LWN: Comments on "Heise reports from SCO Forum"

Heise reports from SCO Forum

bajw — Thu, 21 Aug 2003 07:19:32 +0000

I did a Google on the phrase, and the results showed that the phrase was
"Hell, everything's legal in Mexico. It's the American Way." Google also
indicated that it was said by Uncle Jimbo of Southpark.

Heise reports from SCO Forum

Ross — Wed, 20 Aug 2003 17:18:34 +0000

You made me think of something... this code is only used on IA64!

How many people are actually using that code? Not many I would imagine.
SCO's threatened sue-the-users campaign would probably have trouble
locating any victims.

Heise reports from SCO Forum

keithl — Wed, 20 Aug 2003 08:52:35 +0000

Non programmer here. I looked at ftp.kernel.org for the ate_utils.c
file. I note that ate_utils.o appears in .../arch/ia64/sn/io/Makefile
between 2.4.3 (March 30, 2001) and 2.4.4 (April 28, 2001). The
ate_utils.c file itself shows up in that directory between 2.4.18
(Feb 25, 2002) and 2.4.19 (Aug 2, 2002), complete. The routine
shown in the article, atealloc(), shows up in 2.4.4 as a single
define in .../arch/ia64/sn/io/pcibr.c:

#ifdef __ia64
#define rmallocmap atemapalloc
#define rmfreemap atemapfree
#define rmfree atefree
#define rmalloc atealloc
#endif

While in 2.4.19 it is in a bunch of places, still associated with
variations of pcibr.c. Perhaps someone with far more code
understanding than I can explain what all this code is for, and
how the code got in there, who wrote it, and why the .o is
mentioned in the 2.4.4 makefile over a year before the actual
code arrives, and maybe even how the rmalloc [that is, atealloc]
works when atealloc isn't there yet?

No agenda here, I'm just an egg wondering how all this works, so I
can pontificate at length to the two or three folks out there who
know even less than I do. Perhaps some kernel savvy individual
might inform us what this all means. The cool thing about Linux
is that almost all of this is documented in public somewhere, if
you know where to look, and some individual is always ready with
a clue-by-four to enlighten the rest of us.

Keith

Heise reports from SCO Forum

stuart — Wed, 20 Aug 2003 07:59:59 +0000

Precisely because they aren't in those other areas of the industry :-)

My write-up of your research

ekj — Wed, 20 Aug 2003 07:35:40 +0000

This also happens to demonstrate very nicely why SCO has been so reluctant to identify which wrongdoing, exactly, they are accusing us of. Other than that it involves "IP" and "Contract law".

They know fully that they have no evidence whatsoever that will stand up to even the briefest scrutiny.

This incident demonstrates that. As soon as they release the smallest possible hint of evidence, two slides totaling maybe 40 lines of codes, thousands of nerds descend upon it, dive into Google, and appear minutes later waving embarassing references to fucking K&R (If there is ONE place for code to appear to rule out that said code can be called a "Trade Secret" then it is K&R), aswell as the equally unwelcome fact that SCO themselves BSD-licensed the code in question.

I predict that the same, or something very similar will happen to all other "evidence" that SCO publish. Furthermore I am fairly sure that SCO is aware of this.

There's no other plausible explanation for their behaviour.

Heise reports from SCO Forum

avanaardt — Wed, 20 Aug 2003 01:18:53 +0000

As a newbie, just a word of thanks to all you Linux guys - amazing work of the highest quality. Thanks guys!

(Is it just me, but why do Linux people seem to have a sense of humor sadly lacking in other areas of the industry?)

Heise reports from SCO Forum

Jhebbard — Wed, 20 Aug 2003 00:39:03 +0000

unless the NDA is just a subterfuge to give them time to work an extortion racket - and if it never goes to court they will never be proven guilty - the final issue depends on the judge and his/her attitude towards justice. . As an Engineer (me) sees it, the main purpose of the civil tort system is supposed to be justice, not punishment. Therefore any court interested in justice will throw SCO's case out because the NDA is obviously designed to prevent the defendants from remedying the tort. For example, thousands of lines of offending code could be removed from the Linux kernel and replaced with more ellegant solutions if SCO would only file a proper complaint.

Heise reports from SCO Forum

Jocko — Tue, 19 Aug 2003 23:30:35 +0000

K&R First Edition (1978) has this code on page 175. Sounds like it's the same as the
described Second Edition code.

Heise reports from SCO Forum

tjc — Tue, 19 Aug 2003 21:42:27 +0000

Klingon isn't really all that hard...

The Tengwar appears to be a bit more complex, although very interesting.

It would be nice if there were an xfonts-tengwar package in Debian. :-)

Heise reports from SCO Forum

Soruk — Tue, 19 Aug 2003 21:13:35 +0000

Klingon isn't really all that hard...

My write-up of your research

mmarq — Tue, 19 Aug 2003 20:16:16 +0000

Pretty damn good job of gathering and synthesis for 1 day, i must say... and all the Linux/OSS community should be thankfull to...

I belive, SCO must be pretty amazed to, specially thinking as i do, that when they mentioned thounsands, millions of lines of code, they could be preparing to feed more than one of this "DIVERSIONS" a week!

Makes me wonder...

mmarq — Tue, 19 Aug 2003 19:51:02 +0000

But their business is really pump&dump, and protect the master as in the last M$ vs DOJ settlement, of wich Boies was the specialist...
As hard as it must sound to many, law in the hands of the powerfull is wholly a tool that can be a weapon, and not the letter of justice...

When all the dust settles, all the harm and FUD that there is to be done, would already be done, all the dumps that there is to be done, would already be done..., and with a major DRM/paladium lock up around the corner, guess who gets to win ?

SCO is right now just the prospect of a ghost,... they give every indications of knowing it, and are just playing a show as a matter of diversion for clearing the bosses... and they are PRETTY DAMN GOOD AT IT.

Heise reports from SCO Forum

tjc — Tue, 19 Aug 2003 19:24:37 +0000

I guess I've got an illegal decryption device embedded in my head then...

I've got a parser in my head, and I'm getting a syntax error on line 2 of the second slide.

Heise reports from SCO Forum

dbhost — Tue, 19 Aug 2003 18:44:59 +0000

to repeat a quote I have heard attributed to Krusty The Clown from The Simpsons while in Mexico.

"Everything's legal in Mexico. It's the American way!"

If I have misquoted, or given credit to the wrong mouth that uttered that wisdom please correct me...

Heise reports from SCO Forum

dsmouse — Tue, 19 Aug 2003 18:31:54 +0000

I hope that they DO make a fortune dumpint their stocks. I also hope they get charged with illegal stock market manipulation

Heise reports from SCO Forum

mmarq — Tue, 19 Aug 2003 18:14:37 +0000

So besides a Diversion, all this show could be a trap to !!!...

With the "criminal" minds that are feading SCO more than clue cards, its no wonder!...

What is not good, is that they can predict Linux/OSS movements...

I belive a "REALLY BIG PUSH" to LSB to completly cover the Desktop, would suprise and scare them to dead!

If all this sound to you as nosense, than you better think again, because you are confused about how "fox smart" business minds work.

Heise reports from SCO Forum

rfunk — Tue, 19 Aug 2003 18:03:28 +0000

Oops, I missed a closing brace in this line:

 
                       while((bp-1)->m_size = bp->m_size);

Should be:

 
                       } while((bp-1)->m_size = bp->m_size);

Heise reports from SCO Forum

mmarq — Tue, 19 Aug 2003 17:36:35 +0000

"Identical typing errors in the comments as well as unusual ways of writing would have left traitorous traces, to stated Sontag"

" team for pattern recognition had angeheuert, around ten thousands from program lines to through forests. The few code sequences shown apart from the comments were made to a large extent illegible, alleged, in order to protect SCOs author-genuine"

AFAI can read, "they" are making pretty vague, and in a rogue manner, allegations about the copying "mantra"... there is no conclusive statement that garanties that they are talking about this particular algorithm, and not just showing "pattern" like "Identical typing errors" to prove that Linux/OSS are thieves, even without having stealed anything!...

The clear intention, is to confuse the Linux/OSS community, because it is all ears now, as this almost blowing up of the LWN server is prove, and maintain them as far as possible from defeating is master M$, aligned behind "corporate" interests, and at the same time collect a "FEW" dollars from stock speculation and a few really scared users...

In few words, Plain "Criminal" Diversion...

out there

rfunk — Tue, 19 Aug 2003 17:33:09 +0000

That doesn't work: copyright isn't the same as trademark.

Good point, though actually I was thinking trade secret, not trademark. And SCO has been including trade secret misappropriation as part of their accusations, though it's not clear to me which one they're thinking of with this code.

Rather, we (IBM, really) will need to rely on the Regents of U.C vs. USL agreement granting U.C. full rights to that code as it appears in 2.11 BSD and up.

Except as Bruce has pointed out, we don't know the full terms of the agreement. Are you saying that UC won full rights to everything in 2.11BSD? That's the first I've heard of that.

UCB dropped the advertising clause requirement sometime after that. SCO has no proprietary rights to anything that appears in the last BSD release.

Yes, but this code doesn't appear in the last BSD release. (2.11 is quite a bit older than 4.4.)

Didn't Caldera/SCOX open source SYS V v7

rfunk — Tue, 19 Aug 2003 17:20:48 +0000

Actually I shouldn't have emphasized the GPL incompatibility so much as the basic noncompliance with the Caldera license. See: http://marc.theaimsgroup.com/?l=linux-kernel&m=105607172529951

On the other hand... http://marc.theaimsgroup.com/?l=linux-kernel&m=105611216326103

out there

ncm — Tue, 19 Aug 2003 16:47:49 +0000

rfunk said, "we have to go back to a defense of 'that code has been out there for years anyway'".

That doesn't work: copyright isn't the same as trademark. Rather, we (IBM, really) will need to rely on the Regents of U.C vs. USL agreement granting U.C. full rights to that code as it appears in 2.11 BSD and up. UCB dropped the advertising clause requirement sometime after that. SCO has no proprietary rights to anything that appears in the last BSD release. This is not (as is often reported) because they didn't write much of it, but rather because they were in such hot water over having stolen huge amounts of BSD code, they had to give up rights to the code they had written, or lose the right to sell UNIX at all.

My write-up of your research

BrucePerens — Tue, 19 Aug 2003 16:41:47 +0000

You folks have done a great job today. I've written up your research and am sending it to my press list right now. It's at this link .

Thanks

Bruce

Didn't Caldera/SCOX open source SYS V v7

BrucePerens — Tue, 19 Aug 2003 16:38:20 +0000

Don't throw a fit over the advertising clause. I doubt the GPL copyright holders in the kernel would sue about it in this case. And I think they would be happy to give you a written exception to the GPL for its use. If it mattered.

Bruce

Where is the code ?

mmarq — Tue, 19 Aug 2003 16:37:44 +0000

Actually they dont show any "SIDE BY SIDE", at least in this article, code from linux and from Sytem V...

What they show is a similar "header comment" and everybody is assuming is a piece of code that appeared in a book.

Far as is showed, THERE IS NO EQUAL CODE !

Heise reports from SCO Forum

rknop — Tue, 19 Aug 2003 16:37:07 +0000

I'd love to see those.

The *abuse* that the rodent would take.... :)

-Rob

Heise reports from SCO Forum

rknop — Tue, 19 Aug 2003 16:35:52 +0000

By trafficing in your head, you are a federal criminal.

The only way to be safe is to cut it off and destroy it before the feds look at you.

(And, yes, the DMCA really is that silly.)

-Rob

Heise reports from SCO Forum

daw — Tue, 19 Aug 2003 16:28:04 +0000

A very similar malloc implementation also appears in Kernighan & Ritchie's "The C Programming Language," arguably the most common computer programming textbook ever. I have the second edition (1988), and the code is on p. 187, but I imagine it's in the first, 1978, edition as well.

K&R's version uses a linked list of structs to store the free pointers, while the ancient unix version uses an array of pointers, but the structure of the code snippet is otherwise the same, and it seems likely to me that it had a common ancestor.

FWIW, "The Design and Implementation of the 4.4BSD Operating System" (McKusick et al., 1996) also has a short discussion of the kernel malloc implmentation (no code), which mentions the "first-fit" algorithm.

Heise reports from SCO Forum

gups — Tue, 19 Aug 2003 16:20:05 +0000

Regardless of how stupid a show they're putting on, there always are people stupid enough to buy into their crap.

That's why McBride and gang will be making a fortune dumping their previously worthless stocks, and that's probably all they've ever wanted to do.

It's sad.

Heise reports from SCO Forum

BrucePerens — Tue, 19 Aug 2003 16:19:35 +0000

The same code is available under the BSD license without the advertising clause in code copyrighted by the University. I assume this code is subject of the USL vs. BSDI lawsuit and its following settlement. I think it's time for the University to say something about the settlement terms, which aren't entirely public although the results are well-known. We can subpoena them in one of the lawsuits, if necessary. If this came down to attribution and the advertising clause in the Caldera license, they'd have no case anyway.

Bruce

Heise reports from SCO Forum

lorenb — Tue, 19 Aug 2003 16:17:36 +0000

I have this book and in Chapter 5. The book doesn't have pages per say. This is what's the index:

malloc:

line page
------------
2528 5-2
2534 5-2
2535 5-2
2536 5-2
2537 5-2
2538 5-2
2539 5-2
2542 5-2
2543 5-2

Heise reports from SCO Forum

shadowman99 — Tue, 19 Aug 2003 16:16:16 +0000

Actually I would think SCO would be in trouble with the DMCA before anyone for violating copywrite law by changing the text as originally written. They have moddified the character set without permission of the author.

By changing code in an effort to obfuscate they have done the very thing they accuse our community of.

Bonk Bonk on the head!

Heise reports from SCO Forum

gups — Tue, 19 Aug 2003 16:08:47 +0000

Hahahaha...

That's funny.

Heise reports from SCO Forum

rickfdd — Tue, 19 Aug 2003 16:07:56 +0000

So this evidence is extremely damaging to SCO's case. Why is the stock price not yet tanking?

If only Disney would BSD license one image of Mickey Mouse, we could all make our own "derivative" cartoons staring the damn rodent.

Heise reports from SCO Forum

frazier — Tue, 19 Aug 2003 16:02:03 +0000

Bruce,

That license doesn't look GPL compatible to me due to the following:

The following copyright notice applies to the source code files for which this license is granted.
Copyright(C) Caldera International Inc. 2001-2002. All rights reserved.
Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:
Redistributions of source code and documentation must retain the above copyright notice, this list of conditions and the following disclaimer. Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution.
All advertising materials mentioning features or use of this software must display the following acknowledgement:
This product includes software developed or owned by Caldera International, Inc.

Provided this is the case (and I'm following licenses correctly), inclusion in a GPL'd system would be in violation of Caldera's license. That noted, once again Caldera/SCO has likely redistributed this code (uh, set of comments) under the GPL for their own Linux distribution.

Heise reports from SCO Forum

allesfresser — Tue, 19 Aug 2003 15:59:58 +0000

I guess I've got an illegal decryption device embedded in my head then... I just read the text as it was. (Thanks to whoever made the Symbol font for making the transliteration pretty easy!) How am I supposed to keep from violating the DMCA with such an embedded technology...? No, don't answer that. :-)

Didn't Caldera/SCOX open source SYS V v7

rfunk — Tue, 19 Aug 2003 15:48:35 +0000

One problem: When Caldera opened the old Unix code, it was under a BSD-style copyright *with advertising clause*. This makes it incompatible with the GPL which that file in the Linux code claims to be under.

For those not familiar with this incompatibility: http://www.gnu.org/philosophy/bsd.html

So rather than a defense of "Caldera already opened the code", we have to go back to a defense of "that code has been out there for years anyway", which is the defense ESR has apparently been working to document for IBM. http://newsforge.com/article.pl?sid=03/05/25/1240238

Heise reports from SCO Forum

PaulShirley — Tue, 19 Aug 2003 15:47:36 +0000

That document specifically excludes Sys V derivatives, which seems to be what they're currently attacking.

Heise reports from SCO Forum

allesfresser — Tue, 19 Aug 2003 15:44:20 +0000

Well, that's what their whole farce is about, isn't it? "All your base are belong to us", so no matter where it's copied from, they own it. And so there is no such thing as legitimate copying, no matter what all those nasty evil un-American commie pinko licenses say. What a nice, tidy, convenient hypothesis.

Not.

Didn't Caldera/SCOX open source SYS V v7

allesfresser — Tue, 19 Aug 2003 15:40:53 +0000

The title of the post is tremendously poignant considering the present crusade against the GPL...

Heise reports from SCO Forum

rfunk — Tue, 19 Aug 2003 15:32:54 +0000

Oops, it was the Santa Cruz Operation that blessed it, not Caldera.
Caldera didn't get the rights until later, then they opened it all under
the BSD-style license.