Files
oldlinux-files/ftp-archives/tsx-11.mit.edu/1996-10-07/mail-archive/linux-devel/Volume2/digest151
2024-02-19 00:24:15 -05:00

566 lines
23 KiB
Plaintext

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: Sat, 10 Sep 94 18:13:14 EDT
Subject: Linux-Development Digest #151
Linux-Development Digest #151, Volume #2 Sat, 10 Sep 94 18:13:14 EDT
Contents:
Re: Future of linux -- the sequel (Anthony J. Stuckey)
Re: Don't use Linux?! (Stephen Harris)
Re: Survey: who wants f77,cc,c++,hpf for linux? (Steve Allen)
Re: Survey: who wants f77,cc,c++,hpf for linux? (Bill McKinney)
Re: Linux for DEC Alpha platform?
Re: 320x200 X resolution? (CLAYTON MICHAEL O'NEILL)
Re: News Spool File System - new filesystem type?? (Arthur Tateishi)
Re: 320x200 X resolution? (Christopher Wiles)
Re: Linux Kernel's and ATDISK-Patches (Delman Lee)
Re: Don't use Linux?! (Bogdan Urma)
Developing Distributed Filesystems for Linux? (Lincoln Myers)
Re: Survey: who wants f77,cc,c++,hpf for linux? (Hiroshi)
----------------------------------------------------------------------------
From: stuckey@mrcnext.cso.uiuc.edu (Anthony J. Stuckey)
Subject: Re: Future of linux -- the sequel
Date: 8 Sep 1994 21:27:37 GMT
shendrix@escape.widomaker.com (Shannon Hendrix) writes:
>My little ISA bus 486 has little trouble besting low-end workstations
>in anything (Sun 4/110's and Sparcstation 1's). It will do some integer
>ops faster than a Sun 2. The Sun 2's FPU will eat mine alive but for
>$10000 it should beat it on *ALL* counts. It doesn't.
Ahem. Correct terminology is probably necessary here. A Sun 2 is a
68010-based box, as I recall, which won't eat any 486 alive, even a lowly
SX. This is particularly odd, because you do refer to the Sun 4
correctly.
A SparcStation 2, on the other hand, is a fairly nice machine.
>I'd love to have a workstation but given that by the time I can afford
>one there will be little or no difference between a PC and a workstation
>(as if there really is one now) so who cares? Workstations are generally
>overpriced still. Software for them is even worse.
I'm going to stay out of this side of the argument. I'm unconvinced
that workstation hardware is overpriced, given that you frequently do get
better capabilities than PC-stuff will give you, particularly on I/O. I am
also not convinced that many people or companies take multi-user-ness into
account when pricing their software. Researched, cogent arguments for any
of the sides would be welcome. Flamers should be shot.
--
Anthony J. Stuckey stuckey@mrcnext.cso.uiuc.edu
"And if you frisbee-throw a universe where does it go?" -- Steve Blunt.
GCS/S -d+@ p c(++) l u+ e+(-) m+(*) s+++/-- !n h(*) f+ g+ w+ t+@ r y?
KiboNumber == 1
------------------------------
From: hsw1@papa.attmail.com (Stephen Harris)
Subject: Re: Don't use Linux?!
Date: Sat, 10 Sep 1994 12:07:16 GMT
Michael Schumacher (hightec@sbusol.rz.uni-sb.de) wrote:
: 1. Commercial software products are typically binary-only (i.e., no
You admitted this is not a problem because of shared libraries, and then...
: 2. Linux's libc tends to change its version number almost every week
: (sometimes even more often). Even though changes of the minor
Really? Honest? The *ALPHA* versions change rapidly, yes. Public releases
that most people are using?
Well, 4.5.26 was released Apr 4.
That means it has been stable for over 5 months. Not quite every week :-)
More people are using the stable versions of libc than the alpha versions.
: 3. The kernel versions change faster than the speed of light. If you
: ask for a "stable" version, you'll be teached that there are two
: versions: 1.0 (production) and 1.1 (hacker's paradise). Wanna have a
: stable one? Get 1.0! Okay, but if I want to offer a commercial
: product, it doesn't matter what kernel version *I* am using, but
: what version is used by my potential *customers*! There's a reason
The majority of applications built to run under 1.0.9 with libc 4.5.26 will
continue to work. Hell, 99% of my system was compiled with kernel 0.99pl12A
with libc 4.4.2. And they STILL work.
(talking of 0.99pl12A, I've had that running on 5 client sites now for a year,
and the biggest problem we have had is HARDWARE failure! One day I'll have
to upgrade them :-) )
: 4. The spirit of free software is all around. Free in both meanings:
: free availability of the sources, and free of charge. Which does
[ writing shareware and not gettting any registrations ]
On the other hand it could be that no-one thought your software was worth
while. If common commercial software ran seamlessly under Linux then it
would actually boost linux in the workspace. I bet that if Lotus 123,
WordPerfect and a few other common apps ran under Linux then Linux would be
in the running as a transition from DOS kludged networking to Unix.
: 5. On the other hand, I can tell you how to make lots of money with Linux:
[ making CD copy of FTP archives and selling it ]
: Linux. He buys a "Dream Linux" CD - and is lost. Nothing works "out of
: the box", no reasonable documentation is available, nor hotline support.
: What will happen? I'm quite sure that most of these desperated people
: will close the Linux chapter - forever.
The majority of Linux users either heard of it from this group, and so ask
'whats a good distribution', or heard of it from a friend who... or a friend
of a friend of a friend... In all cases these people wouldn't just go out and
buy any old CD distribution. People are smarter than that!
: processing is out - try TeX", or "you can run xyz under DOSEMU" or "try SCO
: versions of xyz; just recompile the kernel with SYSV support and get the
: iBSC2 package from foo.bar". This can be - at most - a temporary work-around.
: Users don't want to know how to roll a new kernel, they don't want to ftp
Sigh. Currently the DOSemu and IBCS stuff is still in development. The IBCS
can be loaded as a module. When it becomes stable (it's already looking good!)
then the typical user won't need to do anything!
DOSemu is great to run DOS apps, but it's not seamless, and so is really just
a stop-gap IMHO, but a vital one.
: being used in offices and other commercial environments? If we *really*
: want Linux to succeed, we *need* the companies and their commercial products!
As I posted two weeks - QUIT MOANING AND DO SOMETHING ABOUT IT!
--
rgds
Stephen
------------------------------
From: sla@umbra.UCSC.EDU (Steve Allen)
Crossposted-To: comp.lang.fortran
Subject: Re: Survey: who wants f77,cc,c++,hpf for linux?
Date: 10 Sep 1994 19:00:37 GMT
In article <34ssa9$fav@taco.cc.ncsu.edu>,
Bill McKinney <mckinney@math.ncsu.edu> wrote:
>f2c/gcc works about 2/3 of the time for me. When it does work, I
>am not really sure about how efficient the code is, but it runs.
>The problem is that my original code might not compile under f2c/gcc.
>I don't want to spend time "fixing" codes (that have already been working
>and optimized) so they'll run under f2c/gcc and allow me to do development
>on Linux.
I must admit to being a bit confused by Bill McKinney's remarks here,
because I find that f2c does just fine with any Fortran code that
conforms to the ANSI standard.
When f2c does produce results which differ from commercial compilers,
it has always been the case that there is some non-standard assumption
being made by the code (many of these problems being due to the
extensions added by DEC in its early compilers: assuming unintialized
variables will be zeroed, assuming the linker will pad all common block
references to the same length, assuming all variables are SAVEd, liberal
use of the long-illegal Hollerith to store characters, incorrect ordering
of statements and the other familiar Fortran porting problems).
I know of several respected astrophysicists who do all their Fortran code
development on their Linux laptops, and when it is working they move it
onto their commercial workhorse systems to get results fast.
-- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
Steve Allen UCO/Lick Observatory Santa Cruz, CA 95064
sla@lick.ucsc.edu Voice: +1 408 459 3046 FAX: +1 408 454 9863
<a href=ftp://lick.ucsc.edu/pub/www/sla/sms.html>Seabright Morris & Sword</a>
------------------------------
From: mckinney@math.ncsu.edu (Bill McKinney)
Crossposted-To: comp.lang.fortran
Subject: Re: Survey: who wants f77,cc,c++,hpf for linux?
Date: 10 Sep 1994 18:01:45 GMT
Thomas.Koenig@ciw.uni-karlsruhe.de writes:
>Larry Meadows (lfm@pgroup.com) wrote in article <34m769$bju@indy.pgroup.com>:
>
>>1. Are people interested in a commercial compiler suite for Linux on
>> Intel Architecture platforms? The suite would include true compilers
>> for extended Fortran 77, ANSI C, Draft-ANSI C++ with extensions, and
>> High Performance Fortran.
>
>C: gcc is just fine for Linux.
Ditto.
>F77: f2c/gcc works for the kind of things I do with it. I don't run
> large linear systems on a Linux box, anyway (those go on a large
> HP, or an even larger Fujitsu vector computer), but for sheer
> floating point applications, which operate on little data, a
> 486 doesn't do badly compared to a heavily loaded HP.
f2c/gcc works about 2/3 of the time for me. When it does work, I
am not really sure about how efficient the code is, but it runs.
While I don't intend to use my Linux box for large computationally
intensive programs, it would be nice to have f77 (or HPF :) )
available for development. For example, I have several programs
with thousands of lines of code. If I were to modify them, I would
like to do so in Linux: modify some routines, do any debugging,
do some limited testing for small data sets, etc. Once this is
done, I can put the code on a workhorse somewhere else and do
production runs on a large data set there.
The problem is that my original code might not compile under f2c/gcc.
I don't want to spend time "fixing" codes (that have already been working
and optimized) so they'll run under f2c/gcc and allow me to do development
on Linux.
I love Linux and want to use it as much as possible. It has everything
I want, EXCEPT for a true fortran compiler.
--
Bill McKinney
mckinney@math.ncsu.edu
------------------------------
From: jsmith@red-branch.MIT.EDU ()
Subject: Re: Linux for DEC Alpha platform?
Date: 10 Sep 1994 19:40:52 GMT
Peter Hahn (Peter@tequila.oche.de) wrote:
: becker@cesdis.gsfc.nasa.gov (Donald Becker) writes:
: >In article <33u1fh$9ej@bmerha64.bnr.ca>,
: >Hamish Macdonald <Hamish.Macdonald@bnr.ca> wrote:
: >>>>>>> On 29 Aug 1994 08:14:02 EST,
: >>>>>>> In message <33smuq$mkq@scotty.waldorf-gmbh.de>,
: >>>>>>> ralf@resi.waldorf-gmbh.de (Ralf Baechle) wrote:
: >>
: >>Ralf> Expect Linux to be available for all major CPU families in the
: >>Ralf> near future.
: >>
: >>Ralf, you and I both know that the big work in a Linux port is not so
: >>much the CPU specific stuff, but the device drivers.
: >>
: >>As such, only expect Linux to work on PC clones (powered by various
: >>ALPHA, x86, MIPS chips) in the *near* future.
: >Both the PowerPC/PREP and the Alpha are promoted as using the PCI bus.
: >Picking the three major devices I consider necessary for a system: disk,
: >network, and video controllers, I see that x86 Linux is close to supporting
: >PCI bus versions of all three. I'm only intimately familiar with network
: >device driver, but I don't think it would take very long to convert it to a
: >different processor once I know I knew a few details, such as the I/O space
: >mapping. Perhaps most of the changes will be just converting the ASM
: >in*()/out*() functions to memory operations, and checking for byte-sex
: >problems.
: >The remaining essential device drivers to be written are for the
: >keyboard/mouse port, PCI bus bridge/controller, and perhaps a PCMCIA bus
: >bridge. None are trivial, but getting close to having something working is
: >highly motivating...
: As I understand the PCI specification you can build a bare motherboard with
: a SCSI interface chip, a PCI to isa bridge, some timers/clocks and four PCI
: connectors. Slow I/O devices as mouse, keyboard, serial and Ethernet are
: linked via the additional isa bus slots as usual. Now we plug in a PCI
: graphics card and a CPU module and our system runs. We even have two slots
: left.
: This would be the ideal board to give the CPU manufactures good
: competition. And porting is as easy as possible, as you only have to port
: the parts dependig on the CPU module. I don't want to mention the
: upgradability this solution provides.
: I don't know, if someone offers such a board, though :(
Well I'm getting very close to having workign hardware on the board I'd
mentioned on here before..
Some Rough spec/outline:
MotherBoard Specification (Subject to Change)
1 - 21066 @ 166Mhz cpu
512k - 2Megabyte L2 Cache
8 72-Pin Simm Sockets for Main Memory and ECC
1(2*) NCR53C810 PCI Scsi-II/Fast controller
68EN360 (25/33Mhz) Serial Communications Controller
512K Local SRAM
512K Flash ROM
lotsa stuff to put here about this..
0,1, or 2 ethernet ports
ISDN (optional)
2 RS-232C Serial Ports
Audio Codecs: 2 pairs in, 3 pairs out (right&left channels). Can
playback on one channel pair while sampling from another channel pair,
16 bit at up to 50Khz.
High Speed Centronics Bi-directional Parallel Port.
PCI Super Combo
32-bit addressable DMA controller
Interrupt controller edge or level sensitive, programmable
Real Time Clock with 128-byte configuration ram
Timer/counters
IDE hard disk drive interface capable of fast IDE (11M bytes/s)
Floppy disk drive interface
2 16550 serial ports
Parallel port (Bi-directional)
External system bus for add-ons
21140 {21040} PCI Ethernet LAN Controller
100Mbs or 10Mbs {10Mbs only} onchip selectable Ethernet
Twisted Pair, Co-axial, Fiber ports
21030 PCI 2D/3D Graphics Accelerator (optional)
Frame buffer sizes from 2meg to 16meg Vram
Hardware Z-buffer
PCI to ISA bridge chip (4 or 6 ISA slots)
With a PCI to PCI bridge chip this also includes
4 PCI sockets for general I/O and an additional one for gfx.
: Peter
: --
: Peter Hahn Peterstr. 26
: 52062 Aachen Germany
: Peter@tequila.oche.de pch@pool.informatik.rwth-aachen.de
: Voice: +49 241 37151
Jonathan Smith
------------------------------
From: cs339014@bit.com (CLAYTON MICHAEL O'NEILL)
Crossposted-To: comp.os.linux.misc
Subject: Re: 320x200 X resolution?
Date: 10 Sep 1994 16:22:58 GMT
Christopher Wiles (a0017097@wsuaix.csc.wsu.edu) wrote:
: Seriously, IMHO Doom will probably be more useable in the promised
: pixel-doubling mode than in a straight 320x200. Easier to make things
: look innocent when the boss walks in ... "Hey, you're not actually
: _working_ in 320x200, are you?"
The only problem is that the pixel doubling mode (at least in 256 colors)
looks really horrible. However, I'm _extremely_ impressed by the speed of
the thing. Just two real complaints.
1) I wish 8 bit sound sounded decent. He makes a comment that if you
don't have 16-bit sound, you'll wish you did. I don't and I do.
2) I have an annoying problem w/ the shift keys getting stuck down, so
I get in a situation where I'm always strafing or running fast. You get
the idea.
Clayton O'Neill
------------------------------
Crossposted-To: news.software.b
From: ruhtra@turing.toronto.edu (Arthur Tateishi)
Subject: Re: News Spool File System - new filesystem type??
Date: 10 Sep 94 19:45:45 GMT
In article <m7FSkapDlzU6071yn@halcyon.com>,
Michael Dillon <mpdillon@halcyon.com> wrote:
>> 1) eliminate/minimize directory traversal for article access
>> (opens?).
>
>If a program issues an open for comp/os/linux/development/22334 the file
>system is free to use an implementation that does not involve the
>normal UNIX directory tree traversal.
As a friend of mine wisely noted recently, log file systems are almost
ideally designed for usenet news. Coincidentally, I recently came
across an old paper while cleaning out my office. Someone else may
have more contemporary references to share.
"Beating the I/O Bottleneck: A Case for Log-Structured File Systems"
John Ousterhout and Fred Douglis
Tech Report No. UCB/CSD 88/467
October 1988
Computer Science Division (EECS)
University of California
Berkeley, CA 94720
Undoubtedly, they probably have an FTP or WWW server for tech reports
by now. Track down the CS Tech Report FAQ for the location.
Enjoy
arthur
--
Choices don't scare me. However, a lack of choices does.
Arthur Tateishi ruhtra@turing.utoronto.ca
------------------------------
Crossposted-To: comp.os.linux.misc
From: a0017097@wsuaix.csc.wsu.edu (Christopher Wiles)
Subject: Re: 320x200 X resolution?
Date: Sat, 10 Sep 1994 17:09:36 GMT
slouken@cs.ucdavis.edu (Sam Oscar Lantinga) writes:
: I suppose you speak in ignorance. :) It's out, and
: available from sunsite.unc.edu in /pub/Linux/Incoming and at
: dewdrop.water.ca.gov in /pub/doom as the file lnxdoom.tgz
I posted the night before DOOM showed up at sunsite. Nice timing.
: Well, a couple of things... I've tried it in pixel doubling
: mode, and not only is it slower, but the display is broken. It looks
: like you are looking through one of those windows with glass ribs.
: Also... if you should happen to be one of those that actually play
: games at work, you might want to know that Ctrl-Alt-+ and Ctrl-Alt--
: change X11 resolutions on the fly at the Linux console. :)
The pixel-increase modes _are_ broken. Dammit. And, yeah, res swapping
has so far enabled me to keep my job ;)
-- Chris
a0017097@wsuaix.csc.wsu.edu wileyc@halcyon.com wileyc@quark.chs.wa.com
"... but I want to use all eight comm ports SIMULTANEOUSLY!"
PGP 2.6 public key available by finger for the clinically paranoid.
------------------------------
From: delman@mipg.upenn.edu (Delman Lee)
Subject: Re: Linux Kernel's and ATDISK-Patches
Date: 10 Sep 94 13:25:26
In article <Cvr24L.49@cove.han.de> knick@cove.han.de (Michael Knigge) writes:
Will the atdisk-Patch (up to four IDE-Drives) be "standard" in the next
Linux-Releases? I hope so because I need them and the Patch works without
any Problems.....
Probably not. The old atdisk2 patch is not a very clean solution. The
new atdisk2 patch, which is a cleaner solution, is under testing. If
you feel adventurous, you could try the alpha test code at
mipgsun.mipg.upenn.edu:pub/delman/atdisk2-1.0alpha?.tgz
Delman.
--
______________________________________________________________________
Delman Lee Tel.: +1-215-662-6780
Medical Image Processing Group, Fax.: +1-215-898-9145
University of Pennsylvania,
4/F Blockley Hall, 423 Guardian Drive,
Philadelphia, PA 19104-6021,
U.S.A.. Internet: delman@mipg.upenn.edu
______________________________________________________________________
------------------------------
From: bau1@cornell.edu (Bogdan Urma)
Subject: Re: Don't use Linux?!
Date: 10 Sep 1994 17:46:54 GMT
Reply-To: bau1@cornell.edu
Michael Schumacher (hightec@sbusol.rz.uni-sb.de) wrote:
: Hello Linuxers!
: Okay. Before you start sending me endless flames, I want to make sure
: that you know that I *love* Linux. It's probably the best PC Un*x you
: can find between here and the sun. Linux has some nice features, e.g.
: the /proc filesystem, it is fast, it supports lots of hardware, it
: follows the POSIX standard (which makes porting of existing software
: much easier), plus: it's free. Nobody knows the exact number of Linux
: installations, but it's likely to be in the 100000's. One could think
: that companies are willing to consider Linux a reasonable and serious
: platform, and that they would port and offer their products to the
: Linux community. However, they are far away from doing so, actually.
: Here's why:
That's pretty funny, since Maple V Release 3 has just been ported to Linux.
Bogdan
------------------------------
From: lim@soda.CSUA.Berkeley.EDU (Lincoln Myers)
Crossposted-To: alt.filesystems.afs
Subject: Developing Distributed Filesystems for Linux?
Date: 9 Sep 1994 06:07:40 GMT
I've been reading the AFS FAQ and other reports (available from
grand.central.org:/pub/*) about AFS (formerly the Andrew File System) and
DFS (the DEcorum File System, part of DCE) both of which are commercially
available filesystems which are generally faster, more secure, and easier to
manage than NFS, especially over large networks.
Is anyone working on a filesystem for Linux or another freely available UN*X
which will offer the advantages that these do?
AFS was a project at CMU before it became a commercial product supported by
Transarc. Would it be feasable to port an earlier version (pre-3.0?) to
Linux? (Is it freely available? Would it be compatible with current AFS?
Would one want to use it?)
If not, would it be possible to make a freely available implementation of
AFS or DFS for Linux, without infringing on their current owner's
(Transarc's) rights? Is there enough information out there?
I would be willing to help develop something like this if there is any
demand.
Lincoln
------------------------------
From: hiroshi@cgate.hipecs.hokudai.ac.jp (Hiroshi)
Crossposted-To: comp.lang.fortran
Subject: Re: Survey: who wants f77,cc,c++,hpf for linux?
Date: 9 Sep 1994 03:43:22 GMT
In article <34m769$bju@indy.pgroup.com>, Larry Meadows <lfm@pgroup.com> wrote:
>Given the interest in Linux, I thought I'd post a short survey:
>
>1. Are people interested in a commercial compiler suite for Linux on
> Intel Architecture platforms? The suite would include true compilers
I thought the LINUX has a feature of its AT&T 386 unix binary compatible.
If that is true, I think there are some FORTRAN or C or C++ compiler
for intel386 AT&T unix that is commercially available which will run
on linux. Am I wrong? Has anyone running those commercial compilers
on linux?
------------------------------
** 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
******************************