Files
oldlinux-files/ftp-archives/tsx-11.mit.edu/1996-10-07/mail-archive/linux-devel/Volume2/digest202
2024-02-19 00:24:15 -05:00

630 lines
24 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: Wed, 21 Sep 94 19:13:13 EDT
Subject: Linux-Development Digest #202
Linux-Development Digest #202, Volume #2 Wed, 21 Sep 94 19:13:13 EDT
Contents:
Re: Extending the IP Protocol? (Sam Oscar Lantinga)
Re: 1+ Gig SCSI Drives (Marc Singer)
Re: Shared Libs: working toward a permanent solution? (Neal Becker)
Re: Any Linux MOTIF packages out there? (Iulius M Nicolescu)
Where are Patches 13,16 and 41? (Brian Oliver)
Re: Future of linux -- the sequel (Eric J. Schwertfeger)
Re: How to use a host as a router - READ THIS (Frank van Maarseveen)
Re: Shared Libs: working toward a permanent solution? (Donald Becker)
Re: XFig Eats All My Memory. (Russell E. Dube)
Re: Shared Libs: working toward a permanent solution? (Donald Becker)
Re: Don't use Linux?! (Lutz Behnke HiWi)
Re: Don't use Linux?! (Lutz Behnke HiWi)
Re: How to use a host as a router - READ THIS (David - Morris)
Re: Don't use Linux?! (Lutz Behnke HiWi)
----------------------------------------------------------------------------
From: slouken@cs.ucdavis.edu (Sam Oscar Lantinga)
Crossposted-To: comp.protocols.tcp-ip
Subject: Re: Extending the IP Protocol?
Date: 19 Sep 1994 18:02:05 GMT
Stef Van Dessel (stef@INbe.net) wrote:
: Say you have NET1-ethernet- GATE1 <point-to-point-link> GATE2 -ethernet-NET2
: Then GATE1 needs to proxy arp for each host on NET2 and for GATE2. GATE1
: also needs host routes to these machines via the point-to-point link with
: GATE2. Do the same for the opposite direction. It's trivial to write a little
: script that will snarf the IP's from a file, and do the arp publishing and
: route adding.
I have:
NET1_work-ethernet- Linux_box - NET1_work-ethernet- ROUTER1 -
INTERNET_router_router_router_INTERNET - SLIP_server -serial-Home_Linux_box
I want my Linux box on the network at work to act as a proxy
for my Home_Linux_box, making it appear as though my home Linux box is
on the same network as my machine at work. I think the proxy arp will
work, but the issue is routing subnet packets all the way past the
SLIP server to my home linux box.
... I've gotten proxy arp working, but routing is still a
problem, hence the need for the IPOPT_RELAY extension... ?
Is there any way to force non-specific IP packets along a certain
route? Or to broadcast routing information telling the routers that
one particular machine on the subnet can be reached through an
entirely different network?
Thanks for all the help. :)
-Sam Lantinga (slouken@cs.ucdavis.edu)
------------------------------
From: elf@netcom.com (Marc Singer)
Subject: Re: 1+ Gig SCSI Drives
Date: Wed, 21 Sep 1994 17:13:15 GMT
Bruce Varney (varneyb@sage.cc.purdue.edu) wrote:
: I thought I saw something about troubles with large drives
: under linux, but when I went back through news today, I couldn't
: find anything. Could someone please tell me what the problem with
: large drives is.
: Bruce
I, too, have been wondering about this. I believe that there are at
least two problems with >1G support. First, the standard IBM
partition scheme limits the number of heads to 255 and the number of
cylinders to 1024. The math comes out such that drives larger than 1G
cannot be supported without hardware/firmware assist, e.g. Adaptec's
255 head mapping trick. Since Linux uses the IBM style partition
tables, there is nothing simple that can be done.
Second, I suspect that there are some other kernel dependencies
relating to >1G drives. Unfortunately, this is merely speculation.
It comes from troubles I have had with 1.2G and 1.7G drives as ext2
devices.
There is hope, however. I have a Micropolis 2217 (1.7G) drive which I
formatted and partitioned with Adaptec software & hardware. I wanted
to use all of the space, so I formatted as a DOS device, split it at
the 1G boundary to avoid 32K allocation clusters, did the 255 head
magic, and mounted the device as a UMSDOS device in Linux. With some
inconveniences with respect to file and directory attributes (via
samba), the device works fine.
I once read a rumor about a new filesystem standard. I believe that
ALL unices are limited to 2G partition sizes due to the 32 bit file
pointer accepted by the standard OS entry points. Perhaps there is a
movement afoot to go to 64 bit pointers as did Microsoft with Windows
NT.
Marc Singer
------------------------------
From: neal@ctd.comsat.com (Neal Becker)
Subject: Re: Shared Libs: working toward a permanent solution?
Date: 21 Sep 1994 17:59:15 GMT
Keeping on with the old non-pic libs is not a sustainable option. The
requirement to globally register every shared lib is becoming an
impediment to wider usage of linux.
I'm sorry for those who don't want to give up a small percent of their
CPU. Piss an moan in Intel's direction. It is because they designed
the 8086 family without consideration of software requirements that
pic performance is not as good as is could be.
I don't care if I have to give up 10%. We have no real choice for a
sustainable linux future.
------------------------------
From: imn@cs.buffalo.edu (Iulius M Nicolescu)
Subject: Re: Any Linux MOTIF packages out there?
Date: Wed, 21 Sep 1994 18:10:58 GMT
Piotr Kapiszewski (kapis-p@cs.Buffalo.EDU) wrote:
: D. Blake Werts (dwerts@hubcap.clemson.edu) wrote:
: : Just wondering if anyone could direct me to any Linux MOTIF packages out
: : there....
: I think there is something out there. Try talking to a friend of mine who
: actually mentioned it to me.
: imn@cs.buffalo.edu or imn@cedar.buffalo.edu
: -Kapi
: --
: Kapi, 542 Baldy Hall, 645-2448
Please STOP sending me e-mail about Motif.
Here is where I BOUGHT Motif from.
^^^^^^
From holly@chickadee.metrolink.com Mon Jul 18 13:38:25 1994
Received: from ns.metrolink.com (NS.METROLINK.COM [192.153.117.163]) by terminus-est (8.6.8/8.6.4) with SMTP id NAA11232 for <imn@acsu.buffalo.edu>; Mon, 18 Jul 1994 13:38:14 -0400
From: holly@chickadee.metrolink.com
Received: from chickadee.metrolink.com by ns.metrolink.com with SMTP id AA15433
(5.67b/IDA-1.5 for <imn@acsu.buffalo.edu>); Mon, 18 Jul 1994 13:38:10 -0400
Received: by chickadee.metrolink.com (/\==/\ Smail3.1.21.1 #21.8)
id <m0qPwdl-0004aKC@chickadee.metrolink.com>; Mon, 18 Jul 94 13:38 EDT
Message-Id: <m0qPwdl-0004aKC@chickadee.metrolink.com>
Date: Mon, 18 Jul 94 13:38 EDT
To: Iulius M Nicolescu <imn@acsu.buffalo.edu>
Subject: Re: MOTIF_FOR_LINUX
Content-Length: 2851
Status: R
Motif for Linux information is attached below. An order form is
also included.
This package requires 8-16MB of memory and uses about
9MB of disk space.
<><><>=<><><>=<><><>=<><><>=<><><>=<><><>=<><><>
Holly Robinson
Metro Link Incorporated
holly@metrolink.com
4711 N. Powerline Rd. <> Fort Lauderdale, FL <> 33309
<> Phone: 305-938-0283 <> Fax: 305-938-1982 <>
<><><>=<><><>=<><><>=<><><>=<><><>=<><><>=<><><>
Motif for Linux
Metro Link, Inc. is pleased to announce the availability of OSF/Motif 1.2.3
for Linux at $199.00 for a complete runtime and development system.
What is included:
Runtime:
1) Motif Window Manager (mwm)
2) Shared motif library (libXm.so.1.2.3)
3) Motif demos both from OSF and from the net
Development:
1) Shared+Static Motif library
2) Static Mrm and Uil libraries
3) UIL compiler
4) Motif header files
5) Manual pages for Motif function calls
6) Imakefile support
7) Source to OSF/Motif demos
And your choice of any one of the X-window books published by O'Reilly &
Associates, Inc.
Distribution Policy: Motif is distributed on three floppy disks with
one manual and release notes. As this involves shipping costs, you
may optionally pre-pay $199 then receive ftp information to access the
files at Metro Link Internet site. You would not receive a book, but also not
pay any shipping. With this option you would only receive a receipt
in the mail.
Update Policy: A new release will occur after OSF releases the next version of Motif. These
will have an update charge of $35 within the first 30 days of Metro Link's product release, or
$69 thereafter. If the new release occurs within 30 days of your purchase, the update is free.
You can order OSF/Motif 1.2.3 for Linux by calling Metro Link, Inc. at
(305) 938-0283 (voice) or (305) 938-1982 (fax) or by mailing us at
holly@metrolink.com.
===========================================================================
====
Metro Link Incorporated
4711 North Powerline Road
Fort Lauderdale, Florida 33309
USA
Voice: (305) 938-0283
Fax: (305) 938-1982
Email: holly@metrolink.com
OSF/Motif 1.2.3 Order Form for Linux
====================================
(fax or email back to holly@metrolink.com)
Name:
Billing Address:
Shipping Address (if different from Billing Address):
Phone Number (required for international orders):
Fax number:
Email:
Shipping Method (Regular, Express, or Internet ftp information):
3.5" diskettes or ftp access:
Programming Manual or User's Guide or O'Reilly Manual #___
(book not available with ftp option):
Payment method:
Check (US only):
Credit Card (Visa, MasterCard, Amex):
Card Number:
Expiration Date:
Name as it appears on the card:
===========================================================================
====
Sorry,
Iulius
------------------------------
From: boliver@kurango.cit.gu.edu.au (Brian Oliver)
Subject: Where are Patches 13,16 and 41?
Date: Tue, 20 Sep 1994 09:33:14 GMT
Does anyone know where patches 13,16 and 41 have gone from sunsite.une.edu?
I really need to patch linux up to v1.1.20 so I can apply a pcmcia patch,
but I can't seem to if 16 and 13 are missing....
or am I just blind/dumb?
Thanks in advance.
-- Brian
PS: Please email me back at , rather than posting back!
boliver@cit.gu.edu.au
------------------------------
From: eric@pandora.Las-Vegas.NV.US (Eric J. Schwertfeger)
Subject: Re: Future of linux -- the sequel
Date: 20 Sep 1994 14:24:56 GMT
: >> I believe that IDE is *1* bit at a time between the controller and
: >> disk.
: Alan> Then why are their D0-D7 on the cable 8)
: Ah. It sounds like a definitive answer here. Not 16 bits, not 1 bit,
: but 8 bits, just like plain-ordinary SCSI (not wide).
No, there's DD0-DD15, 16 bits. The reason IDE is slower is because without
ATA-2 DMA Mode (preferably multiword mode 2), the CPU has to do the
transfers itself, whereas most SCSI disks do busmastering. The ATA-2 draft
spec (which is already being used) has transfer rates up to 13M/Sec, which
puts it between Fast SCSI II and Fast/Wide SCSI II. I'm seriously
considering getting a busmastering IDE controller just to write DMA drivers
for linux on it :-)
------------------------------
Crossposted-To: comp.os.linux.misc,comp.os.linux.admin,alt.uu.comp.os.linux.questions
From: fvm@tasking.nl (Frank van Maarseveen)
Subject: Re: How to use a host as a router - READ THIS
Date: Wed, 21 Sep 1994 12:11:49 GMT
Jay Ashworth (jra@zeus.IntNet.net) wrote:
[text deleted]
> A typical route information output in such a case might look like this...
>
> Destination Gateway Netmask Flags MSS iface
> 127.0.0.1 127.0.0.1 255.0.0.0 UH 1536 lo0
> 199.245.227.0 199.245.227.254 255.255.255.0 U 1436 eth0
> default 198.147.221.1 255.255.255.0 U 1436 ppp0
Why has the loopback route a destination 127.0.0.1 instead of 127.0.0.0
in accordance with the specified netmask?
The loopback interface could have IP address 127.0.0.1 and connect to
a loopback network 127.0.0.0 with netmask 255.0.0.0 (conceptually)
About assigning the same IP address to multiple interfaces on the same
host:
One could argue that this is "correct" as long as the outer world cannot
see any difference. Consider all interfaces as one big specially designed
interface with the software (linux) hiding the differences. So, as long
as the software supports this concept it is correct IMHO.
______________________________________________________________________
Frank van Maarseveen _____ _ _ fvm@tasking.nl
Tasking BV /_ / |_/ /
Plotterweg 31 / \/_/ _/ phone : +31 33 558584
Amersfoort, The Netherlands fax : +31 33 550033
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
When I hear of Schrodingers cat, I reach for my gun --- S. W. Hawking
------------------------------
From: becker@cesdis.gsfc.nasa.gov (Donald Becker)
Subject: Re: Shared Libs: working toward a permanent solution?
Date: 20 Sep 1994 23:37:35 -0400
In article <35ksr8$nbj@news.cais.com>,
Eric Youngdale <ericy@cais2.cais.com> wrote:
>In article <35kk1o$fbm@cesdis1.gsfc.nasa.gov>,
>Donald Becker <becker@cesdis.gsfc.nasa.gov> wrote:
>>Remember that one goal is to maximize the number of pages that may be shared
>>among processes. Resolving the addresses at load time changes bunches of
>>text pages, preventing them from being shared.
>
> No, this is not true. The way that ELF/PIC is designed, there is
>a slot in the GOT for each PLT entry. Thus it is just a pointer in the .got
>section which is modified when the .plt entries are initialized.
The previous poster was talking about doing real relocation/linking/loading,
not PIC ELF libraries. I was pointing out that that kind of linking would
involve modifying just about every page, rendering it unsharable. Although
it would ultimately generate fast code, sharing pages is far more important
for overall system performance.
--
Donald Becker becker@cesdis.gsfc.nasa.gov
USRA-CESDIS, Center of Excellence in Space Data and Information Sciences.
Code 930.5, Goddard Space Flight Center, Greenbelt, MD. 20771
301-286-0882 http://cesdis.gsfc.nasa.gov/pub/people/becker/whoiam.html
------------------------------
From: russ@wpc18.ed.ray.com (Russell E. Dube)
Subject: Re: XFig Eats All My Memory.
Date: 21 Sep 1994 12:13:09 GMT
Lucas James Sheneman (sheneman@cs.uidaho.edu) wrote:
: I am running Linux 1.1.49, XF86_VGA (from XFree86-2.1.1). When I run
: xfig, it rapidly eats up all of my available memory and never actually
: pops up. I have 20MB RAM and 25MB swap.
FWIW, I have the same problem with 1.0.9 and Slackware 2.0. Eats my
available memory (16M machine), never pops up. Please post since
undoubtedly others have a need to know. Thanks in advance.
--
Russ Dube -- My opinions, not those of my employer ...
------------------------------
From: becker@cesdis.gsfc.nasa.gov (Donald Becker)
Subject: Re: Shared Libs: working toward a permanent solution?
Date: 19 Sep 1994 14:07:52 -0400
In article <S.Herbert.7.2E7D56E2@shef.ac.uk>, <S.Herbert@shef.ac.uk> wrote:
>In article <Cw8B6I.I62@info.swan.ac.uk> iialan@iifeak.swan.ac.uk (Alan Cox) writes:
>
>>0-30% slower depending on the exact code involved. Certainly not acceptable.
>
>If the loss is < 5%, then the benefits to developers makes it acceptable.
>The current way of building, and *maintaining* shared libraries is a
>nightmare - it's impossible to predict how much the structures of a new
>library will change as the library grows, making static libraries look more
>attractive from the large savings in the time of the developer.
For common libraries used by lots of people and program, such as libc,
it's well worth the extra development time. (IMHO, since I'm not the one
doing the work.)
>As for speed concerns, is it not possible for the library loader to resolve
>addresses in the library at load time, leaving applications to link via a
>hash table? That way, extra overhead during actual run time is avoided -
>you have a one-off penalty when you load the library.
Remember that one goal is to maximize the number of pages that may be shared
among processes. Resolving the addresses at load time changes bunches of
text pages, preventing them from being shared.
Here are a few points to consider and debate:
o PIC (position independent code) on x86 processors is
significantly slower than normal code.
o Small processes, like those typically run by shell scripts, spend
most of their time in startup and library calls.
o I/O-oriented program often use printf(), which is library code.
vs.
o Library calls like read() are just a few instructions to set up
the syscall. The slowdown of PIC is unnoticable.
o CPU intensive applications spend little time in library code.
and tend to use calls like read() when they do.
--
Donald Becker becker@cesdis.gsfc.nasa.gov
USRA-CESDIS, Center of Excellence in Space Data and Information Sciences.
Code 930.5, Goddard Space Flight Center, Greenbelt, MD. 20771
301-286-0882 http://cesdis.gsfc.nasa.gov/pub/people/becker/whoiam.html
------------------------------
From: behnke@tu-harburg.d400.de (Lutz Behnke HiWi)
Subject: Re: Don't use Linux?!
Date: 19 Sep 1994 14:58:55 GMT
In article <JKVG.94Sep9231859@kamet.ccs.neu.edu>, jkvg@kamet.ccs.neu.edu (Jagadeesh Krishnamurthy Venugopal) writes:
|>
|> 1) Linux is of the hackers, by the hackers, for the hackers. There is
|> absolutely no need to make it commercially successful. In the search for
|> commercial success I am afraid we might lose track of the original ideal-- a
|> free os which is the hacker's dream. Well if one wants commercially available
|> releases of UNIX there are plenty-- UnixWare, Solaris, etc. Let us have Linux
|> just for us. Once the commercial market is aimed at, technical advances get
|> sacrificed at the altar of compatability (Consider that the largest selling OS
|> today is the most wretched). It would be a sad day for Linux if it happened.
BS!
Just because the the OS is free doesn't meen the software has to be.
Linux (if you don't go for the newest stuff all the time)
is stable enough for the comercial sector to go for it. The thing about the
libc is true. But the argument 'Linux for the hackers' is outdated for long.
The compatibility is nothing to brake progress. Just go for the max_version-1 Kernal
an UR okay.
Nothing lost to comercial applications of the worlds greates OS.
my 2 pfennig
mfg Lutz
------------------------------
From: behnke@tu-harburg.d400.de (Lutz Behnke HiWi)
Subject: Re: Don't use Linux?!
Date: 19 Sep 1994 15:13:00 GMT
In article <Cvxs44.55u@rahul.net>, Kevin Martinez <lps@rahul.net> writes:
|> hightec@sbusol.rz.uni-sb.de (Michael Schumacher) writes:
|>
|>
|> > 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.
|>
|> This is a joke? Have you ever tried to get hotline support for Netware?
|> $200 up front, then wait for them to call back! (2 days in the case I am
|> familiar with). Try getting hotline support from Macrosoft: tell them
|> that "xcopy" trashes your command history. Will they lift a finger?
|> h0h0h0h0 (and you are on a 900 number paying by the minute!)
|>
[some other burdened comments about software-support deleted]
I see a rather simple answer to this problem of not enough support.
(If there is one)
Wy not have a list of people willing to give tech-support for each country
included in every CD-ROM arcive.
This is the way Richard Stallman envisioned the support for GPL products.
I am a student in one of the two major comercial centers of Germany, so
would be glad to help any company or private person in need of some
help installing the Linux package on there PCs. I am pretty shure
I cost less than the support some companies, I are availiable almost
around the clock, and there is no need for the client to understand the
lingo. That beats MS anytime.
(My ear is still clogged from the sermon the +@#$#$@@$$% guys from Micro-
Soft, Lexmark and IBM have giving over that #$%$%#$^@#$@ printer connection
from MSWinWord6.0)
But for the flexible, availiable help from privete persons to privste persons
there is no way to advertise. I. by myself can't afford large ads in
selectone(["CHIP", "ComputerWeekly", "c't", etc....])
How about a list of people with real world telefone numbers on each CD.
maintained by someone, and put in sunsite. (therefore mirrored on each
"DREAM-CD").
mfg Lutz (my 2 pfennig)
| Lutz Behnke | behnke@tu-harburg.d400.de |(Germany) +40 / 630 39 38 |
| TU Hamburg Harburg, Hamburg, Germany, Europe, Earth, Sol-System |
|----------When the Evil Spirit armed the Tiger with claws,----------|
|----------------Brahma gave wings to the Dove-----------------------|
------------------------------
From: dwm@shell.portal.com (David - Morris)
Crossposted-To: comp.os.linux.misc,comp.os.linux.admin,alt.uu.comp.os.linux.questions
Subject: Re: How to use a host as a router - READ THIS
Date: 21 Sep 1994 19:13:07 GMT
Re. why not 127.0.0.0 instead of 127.0.0.1 -- the destination address must
be a 'host' address and the host address can't be zero (0).
------------------------------
From: behnke@tu-harburg.d400.de (Lutz Behnke HiWi)
Subject: Re: Don't use Linux?!
Date: 19 Sep 1994 15:23:43 GMT
In article <34uu83$o4@lucie.wupper.de>, kay@lucie.wupper.de (Kay Hamacher) writes:
|> In article <34pq45INNojt@sbusol.rz.uni-sb.de>, hightec@sbusol.rz.uni-sb.de (Michael Schumacher) writes:
|> The real problem is in the mind of most people : Everything which is good must
|> be very expensive! So Solaris with high prices must be better than UnixWare
|> which seems to be better than Linux, as Solaris is more expensive and more
|> packed as UnixWare which is indeed more expensive than Linux. Or see Windows :
|> It is not good, it is bad. But : 1) the companies see only the stupid customer
|> buying his first computer. So they say : "hey, let's get Windows. You can simply
|> move the mouse-pointer round and do everything" And why ? Because they want
|> to get their percentage of the Windows-price. No computer-dealer would say to
|> the normal customer : "Use Linux, it is better", as he earns no extra money with
|> this kind of software. and 2) that is the point of view : everything must be
|> orientated on the big sell-numbers and not on the performance.
|> It is similar to cars : profis know where to buy good stuff to tune their
|> car or to cause the car not to need such high volumes of gasoline. The
|> average user of the car does not exactly know what does this machine (as I do not
|> know this), but it works : Fine ! I have not the time to intensive my knowlege
|> about cars and it makes me no fun. But there are people having fun on doing
|> this. So why should there be no people having fun in compiling every week
|> a new kernel for the system they use ? It is not car-tuning it is simply
|> computer-tuning. What is wrong on this ?
I like your analogy:
Its extremly good for the enviroment to have people driving around in old
clunkers that burn a gallon a mile. (no smile, face in rage!)
So just you and I know that Linux is better than Windows means that we are the
profesionals and may be proud of that knowledge.
Now look at it from the other angle: I am trying to make a living of
computers. I have to listen to the whailing of the users all
the time: Why is such and such not working (in a Windows envirment (see former post))
I would like to bring the good news to John Q Public too. But as long as there is
no WYSIWYG text-system for them they will not by it. And as long MS is the
only one able and willing to have an aggressive Ad-Campaign running,
all those lemmings will have to follown the light.
And lets face it: This textsystem may only come from a comercial company.
(pleeezzee surprise me on this one)
|>
|>
|> Kay
lutz
| Lutz Behnke | behnke@tu-harburg.d400.de |(Germany) +40 / 630 39 38 |
| TU Hamburg Harburg, Hamburg, Germany, Europe, Earth, Sol-System |
|----------When the Evil Spirit armed the Tiger with claws,----------|
|----------------Brahma gave wings to the Dove-----------------------|
------------------------------
** 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
******************************