Subject: Linux-Misc Digest #611
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:     Fri, 28 Jan 94 22:13:12 EST

Linux-Misc Digest #611, Volume #1                Fri, 28 Jan 94 22:13:12 EST

Contents:
  Re: Anybody using a UPS? (Steve Dawson)
  Re: MSDOS Better than Linux (robert logan)
  Whats the JANA/Morse address? (robert logan)
  Re: Wierd slow machines... (Nicholas Ambrose)
  Arrow keys not working in elvis. (Martien Hulsen)
  Re: port DOS games to Linux..? (really: micro sleep) (Rick)
  Re: Linux Pentium workstations $3589 (Donald J. Becker)
  Re: accessing DD MS-DOS floppies (Benson L Chow)
  Re: port DOS games to Linux..? (Harm Hanemaaijer)
  Re: Linux as X-Terminal? No! (Dominik Kubla)
  can't compile math stuff w/SLACKWARE install (JEREMY)
  Comments on JFH II's Shadow passwords software (Steve Allen)
  Re: MSDOS Better than Linux (Scott Derrick)
  Re: MSDOS Better than Linux (Bob Bagwill)

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

From: sdawson@sbcs.sunysb.edu (Steve Dawson)
Subject: Re: Anybody using a UPS?
Date: 28 Jan 1994 16:41:10 GMT

In article <759704815.44snx@work.invlogic.com>, mmclagan@work.invlogic.com (Mike McLagan) writes:
> In article <CK8xEr.7In@cuug.ab.ca> barkers@cuug.ab.ca writes:
> >
> >[...]
> >
> >What's the popular opinion? How many people have or feel a need for a UPS?
> >
> [...]
>
>    While we're on the topic, are any of the UPS's smart communications 
> supported by Linux?  Or is it a write it yourself proposition?
> 

There is support for UPS monitoring in the SysVInit-2.4 package.
I'm not using SysV init, and didn't want to switch to it, because I didn't
want to worry about other things that may depend on which init is used
(I don't know that there are any such dependencies; I just didn't want to
worry about it), so I wrote my own simple daemon that checks the UPS and
calls shutdown directly when the AC power fails.  I also modified shutdown
to tell the UPS to power off when the system is finally down (in the event
that the system went down due to power failure).  I didn't know anything about
init, shutdown, etc. before doing this, so my point is just that the
"write it yourself proposition" isn't so bad in this case.

-Steve Dawson
sdawson@sbcs.sunysb.edu



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

From: rl@dmu.ac.uk (robert logan)
Subject: Re: MSDOS Better than Linux
Date: Fri, 28 Jan 1994 12:31:23 GMT

Julian D Glover (univ0020@black.ox.ac.uk) wrote:
: In terms of real world work you lusers should realise that MS-DOS and
: MS-Windows is far better than some half assed Unix toy, get a life and
: pay for your software like everyone else you spongers.

Obviously the quality of OxBridge intake has gone down
over the past few years. I must attribute this to the
UK government - right wing pig-dogs that they are. Dont
flame this poor soul - he doesnt know or understand - he
is just part of the UK higher educational pulp machine.

bert
--
Where's your other hand.   | 
Between two pillows.       |     rl@dmu.ac.uk 
Those aren't pillows!      |

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

From: rl@dmu.ac.uk (robert logan)
Subject: Whats the JANA/Morse address?
Date: Fri, 28 Jan 1994 12:38:28 GMT

I recently(*) ordered a copy of the CDROM offer
from some people at morse.net - Ive lost the address
and cant get in touch. It hasnt arrived - although
theyve debited my credit card for the cash as of
22nd Dec 1993. Im in the UK, and it takes some time
to get US deliveries - but not this long - opinions,
email address or whatever would make me happy.

Ive got a floppy (ha) distribution but I want a hard
one - this is getting annoying.

bert - not too coherent today.
--
Where's your other hand.   | 
Between two pillows.       |     rl@dmu.ac.uk 
Those aren't pillows!      |

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

From: na2@doc.ic.ac.uk (Nicholas Ambrose)
Subject: Re: Wierd slow machines...
Date: 28 Jan 1994 12:35:59 -0000


In article <2i9ckv$1uo@canoe.gandalf.ca>, ptomblin@gandalf.ca (Paul Tomblin) writes:
|> In my job, I've got 486DX2/66s that turn in about 32 bogomips, and 486DX/33s
|> that run about 16 bogomips, one 386DX/40 that's just over 7, and several
|> cobbled together Intel Motherboard 486DX/25s that run 0.64 bogomips!
|> 
|> Any ideas what's wrong with the 25s?  What should I be looking for?
|> 
|> Paul
|>
possibly cache turned off ? or perhaps try pressing the 'turbo' button ?
Nick
-- 
Every creature has within him the wild, uncontrollable urge to punt.

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

From: martien@dutw85.wbmt.tudelft.nl (Martien Hulsen)
Subject: Arrow keys not working in elvis.
Date: Fri, 28 Jan 1994 09:36:26 GMT

When I do a shell escape in elvis with 

  :! <any shell command>

my arrow keys do not work anymore. This problem only appears in a
terminal window of X11/fvwm (rxvt, xterm or color_xterm) not under a 
console login.  Does anyone has a solution for this? It is rather annoying.

For the record: I use Slackware 1.1.1.

--
Martien Hulsen                                tel: +31-15-784194
Delft University of Technology,               fax: +31-15-782947
Laboratory for Aero and Hydrodynamics,
Rotterdamseweg 145,
2628 AL Delft,  The Netherlands.              email: martien@dutw9.tudelft.nl

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

From: pclink@qus102.qld.tne.oz.au (Rick)
Subject: Re: port DOS games to Linux..? (really: micro sleep)
Date: Fri, 28 Jan 1994 07:54:02 GMT

bassman@isoit034.bbn.hp.com (Mr. Bassman) writes:

>       How do I make my process sleep for a very small, but accurate amount
>of time ?  Did I miss something somewhere ?  (Probably...)  It'd be useful
>to be able to set alarm() in the same way. I know I could read the clock
>and do loads of compares, but this is busy-waiting, and I'd rather have the
>process sleep (for several reasons).

There have been a number of followups to this post, showing the usual
ways of getting sub-second delays on U*X systems.  However, be aware
that Linux is not a real-time operating system, and it's unlikely that
the word `accurate' is going to be applicable.  Apart from interrupt
latency and process prioritization, the accuracy is also going to be
determined by the quantum of the jiffy, the unit in which the kernel
keeps track of time.  I haven't looked at this code in a long time, but
I believe that it's a fairly large time interval (relative to micro or
milli second accuracy, that is).

However, if you want to sleep for *close* to 0.27 seconds, the posted
solutions should be satisfactory.

Rick.

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

From: becker@super.org (Donald J. Becker)
Subject: Re: Linux Pentium workstations $3589
Date: Thu, 27 Jan 1994 05:25:53 GMT

In article <1994Jan24.192041.27713@msc.cornell.edu>,
 <tan@msc.cornell.edu> wrote:
>What _I_ wonder is: do I wait for a few months for Pentium prices to 
>drop, or settle for a 486/66 now and put the money in RAM?  Sigh.

I considered this issue recently while putting together a proposal for a
Linux-based cluster workstation.  The metrics used for selecting a cluster
might provide you with a rational basis for a (usually irrational)
single-processor purchase:

        As in many other markets, performance drops sharply at the low end
        and prices rise steeply at the "tech-groupie status-power-user" high
        end. Doubling the price for an extra 20-30% performance (cf. DX2-66
        to Pentium) isn't worth it when you can just add more processors.

        Getting the fastest system, as long as the price-performance remains
        linear, will minimize overhead.

My conclusion was that the best performance/price was a 486/66 PCI
motherboard with built-in SCSI-2 (1).  That's available for $900, while
Pentium motherboards are close to $2K.

Also under consideration was the $250 IBM SLC2-66 ISA motherboard, but that
wouldn't be a good match for the planned 100Mbs ethernet, has a separate
FPU, and has a less-obvious migration path to a clock-tripled processor.
This would be a great processor coupled with a 10Mbs network, and moderate
memory and disk (say, 8M and 340M).

A clustered workstation is the kind of system that demonstrates the
advantages of Linux.  SRC has two clusters of sparc workstations, one each
of 16 and 32 processors connected by a Kalpana etherswitch.  While cost was
not a consideration in their construction, imagine building a similar
low-cost system.  Clone Sparc-2 (2) processor boards with on-board SCSI-2 and
ethernet(3) are only around $1600, but that doesn't include the per-CPU
SunOS RTU license.  Add to that the difficulty/expense of getting and
modifying SunOS.  Technology as young as cluster multiprocessing will
require frequent updates, so there might be additional SunOS version upgrade
license fees per-CPU.  The software complications and expenses quickly make
a cluster unattractive.

(1) The motherboard, from MS Engineering, also has 3 PCI slots, IDE, FDC,
2S, 1P etc., but the on-board SCSI-2 is the winning feature.

(2) A modern Sparc-2 processor board is somewhat slower than a DX2-66 for
"typical" applications.

(3) A transceiver for the Sparc board will cost about the same as a ISA bus
ethercard ($30 vs. $40).  Also 100Mbs ethercards will be cheaper and
available sooner for EISA|VLB|PCI than for SBus.
-- 

Donald Becker                                          becker@super.org
IDA Supercomputing Research Center
17100 Science Drive, Bowie MD 20715                        301-805-7482

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

From: Benson L Chow <bc3c+@andrew.cmu.edu>
Crossposted-To: comp.os.linux.help
Subject: Re: accessing DD MS-DOS floppies
Date: Fri, 28 Jan 1994 15:18:42 -0500

Apparently, an /etc/mtools with:

A /dev/fd0 12 0 0 0
B /dev/fd1 12 0 0 0

seems to autodetect... But don't mformat a: or mformat b:, it will fail
in this respect.

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

From: hhanemaa@cs.ruu.nl (Harm Hanemaaijer)
Subject: Re: port DOS games to Linux..?
Date: Fri, 28 Jan 1994 14:22:32 GMT

In <1994Jan25.191911.12367@nicmad.uucp> wanggaar@nicmad.uucp (Mike Wanggaard ) writes:


>After recently receiving a copy of DOOM, I began to wonder what it would take
>to port a game of that magnitude to Linux native.  It seems so painful to
>think that they are both running on the same instruction set, and yet I
>can't run DOOM from within Linux.

As another poster pointed out, a Linux port of Doom (either X or raw VGA
(svgalib)) is not totally out of the question.

>I am looking for any insight on the topic; I have very little programming
>experience in Unix (and none in DOS) but I am interested in at least a 
>superficial way.  The following is a list of questions I think explain my
>question a little better:
>
>How does multitasking affect game programming?

For a real-time animated VGA type of game, to have smooth animation you
will have to have no CPU-hogging processes running. Usually, if you are not
running a compiler or having network traffic, everything will be sleeping
and basically not take any CPU time. Running CPU-intensive stuff in the
background while doing real-time animation causes 'bursts' of animation
frames. A game can also sleep if there's no animation (waiting for
keyboard/mouse input).

>How would the executable format of DOS vs. Linux affect things?

If you define the DOS executable format to be 16-bit real mode code, then
the Linux executable is faster and may also be smaller because of shared
libraries. The big win for Linux is to have a couple of megabytes of linear
memory to play with and an optimizing C compiler.
As for running a DOS executable in Linux, DOSEMU does a good job of it but
I/O is virtualized and thus slow.

>Would packages such as SVGALIB or VGALIB be able to provide the necessary
>graphics routines (or perhaps even better)?

Yes. All you need is to set a mode and to copy a framebuffer to the screen.
All of the graphics drawing is done in the system memory framebuffer. 

>Am I mistaken in believing that no Linux games have been developed?
>(I know about xtetris and mahjongg; I mean big games like this...)

Well, there's the Boulderdash game which uses svgalib, and there 
are people interested in programming VGA-style games for Linux. I would say
it's much easier than in DOS.

>On a side note...what is the DOSextender of DOOM doing?

Letting DOOM run in a processor mode that is not flawed by a 15 year old
instruction set design. It's basically the same mode that Linux runs in.

hhanemaa@cs.ruu.nl

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

From: kubla@bernhard.zdv.Uni-Mainz.DE (Dominik Kubla)
Crossposted-To: comp.windows.x.i386unix
Subject: Re: Linux as X-Terminal? No!
Date: 28 Jan 1994 19:02:37 GMT

In article <1994Jan26.140307.21474@taylor.wyvern.com>, mark@taylor.wyvern.com (Mark A. Davis) writes:
"> "
"> "But how would one do that?  Can you get a Linux boot ROM for an ethernet
"> "controller?  Most people wanting to turn a clone into an Xterminal had
"> "been looking at trashing local media and putting tiny-Linux on ROM.  The
"> "remote boot idea is tremendously more logical, as it would require much less
"> "ROM; and it would follow the way most Xterminals get their Xserver files
"> "anyway....
That is possible! The same code can be used to either produce a bootable
floppy (useful demonstrations), an DOS executable or ROM boot code.
Look at the netboot package for linux. The author designed it to be easy
extendable to other bootimages, eg NetBSD or something completely different.

All it needs are Turbo-C and Turbo-Assembler.

-- 
Cheers,
  Dominik

+---------------------------------------------------------------------------+
| eMail: kubla@goofy.zdv.Uni-Mainz.DE                                       |
| sMail: Dominik Kubla, Steinsberg 34, 56355 Nast"atten, F. R. Germany      |
+---------------------------------------------------------------------------+
|                                                                           |
|        "Linux: The choice of a GNU generation"      --S. Frampton         |
|                                                                           |
+---------------------------------------------------------------------------+
DISCLAIMER:  Everything written above are the expressed thoughts of the
author and in no way connected to 'Johannes Gutenberg Universit"at', Mainz
(Germany). This way, they do not have to care about what I say ...

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

Subject: can't compile math stuff w/SLACKWARE install
From: z_huffmanjj@ccsvax.sfasu.edu (JEREMY)
Date: 27 Jan 94 19:51:47 CST


i've ben trying to compile some simple math equations on my linux box, but
have been running into some problems.
for example: foo = pow(x,y);  comes back with 
 /usr/tmp/cca001801.o: Undefined symbol _pow referenced from text segment

anyhow i checked out math.h and everything looks good there.

i've tried on two linux boxes both have Slackware 1.1.1. & GNU C, C++ v2.4
i never had this problem w/ SLS 1.03 & GNU v2.4.

any help would be greatly appreciated,

jeremy
--
jeremy@devo.sfasu.edu

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

From: sla@umbra.UCSC.EDU (Steve Allen)
Crossposted-To: comp.os.linux.admin
Subject: Comments on JFH II's Shadow passwords software
Date: 28 Jan 1994 19:41:08 GMT

In article <JRS.94Jan28024832@lepton.world.std.com>,
Rick Sladkey <jrs@world.std.com> wrote:
>Hacking is fixing bugs in other people's programs.  Hacking is its own
>reward.

Absolutely.

However, for my purposes I need a shadow password suite which will operate
on SunOS, IRIX, OSF/1, AIX, Linux, and others yet-to-come.  I don't have
time to maintain a different scheme on each kind of system.
I find that JFH's shadow is my best option across all platforms.
It is not perfect for every system, but it is continuing to evolve to
suit the needs of these many systems.

The login and authentication procedures for an OS are a region where
"there be dragons" which are different for every vendor--and different
for every OS release from some vendors.  Getting a shadow password code
which works on all of these systems requires a synoptic approach.
It is very important that a feature for one system not have a side
effect which makes it impossible to recover from a disaster on another.
Helpful hacks intended for one system can easily break another, or
compromise the security which shadow should provide.

I hope that the Linux community doesn't find it necessary to reinvent
shadow.  The result would probably take years of tweaking before it became
useful on the broad spectrum of other systems where JFH's code now works.

I've diverged this thread too far to followup in a Linux group.
Followups and flames to /dev/null
--  --  --  --  --  -+-   --  --  --  --  --  --  --  -+-  --  --  --  --  --
Steve Allen          | "Shameless Morris Enthusiasm"   |   sla@lick.ucsc.edu
UCO/Lick Observatory | Yep, that's me.  6 up for       |   +1 408 459 3046
Santa Cruz, CA 95064 | Orange in Bloom, Sherborne, eh? |   Seabright Morris

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

From: scot@as215-ws-15.ucsc.edu (Scott Derrick)
Subject: Re: MSDOS Better than Linux
Date: 28 Jan 1994 19:43:45 GMT
Reply-To: scot@cats.ucsc.edu


   Julian D Glover (univ0020@black.ox.ac.uk) wrote:
   : In terms of real world work you lusers should realise that MS-DOS and
   : MS-Windows is far better than some half assed Unix toy, get a life and
   : pay for your software like everyone else you spongers.

   Actually, I think be brings up a valid point.  As another netter pointed
   out, a lot of people actually don't like Linux.

   The main usage of MS-Dos at places, are games (which Linux has), and more
   importantly "hot" software packages (mucho dollars) like Excel, WP, etc.
   When DOSEMU is full-fledged, this problem will be solved.  And Linux can
   move onto compete in that market.

I agree wholeheartedly. I'm new to Unix(1 yr) and even newer to Linux.
I would only use Linux as my OS if it would run allof my favorite
Apps. Or at least offer reasonable alternatives. I use a speadsheet,
QP for Windows, and a database Paradox, and can't run either of them
on Linux. There are no alternatives that come even close to either of
these apps. 

As distastfull as Windows or Dos is to most off the Linux community,
Linux will never be a serious contender until it offers emulators or a
whole slew of alternative applications. I think teh emulator option is
the only way to go. There is just too much good software out there
written for Windoz & Messy Dos.  

I personally think that OS/2 is more on the right track than NT or
Linux.  But its such a butt ugly environment, and a pig to boot.




--
===============================================
Scott Derrick              | Yahoo Productions
scot@cats.ucsc.edu         |
(408) 335-7373             | "Make it so!"
===============================================   

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

From: bagwill@sst.ncsl.nist.gov (Bob Bagwill)
Subject: Re: MSDOS Better than Linux
Date: Fri, 28 Jan 1994 18:01:45 GMT

Kumar K.E.V. (kkev@professional.kkev.dome) wrote:
: Here at UMKC, we are getting a "new" 486DX2/66 lab, and we are
: contemplating running UN*X (Linux, maybe)( on all of them. The problem
: is that this lab is "general purpose" and can be used by any UMKC
: student, UN*X would probably not be a good choice. WABI is one option.
: But, Windows is a lot better, for that purpose.

Don't forget that you can split the disks, with DOS/Windows on one
partitiion and Linux on the other. LILO will let you boot either
partition. It only takes a ten seconds to reboot DOS, and 45 seconds
to boot Linux (less if you don't include self-test time and don't have
SCSI).

--
Bob Bagwill
rbagwill@nist.gov

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


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