See also: last week's Back page page.

Linux Links of the Week

A related site is EFF's Campaign for Audiovisual Free Expression. Check it out for news from that front of the fight for freedom.

Section Editor: Jon Corbet

This week in history

However, writing new drivers for the thousands of peripherals on the market is a daunting task. Hence, Project UDI is hoping the Linux community will help... A reference platform will be distibributed as freeware for Linux, and the Project UDI members will be counting on the Linux community to work on device drivers...
-- ZDNet.

The Linux community showed little enthusiasm for the idea of providing device drivers for the convenience of proprietary Unix vendors, and UDI faded away.

The development kernel release remained at 2.1.122. Linus called for a change in how the network drivers worked, because it was all wrong at the time. The changes called for happened, but not until 2.3.43.

IBM finally got around to announcing that its DB2 database would be made available for Linux. Sybase, too, got in on the act with its release of "Adaptive Server Enterprise."

One year ago (September 23, 1999 LWN): A previously obscure company called LinuxOne released a new distribution (called by some "Red Hat with the serial numbers filed off") then promptly filed for an IPO. Needless to say, this move was not well received. One year later, now, the IPO has not happened.

Corel Linux went into beta test. The event was overshadowed, however, by a rather severe nondisclosure agreement that beta testers were expected to sign. Linux-Mandrake 6.1 was made available for download.

LinuxOne was not the only IPO filing that week; Andover.Net also put in for an offering. They had rather more success at it.

Letters to the editor

Date: Thu, 14 Sep 2000 14:18:20 -0600
From: yodaiken@fsmlabs.com
To: lwn@lwn.net
Subject: RTLinux story

Your story was very nice, but, of course, I have to complain about
something.  And the something is

	RTLinux makes many changes to the standard kernel source, while
	RTAI takes a minimalist approach to kernel changes.

RTLinux has always done only minimal changes to Linux: working only at the
lowest level of the architecture dependent interrupt handling. RTLinux on
PowerPC requires no changes at all to Linux kernel and on x86 and Alpha the
changes are all localized.  In fact, from the beginning of this project,
making it easy to track the kernel and staying out of the core operating
system have been priorities for us.

-- 
---------------------------------------------------------
Victor Yodaiken 
Finite State Machine Labs: The RTLinux Company.
 www.fsmlabs.com  www.rtlinux.com

From: Massimo Dal Zotto <dz@cs.unitn.it>
Subject: Re: Cisco patents NAT
To: lwn@lwn.net
Date: Wed, 20 Sep 2000 22:06:39 +0200 (MEST)

Hi,

the Cisco patent on NAT was filed on 3 Nov 1995 and my old copy of the
linux-1.3.20 kernel, dated approximately August 1995, already contains many
references to the constant CONFIG_IP_MASQUERADE, so we have a clear case of
prior art in the linux kernel itself.

Should this be enough to invalidate the Cisco patent or do we need some
legal battle to defend our version of the same idea?

-- 
Massimo Dal Zotto

+----------------------------------------------------------------------+
|  Massimo Dal Zotto               email: dz@cs.unitn.it               |
|  Via Marconi, 141                phone: ++39-0461534251              |
|  38057 Pergine Valsugana (TN)      www: http://www.cs.unitn.it/~dz/  |
|  Italy                             pgp: see my www home page         |
+----------------------------------------------------------------------+

Date: Wed, 20 Sep 2000 15:03:18 -0500
From: Dave Finton <surazal@chef.nerp.net>
To: letters@lwn.net
Subject: Am I the only one?


Lately I've noticed a big trend with Linux news sites.  Due to it's
rather populist roots, Linux advocates have always been extremely
passionate about their beliefs (I know this because I'm from the same
camp myself).  No more is this true than with those who run popular
Linux-oriented web sites, such as Slashdot and LinuxToday.  This, too, 
is the result of a long history, where these types of web sites were
the voices of Linux and Open Source in the earlier days.

Nowadays, Linux is jockeying for position in multiple markets
(successfully in many cases) and the concept of "Open Source" or "Free 
Software" (or whatever your preferred buzzword is) has taken root even
in business executives' minds.  Linux has hundreds of millions (if not 
billions) of dollars surrounding it when counting public and private
companies together.  Times have changed, in other words.

So why haven't our advocacy methods changed at all?

Take the recent ruckus with online polling between LinuxToday and
MSNBC.  MSNBC has so far refused to comment in great deal concerning
the mess.  They have so far refrained from trying to sully
LinuxToday's reputation.  This is simply because LinuxToday is doing a 
fine job of shooting itself in the foot on its own.

What the Paul Ferris's and the Rob Malda's of the world have to
realize is that it ain't the best strategy to provoke an opponent to
react to outrageous actions they've committed and then whine when the
opponents do react in some way.  I can't fault MSNBC for rigging its
own polls, because LinuxToday forced them to.  Yes, we all know online 
polls are a joke, but I disagree with the methods being used to combat 
their use in public discourse.

What is LinuxToday trying to accomplish with this?  Are they pointing
out that MSNBC is owned by a corporation with a conflict of interests
a mile long when it comes to technical journalism?  Well no duh.  Are
they trying to say that MSNBC was at fault for reacting to a
(what could be interpretted as) legitimate attack on their servers?
Maybe, but personally I am beyond caring at this point.  Are they
trying to improve their corporate image?  If so, man do they ever need 
help in that department.  Personally I find the behavior of the Linux
advocates highly questionable, simply because it just looks like (to
me) a provocation done for the sake of provocation.

Even though I single out LinuxToday in this rant, a lot of news
publications devoted to Linux are guilty of the same misdeeds.  And
when someone like me (a Linux advocate who's more than likely ventured 
out into the land of zealotry on more than one occasion in the past)
finds that behavior extreme, what does it say to "the unwashed
masses"?  Do they think that this is supposed to make themselves look
good to the public?  If so, how?

My criticism is harsh, but I think the point needs to come across some
of the "great leaders" of this big movement that we need to move on.
When trying to paint a friendly face on Linux and Open Source, it's
best to put your best face forward, not scream and hollar like
children whenever things don't go the way we want them to.  We must
all learn to adapt.

If we don't, soon we'll find ourselves starting at square one all over 
again.

-- 

                          - Dave Finton

---------------------------------------------------------
| If an infinite number of monkeys typed randomly at    |
|   an infinite number of typewriters for an infinite   |
|   amount of time, they would eventually type out      |
|   this sentencdfjg sd84wUUlksaWQE~kd ::.              |
| ----------------------------------------------------- |
|      Name:      Dave Finton                           |
|      E-mail:    surazal@nerp.net                      |
|      Web Page:  http://surazal.nerp.net/              |
---------------------------------------------------------

Date: Thu, 14 Sep 2000 11:44:25 -0600
From: Jim Easter <jre@sni.net>
To: letters@lwn.net
Subject: A hit with a bullet?

Regarding "MP3.com yanks song with illegal DVD-hacking code (News.com)":

Years ago, my family lived next door to a man named Dolf, who had grown
up in Poland during the 1940s and 50s.
We were talking about popular music one day, and Dolf told us about how
new songs had gotten exposure in the Poland of his youth.  He said that
the coffeeshops and bistros of Polish cities had a kind of permanent
open-mike policy, and that aspiring musicians would hang out there in
search of fame.  Some got well-known over the course of time, but the
best shortcut to renown was the government censor.  Each coffeehouse had
a government representative sitting at the side of the stage, whose job
it was to stop performers from singing political satire which strayed
into forbidden territory.  When that happened, the censor would step
forward and firmly state that the song in question could not be sung.
According to Dolf, the coffeeshop would then empty out as people ran to
their friends' houses with the new song: "Didja hear the song the censor
just shut down?  It goes like this ..."

I can't speak for anyone else, but this story was all the encouragement
I needed to download Joe Wecker's hilarious song from the (perfectly
legal) Gallery of DeCSS Descramblers [1] maintained by Dr. David
Touretzky.  Give it a listen.

It should be noted that MP3.com is a reluctant censor, acting out of
fear rather than malice, but the effect is the same as if the DVD CCA
had been sitting at the side of the stage.  Fortunately, the end result
is that these threats tend to backfire.  You can't buy that kind of
publicity!

[1]  Touretzky, D. S. (2000) Gallery of CSS Descramblers. Available:
http://www.cs.cmu.edu/~dst/DeCSS/Gallery, (14 Sept 2000)

Date: Fri, 15 Sep 2000 04:56:52 -0500 (CDT)
From: <shane@time-travellers.org>
To: Jon Corbet <lwn@lwn.net>
Subject: Teensy ELF Executables

[Regarding last week's link of the week on teensy ELF executables: 
    http://www.muppetlabs.com/~breadbox/software/tiny/teensy.html
]

Given that this isn't really an ELF file that you end up with, why not
just use the following:

  $ echo -n 'exit 42' > a.out
  $ chmod +x a.out
  $ ./a.out ; echo $?
  42
  $ wc -c a.out
        7 a.out

Frankly, I'm tired of those bloated 45 byte executables.  BTW, this works
on all Unix variants.

Shane

p.s. Yes, this requires a interpreter, but given that /bin/ash is about 60
Kbyte and /boot/vmlinuz is about 600 Kbyte, I'd say that's fair.  :)

Date: Wed, 20 Sep 2000 12:34:22 +0100
From: Richard Kay <rich@driveout.demon.co.uk>
To: jim@mischel.com, letters@lwn.net
Subject: Failure of the free software movement ?

Your article in Dr Dobbs Journal might describe the as yet incomplete
success of what to some seems an extreme ideology perhaps. The many
examples of adoption of open-source development methodologies which you
record for perfectly good commercial reasons also suggests very
considerable pragmatic success of this movement.

I also don't see the core free software ideology as anti-commercial, though
a few who claim to support this may seem to do so for such motives. Another
factor you miss is the noticeable drift of important semi-free software
such as Netscape/Mozilla, Star-Office and KDE towards becoming completely
free software, also for very pragmatic commercially-driven reasons.

Your anti-GUI ease of use argument is way out of date.  I might accept your
critique that Linux is less easy for newbies to use in respect of my more
than 2 year old Slakware 3.4 FVWM95 system. Having very recently upgraded
this using the new Mandrake 7.1 basedKDE desktop this gives such a
significant improvement on the GUI ease of use front that this is an
improvement over Windows 98 and NT4 in this area, let alone Windows 3.1.

There exists a core issue to do with the principles behind free software
which your article fails to address. The progress of the free software
movement described and the general advantages to inventors, authors,
programmers and artists of having access to free distribution not mediated
by powerful corporations is leading many of us to question in whose
interest the copyright and patenting system works. This leads to a definite
political question: should the state be involved in the protection of the
private vested interests which patent and copyright laws involve ? To what
extent do these laws protect the interests of inventors and authors as
opposed to those of corporations and publishers ?

There was clearly a case for such state protection 200 years ago when a
well capitalised printer could readily cream the commercial value of an
impoverished author's work, but copyrights and patent laws are no longer
seen to protect those in whose interests they were originally passed. It is
now up to those who would seek to maintain these protections to justify why
the general public should be deprived, by the force of law, of their
pre-existing natural right to make copies of works considered worth
copying.

Richard Kay