
>...
>IMHO, *BSD has a LONG way to go before its ready for users like me.  I
>guess what they need most is a kind of Slackware install, where someone
>can get the system and many utilities installed quickly.  I was

I've done many installations of FreeBSD, haven't tried NetBSD yet...

>VERY VERY disappointed to find that, even though I downloaded every
>distribution file for NetBSD at iastate.edu, I didn't even have a good
>way to talk to my modem -- kermit is not there, I can't get tip
>to work, and there's just not any instructions anywhere.  Plus, when

cu/tip are in FreeBSD.  Now I don't know what version of NetBSD you
were trying, but I'd expect it's cu to work as expected.  All the
manpages I've seen for these commands are complete and correct.
If you can't even look at a manpage long enough to see the '-l'
option to cu(1) then I have no sympathy.

>there are instructions, they're wrong.  The FAQ for makeing a new
>*BSD kernel tells me to switch to a directory that doesn't exists on
>my machine, and config a file that isn't there.  Yes, I did manage to
>find the right place, and create the GENERICISA file from the 

Where did you find it?  This is truly a mystery to me...

>GENERICAHA, but it wasn't in /sys/i386/conf.  Plus, there's no
>description of all that junk in the configuration file -- the FAQ
>says, "Perhaps someone should tell us what all these options actually
>mean."  I thought that's what the *!@# FAQ was for!

Agreed.  FreeBSD-1.0 has a /sys/i386/doc directory that describes most
of the common entries for a kernel config file.  I can't speak for
NetBSD on this, though.

>It's not just the FAQ's, either.  The man pages are screwy.  I wanted
>to add some users.  In linux, "adduser" does the trick.  So I do a 
>"man adduser", and it says that this is a command for adding new users,
>and furthermore that it has been around since 3.0BSD.  Great! So I
>try "adduser" and whammo! "adduser: Command not found."  Yes, I
>did finally find out to use "vipw" (Thanks, O'Reilly), but that's
>just plain dumb.

Sorry, the adduser manpage says it's a *PROCEDURE* for adding new
users!!!  I think you misread it?

>utility programs that I never even used.  In *BSD, none of the things
>I was looking for were there -- no virtual consoles, no comm package,
>no "adduser", no good mail reader, no "pico", no nothing. *BSD may
>have fine networking code, but I won't know about it until they make
>it easy enough to use.

Oh, come on!  virtual consoles ARE there. See above about comm packages
(tip/cu).  What's the matter with mail(1) for a reader, anyways?

>Now, I do realize that many of the *BSD people like it fine.  But I
>will bet that they have a firmer grasp of Unix sys administration

I was a total DOS-weenie when I first installed FreeBSD (after trying
linux).  I liked the fact that it didn't have a slew of things that I
didn't know how to use or what they were for.

>than I do.  I won't say its a bad system...I don't know.  I will say
>that its packaging and installation process for novices like
>myself is absolute crap.

Why?  Because it doesn't include everything under the moon?  Some
people don't wan't to have 1000 utilities on their system and know
what only 5 of them are for.  With FreeBSD I can get the source
and build any utility I find a craving for.

>...
>
>Rob Newberry
>rob-n@clark.net
>

-Shawn Carey

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

From: nemosoft@void.tdcnet.nl (Nemosoft Unv.)
Subject: Re: Maximum serial port speed (REPOST due to non-propogation)
Date: Sat, 26 Mar 1994 21:45:04 GMT

Bill Heiser (bill@bhhome.ci.net) wrote:
: nemosoft@void.tdcnet.nl (Nemosoft Unv.) writes:

: ># We can't get faster than this...
: >115200 #  B50 # B50 CLOCAL SANE -ISTRIP # Fast login: # 115200

: Why not?  Does this mean we wont' be able to use 230000 for use
: with Hayes special serial card and 28.8 modems?

Hmm, the 16550A can go faster, yes. But, first I doubt it if you can get a
effective 7.9:1 compression on your modem all the time; second that requires
that the 16550A runs at a higher clockspeed, and I don't think there are
boards yet with such features, plus you'll need different division rates for
the "old" speeds like 2400/9600/etc. Incompatability all around !

: Also, I hope your question about your changes being moved to the
: kernel sources comes to pass.  This could be a real hassle, making
: all of these changes every time we upgrade to a new kernel!

Every time ? They only need to be patched once. 

: -- 
: Bill Heiser:    bill@bhhome.ci.net,  heiser@world.std.com

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

From: nemosoft@void.tdcnet.nl (Nemosoft Unv.)
Subject: Re: Maximum serial port speed (REPOST due to non-propogation)
Date: Sat, 26 Mar 1994 21:57:50 GMT

Rob Janssen (rob@pe1chl.ampr.org) wrote:
: In <1994Mar23.011834.6267@void.tdcnet.nl> nemosoft@void.tdcnet.nl (Nemosoft Unv.) writes:

: >One point: joe gets the current speed from it's termios structure, and by
: >using B50 and B75 as the "speeds" its output becomes extremely sloooooow.

: That's why you shouldn't do it like that...  it creates unexpected problems
: with some applications.

Yes, I admit that (how can I not ?). However, it's the only application
sofar that has given problems, and there was an easy workaround for it.

: >Maybe this could be stuffed in permanently in the new kernel sources ?

: No.
ah... *sniff* :-)

: There is an extension available in the Linux kernel that allows you to
: set 57600 and 115200, and in fact even any speed that can be made by
: integer division from 115200 (which is the base frequency of the baudrate
: generator). 

: setserial can re-define the 38400 entry to be 57600 or 115200.  to use
: other speeds, you need some own code.  I have modified the "getty" and
: "DIP" that I use to support the random speeds, and now I can run 76800
: on my login/slip line (with a modified serial card and a ZyXEL).

Well, I consider 'setserial' a big kludge. Telling your kernel that "38400
on that port really should be 57600 or something like that" is really weird.
Not to say confusing.  And with these patches, you contradict yourself: now
you need also to patch some programs. With mine, all you had to do is select
a speed of 50 (bizarre, I know), which every program understands, and there
you go at 57600. The other speeds in between are nice, but not really
needed, I think.

[patches removed for bandwidth]

 - Nemosoft.


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

From: jvoosten@compiler.tdcnet.nl (J.S. van Oosten)
Subject: Re: How do I remove LILO after ptting DOS back on my machine
Date: Sat, 26 Mar 1994 22:13:09 GMT

Jim Anderson (anderj@mun18.fi.gs.com) wrote:

: J> RTFM (or better: RTFF): On MS-DOS do "FDISK /MBR". This quite
: J> permamently removes the Lilo booter from your disk.

: If, after "FDISK /MBR", you want to again have linux become the
: default booted OS (assuming, of course, that it's still on the drive),
: how do you reinstate the linux MBR?

By running Lilo again, of course :-). Ok, getting to your linux back in the
first place may be a bit of a problem, you have to boot from floppy then.
Or: before using FDISK save the MBR with a utility like DE from the Norton
Utilities to a file, and restore it when you need it. But be careful with
this !

J. v. O. 

--
Sometimes I wonder if this whole universe isn't anything more than a
simulation on a BIG computer. And they just might press Ctrl-C any time
now...
-- 
My PGP public key [version 2.3] (you know when, why and how...) :
mQCNAi1lYqsAAAEEAMCgUKS7DxyGF8D7QIGYXxRuh2n9Q2+5gIrrb1n9iOl4Xlgo
cO8Y3DE71J5K6WhlpEGDqXZIwY/Xx8mxq80ZHJ3n0pHOUxOQGdxxMT1mrKotjE4Y
wmGqnQhMhpcCKgT/5+5xhuMEluyGQqjyud3PCDogJCC/Sia7eO9+56e/13btAAUR
tC1KLlMuIHZhbiBPb3N0ZW4gPGp2b29zdGVuQGNvbXBpbGVyLnRkY25ldC5ubD4=
=3brb

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


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