decoration decoration
Stories

GROKLAW
When you want to know more...
decoration
For layout only
Home
Archives
Site Map
Search
About Groklaw
Awards
Legal Research
Timelines
ApplevSamsung
ApplevSamsung p.2
ArchiveExplorer
Autozone
Bilski
Cases
Cast: Lawyers
Comes v. MS
Contracts/Documents
Courts
DRM
Gordon v MS
GPL
Grokdoc
HTML How To
IPI v RH
IV v. Google
Legal Docs
Lodsys
MS Litigations
MSvB&N
News Picks
Novell v. MS
Novell-MS Deal
ODF/OOXML
OOXML Appeals
OraclevGoogle
Patents
ProjectMonterey
Psystar
Quote Database
Red Hat v SCO
Salus Book
SCEA v Hotz
SCO Appeals
SCO Bankruptcy
SCO Financials
SCO Overview
SCO v IBM
SCO v Novell
SCO:Soup2Nuts
SCOsource
Sean Daly
Software Patents
Switch to Linux
Transcripts
Unix Books

Gear

Groklaw Gear

Click here to send an email to the editor of this weblog.


You won't find me on Facebook


Donate

Donate Paypal


No Legal Advice

The information on Groklaw is not intended to constitute legal advice. While Mark is a lawyer and he has asked other lawyers and law students to contribute articles, all of these articles are offered to help educate, not to provide specific legal advice. They are not your lawyers.

Here's Groklaw's comments policy.


What's New

STORIES
No new stories

COMMENTS last 48 hrs
No new comments


Sponsors

Hosting:
hosted by ibiblio

On servers donated to ibiblio by AMD.

Webmaster
The Code - Updated
Tuesday, August 19 2003 @ 01:37 PM EDT

There is a comment on Yahoo! Finance from a poster whose identity is unknown to me. Normally, that would mean I wouldn't mention it on Groklaw, as I have scrupulously avoided anything I couldn't evaluate and verify. But this is something I'd be delinquent not to report, I think, now that it's on the internet anyway and public. So I will just say that I am putting it up with a request that those who can appropriately evaluate it do so. Here is the message in full:
SCO's "proof". A joke.
by: d1rkinator 08/19/03 09:10 am
Msg: 29448 of 29609

  The code SCO finds offending:

www.heise.de/newsticker/data/jk-19.08.03-000/imh0.jpg
www.heise.de/newsticker/data/jk-19.08.03-000/imh1.jpg
Its location in Linux:
/usr/src/linux-2.4.20/arch/ia64/sn/io/ate_utils.c
And its heritage:
minnie.tuhs.org/UnixTree/V7/usr/sys/sys/malloc.c.html
Ok, SCO: This was easy. Now, show us the other many examples.
This is definitely one of those days I wish I were a programmer. Feedback, those of you who are?

UPDATE: Lots of analysis of this available now by Bruce Perens, and LWN and Slashdot. Thank you, Groklaw readers. You are, once again, amazing. I knew you'd come through, but I'm amazed at the speed.


  


The Code - Updated | 57 comments | Create New Account
Comments belong to whoever posts them. Please notify us of inappropriate comments.
radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 10:56 AM EDT
A more authoritative source and complete analysis:

http://perens.com/Articles/ SCOCopiedCode.html


David Person

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 10:58 AM EDT
Here's another analysis (from LWN):

http://lwn.net/Articles/45019/


Jeremy Stanley

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 11:00 AM EDT
Basically, the comment and code are ancient. They've evidently been released
for public use multiple times, and are even part of the case AT&T lost. style="height: 2px; width: 20%; margin-left: 0px; margin-right: auto;">Jonathan
Williams

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 11:06 AM EDT
Additional commentary:

Slashdot - "Stolen" SCO Linux Code Snippets Leaked


Jonathan Williams

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 11:23 AM EDT
I've seen versions of this code in version 6, version 7, version 32v from Bell Labs, 3BSD, 4BSD, 4.1BSD, and 4.2BSD.

A version of this code also appeared in "The C Programming Language by Kernighan and Ritchy, pulished in '77 or '78.

What appears to be Greek in the first slide, is *not* source code, but a contiunation of the original comment. SCOGs obfuscation appears to be greek, and somebody has posted a translation in another thread here on GROKLAW. The obfuscated comment seems to be about changing naming conventions for modules.

#include <stdio.h>

main (){ printf ("standard disclaimern"); printf ("not a lawyer, not a paralegal, just a codern"); }


D.

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 11:29 AM EDT
Kind of makes you think that, if anything, the comment was lifted from Linux and
then modified, doesn't it?
Jonathan Williams

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 11:29 AM EDT
Also, see this slashdot comment: http://slashdot.org /comments.pl?sid=75246&cid=6732781

It does appear that this code was in V5 and V7 -- but that code was released by Caldera under a BSD-like license in January 2002 (http://www.tuhs.org/Archi ve/Caldera-license.pdf) in addition to being released under the BSD license from 2.11BSD in 1986 (http://unix-archive.pdp11.org.ru/PDP-11/Trees/2.11BSD/sys/sys/subr_rmap.c)< p> To top it all off, it looks like the patch that introduced this code came from SGI, not IBM...


Jeff Randall

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 11:34 AM EDT
PJ, It was more than just your Groklaw readers; hundreds all over the world did the research.

I have the feeling the open source movement can handle more of SCO's claims when they become public. :)


MathFox

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 11:46 AM EDT
The patch was submitted to the linux tree by HP not SGI, which is very funny considering HP sponsered the event.
Supa

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 12:09 PM EDT
From the Screenshot:

if (size == 0) return) ((ulong_t NULL);

Now i'm no expert in programming, but last time I checked, that wont complie. Interesting ??typo?? for a duplicate "copy"


SD

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 12:11 PM EDT
bah, it didnt take one of the carriage returns.

should read (hopefully works this time)

if (size ==0) ) return) ((ulong_t NULL);


SD

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 12:23 PM EDT
From the back of the title page from "Lion's Commentary on UNIX 6th Edition with Source Code":

Portions reprinted by permission of The Santa Cruz Operations, Inc. (SCO). to the extent SCO has an intellectual property interest in the material contained herein, SCO has granted a license to publish solely for the purpose of creating an educational work. SCO grants no license for any other use of this material. None of the computer software source code contained in this publication may be copied, distributed, modified, transferred, or used in any way. It has been printed here solely for educational purposes in the study of an historical landmark in computer history.

The malloc.c source is found on sheet 25 in the book. But the above license no longer matters since the source code for it was released under a BSD-style license after the book was published. Here is the malloc.c that was published in the Lion's Book:

ftp:// ftp.tribug.org/pub/tuhs/PDP-11/Trees/V6/usr/sys/ken/malloc.c

One of the many variations of malloc.c that are available:

ftp:// ftp.tribug.org/pub/tuhs/PDP-11/Trees/V7/usr/sys/sys/malloc.c

The license that Caldera used to release the old UNIX code:

ftp://ftp.tribug.org/pu b/tuhs/Caldera-license.pdf

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.


dentonj

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 12:41 PM EDT
I loved this from Peren's article: SCO's "pattern analysis team" found this code and correctly concluded that it was identical to code in Unix. But they didn't take the additional step of checking whether or not the code had been released for others to copy legally.

Actually, you don't need a "pattern-analysis team" - you can just type lines of the allegedly copied program text into google.com, and google will show you where that code has been posted to the net.


Greg T Hill

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 12:56 PM EDT
I'm concerned by the advertising clause in the Caldera BSD license for the relevant Unices. Unless the code was ALSO released under a more free license, isn't this code in fact a copyright breach because it does not adhere to the advertising clause (mentioning Caldera and including their disclaimer)?

It's a red herring that the code appears in early BSDs because, afaik, only BSD4.4-lite actually became free after the big AT&T case. Earlier BSDs, which are tainted by UNIX code, are not free?

Somebody (with knowledge of the facts) please correct me.

Dr Drake


Dr Drake

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 01:00 PM EDT
The fact that the code that was allegedly copied into Linux has been released to the public under a BSD-like license is interesting but a few points should be made:

*) BSD-licenses are very similar to the GPL in that they grant the recipient broad rights beyond the rights granted to the recipient under copyright law. However, these rights are granted on certain conditions. Without these conditions it could be argued that such license would effectively place the affected code in the public domain. The conditions of both the GPL and BSD licenses make very clear that the work is not placed in the public domain but that it is still a copyrighted work.

*) The fact that the mentioned code is publically available under a BSD license makes it very hard to maintain that the code should be classified as trade secret. It simply isn't a secret. But merily the fact of being publically available does not imply that the code is in the public domain. It better not, because following the same reasoning it can then be argued that all GPL'ed code is in the public domain as well.

*) If one was to assume that the code has been copied from the code published by Caldera under the BSD-license then it seems that by doing so the license terms of the BSD-license have been violated. "Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer." Thay may not seem a big deal, but I believe it was of material importance in the well-known BSD lawsuit.

*) An interesting question is "Who is the real copyright owner?" There is remarkably similar code available in three different versions with copyright claims by SCO (Caldera), Uni. of Cal. and SGI. The similarity of the code is such that it seems very likely that copying took place. To me it seems that the same code can only be legally copyrighted by multiple parties if parties have scooped it up from the public domain. But is there actual evidence of this code being available in the public domain?

*) Some people claim that the code published in The Lion's book has been placed in the public domain, but that seems unlikely since it contains a copyright notice: ". . . SCO [i.e. the Santa Cruz Operation, not The SCO Group] has granted a license to publish solely for the purpose of creating an educational work. SCO grants no license for any other use of this material . . . ."


inc_x

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 01:10 PM EDT
Bruce Perens has remarked on the advertising clause issue:

"(Posted Aug 19, 2003 16:19 UTC (Tue) by BrucePerens) (Post reply) 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"

and

"(Posted Aug 19, 2003 16:38 UTC (Tue) by BrucePerens) (Post reply) 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"


Jonathan Williams

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 01:11 PM EDT
inc_x: adding to your observations:

The algorithm used was allready well-known in 1960. It is likely that implementations of the algorithm can be found that predate both UNIX and C. One should seriously consider the possibility of several independent translations of the same "pseudocode" into C.


MathFox

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 01:14 PM EDT
I would imagine the code in question was not copied from Unix to Linux or vice versa, but rather they both have a common (and very old) source.

Regardless, as I understand it the code in question was removed from Linux a couple of months back, anyway, largely because it was poor code :)


Jonathan Williams

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 01:20 PM EDT
replies to:

Jonathan Williams - I honestly don't really understand Bruce's first comment. "I assume that" and not knowing the terms of the BSDI lawsuit make his answer less than reassuring. As to his second comment, that's on a different topic - the question of whether the code violates the *GPL* because it has simultaneous nore restrictive licensing. That's not relevant to whether it violates the BSD license.

inc_x - For sure there could be independent derivations of similar code, in which case there's no copyright breach. But the similaries include variable names and even, from what I've seen, indentation (please correct me if I'm wrong) to the Sys7 version. That makes an independent derivation of Linux implausible, although it leaves the option of both Linux and Sys7 both being copied from another source of code (but not pseudocode).

Despite the triumphant comments on Slashdot, I am not convinced that the community has yet refuted any claims of copyright breach with this particular code. I think we should not get carried away. I agree that it's very simple code and there's likely to be a copyright-free source for it somewhere, but until it's found I remain skeptical. I take more comfort from the fact that it isn't in kernels 2.5 and later (it was removed from 2.4.21 or somewhere close to that).

Nonetheless I would like to see some mainstream press articles suggesting that SCO's presentation is misleading. Does anyone have any contacts whom they can give the evidence-so-far to?


Dr Drake

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 01:40 PM EDT
Dr. Drake,
I believe Bruce meant "BSD", not "GPL".
Jonathan Williams

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 01:40 PM EDT
A talkback at LinuxToday (http://linuxtoday.com/news_story.php3?ltsn=2003-08-19-020-26-OS-CD-EV-0004 ) cites a couple of cases on "mitigation of damages", which appears to scuttle SCO's license program, (not to mention their damages vs IBM) but good.

Ford Motor Co., 458 U.S. at 232 n.15 (quoting C.MCCORMICK, LAW OF DAMAGES 127 (1935)) "[w]here one person has committed a tort, breach of contract, or other legal wrong against another, it is incumbent upon the latter to use such means as are reasonable under the circumstances to avoid or minimize the damages. The person wronged cannot recover for any item of damage which could thus have been avoided."

If this pricipal still holds, then SCO's case against IBM, in light of SCO's refuseal to identify to the kernel crew what code needs to be removed, is toast.


Larry

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 01:43 PM EDT
And, of course, the fact that SCO has released the Linux source under the GPL
pretty much makes the issue moot, from what I understand. It may not be
entirely fair (assuming there is, in fact, infringing code), but SCO should have
checked the code before they released it. Frankly, you'd think SCO would have
thought to do so, considering the well-known players who have worked on Linux,
who also had licenses for Unix...
Jonathan Williams

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 01:46 PM EDT
Jonathan - I'm sure Bruce did mean GPL; that's the context of the conversation. He can't have meant BSD, because the copyright owners of the BSD'd Sys7 code in question is Caldera, now SCO. They surely aren't giving anyone licenses for anything, at least not without a fee of $699 per CPU.

Also, I feel that Bruce is quite wrong to say that "If this came down to attribution and the advertising clause in the Caldera license, they'd have no case anyway". As far as I understand it, a very important part of the judge's decision on the USL/BSDI case was that one party had not adhered to an attribution clause of the other's license.

We cannot be cavalier about this - a poorly argued defence, subsequently effectively rebutted by SCO, will make us look all the worse. Indeed, it just adds fuel to their argument that "Linux coders don't believe in or understand copyright". We need a watertight defence which gets circulated in the mainstream press and to which SCO are unable to respond - that will get the message across to the public quite well.

Maybe we need a local paralegal to provide one for us? ;)


Dr Drake

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 01:47 PM EDT
Someone take a copy of this: http://perens.com/Articles/ SCOCopiedCode.html Print it out. Then hand it to everyone at the SCO meeting that walks the street outdoors. Stuff all the car windshield wipers. Post it on all the walls. Everyone at the SCO show/meeting needs to know about this!
annon

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 01:51 PM EDT
Jonathan - in reply to your most recent comment which slipped past as I was writing my previous one. That's not a good legal argument. SCO could easily be forgiven for not noticing some of their code in the 5 million lines of Linux (well, perhaps if they claim "milllions" of them are theirs, but not if they keep their claims to a lower number). If they accidentally redistributed GPL'd code, it does not suddenly clear whoever first added their code in to Linux.

I'm not saying I agree with SCO in any way at all. But the "SCO distributed the code under GPL" argument is only strong because they *continued* to do so after they knew about the violations. And they could claim that they were just acting in good faith to uphold the GPL as best they could.

I know that there are gaping holes here, but a truly unknowing publishing of under GPL of one's own code cannot possibly be binding.


Dr Drake

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 02:09 PM EDT
It should probably be noted that this code rather neatly fits Bill Claybrook's description of some code on the periphery of the kernel that was probably most relevant to SGI and the code being something that would be familar to most old Unix hands.

It should also be noted that as this code appears in arch/ia64/sn it will only be included if you are building a kernel for an Itanium machine, that means mainly SGI Altrix and HP SuperDome although I think IBM have got one too now. The SCO binary license is irrelevant to x86 users in this case as the only copy of that code they will have is the source (if they hav installed kernel source)

Further points to note are that it was only introduced in kernel 2.4.19 http://www.funet.fi/pub/Linux/PEOPL E/Linus/v2.4/patch-html/patch-2.4.19/linux-2.4.19_arch_ia64_sn_io_ate_utils.c.ht ml It would be worth identifying the kernel release where it was removed again.

The earliest copy I can trace is from 1973 at http://mi nnie.tuhs.org/UnixTree/Nsys/sys/nsys/dmr/malloc.c.html where it appears to have been contributed by its original author (Dennis Ritchie) with no copyright notice at all. It is also reputedley in the first edition of "The C Programming Language" by Kerninghan and Ritchie but I only have the second edition so can't verify that.


Adam Baker

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 02:13 PM EDT
Point taken regarding Bruce's comments about GPL license holders.

There are several important aspects to this, I think: 1) the pedigree of the code in question is not clear for either kernel, and if that's the case for the example SCO chooses to showcase, what about the rest? 2) What contractual rights did SGI have regarding UNIX code, since it would appear they were the ones who contributed the code in question? 3) It would appear, from a BSD license point-of-view, that the worst that can be claimed is that code was used without proper attribution. It's rather difficult to claim, however, that trade secrets were infringed, or that substantial damages were incurred by the use of the code itself, since it had been released for free use; 4) The offending code was removed faster than promptly, once even a whiff of its presence was known; 5) SCO has not made a good-faith effort to resolve the alleged infringement.

IANAL, of course, but it seems to me that the above points would be relevant to any case.

Whoever added any infringing code would probably be liable; however, it seems to me that this would still be a separate issue from whether SCO could now charge LINUX users for their IP. That's what I was getting at regarding their distribution of the code under the GPL -- and as you've noted, they're now claiming that up to 25% of Linux is infringing, and yet they've continued to distribute the code.


Jonathan Williams

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 02:23 PM EDT
Dr. Drake, what you're talking about has been discussed here a LOT.
1. Caldera had employees working full time to put UNIX code into Linux (Hellwig, Avazian)
2. Ex-CEO Ransom Love gave speeches to the effect "Caldera is putting SysV into Linux"
3. old SCO, Santa Cruz Operation had employees contributing code
4. old SCO's management knew about and encouraged #3.
5. Sun has put out their own version of Linux under GPL. If Sun's licenses are as good as SUN claims, SCO's claim of no-GPL is again voided, this time by SUN.
6. Sun is doing ANOTHER version of Linux, Mad Hatter. SUN is AGAIN releasing UNDER GPL everything SCO claims is NOT GPLed.
7. >> "SCO distributed the code under GPL" argument is only strong because they *continued* to do so after<<

No. this is the pregnant cow defense, and we've heard from attorneys that the only reason the pregnant cow defense originally worked is that the purchaser had not yet taken delivery of the cow, the sale had not been completely finalized, the 2 parties to the contract had not lived under the contract for any length of time.

We've been assured that if the purchaser had taken delivery and both purchaser and seller had lived under the sale for even a few weeks, the "I thought she was barren" defense would NOT be valid.

Apply this to SCO/Caldera, which shipped Linux under GPL for 9 years (the 'pregnant cow' would have died of old age), and shipped IBM's allegedly unauthorized contributions for 2 years, and shipped Santa Cruz Operation's contributions for several years, and that pregnant cow defense is not totally shredded.

and that's the tip of the iceberg. We don't know yet if Novell can compel SCO to waive ALL their contractual rights to IBM's IP or Dynix IP, for example.

Points 1-5 above are applicable as to laches and waivers on SCO's claims, as well as evidence of SCO's support of the GPL.

IANAL, I'm just reporting all that I've read on some of these issues.


Sanjeev

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 02:27 PM EDT
This is not a coding issue; it's a legal issue. As such, us programmer types have no valid input...
Chris Curran

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 02:28 PM EDT
correcting myself ....
>>> and that pregnant cow defense is not totally shredded.

should of course be "and that pregnant cow defense is totally shredded, IMHO (from my keyboard to the judge's mind ... ) "

</self-flagellation over proofreading>

Jonathan Williams, add a


at the end of each numbered point.

(I hope that html turns out the way I intend)


Sanjeev

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 02:29 PM EDT
Adam; I think you should read the ruling in the USL-BSD case at http://cm.be ll-labs.com/cm/cs/who/dmr/bsdi/930303.ruling.txt AT&T forgot to add the copyright notices that were required to obtain copyrights before 1978:

Consequently, I find that Plaintiff has failed to demonstrate a likelihood that it can successfully defend its copyright in 32V. Plaintiff's claims of copyright violations are not a basis for injunctive relief.

and

In summary, I find that I am unable to ascertain whether any aspect of Net2 or BSD/386, be it an individual line of code or the overall system organization, deserves protection as Plaintiff's trade secret. Since Plaintiff has failed to provide enough evidence to establish a "reasonable probability" that Net2 or BSD/386 contain trade secrets, I find that Plaintiff has failed to demonstrate a likelihood of success on the merits of its claim for misappropriation of trade secrets. No preliminary injunction will issue.

I think it is worth to realise the implications of the copyright ruling. This is not a verdict but it provides an indication of the direction the verdict would take. If AT&T couldn't enforce its copyrights because of mistakes it made, SCO is very unlikely to enforce the same copyrights.

#!/bin/python
print "I am a programmer, not a lawyer"


MathFox

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 02:31 PM EDT
The key point of the USL/BSDI case as I understand it was that both sides had got some bits of the others code in them and neither side was complying with the requirements (USL code = keep it secret, BSD code = attribution reqd)

A post apparently from the best known Linux contributor who used to work at Caldera (http://www .ussg.iu.edu/hypermail/linux/kernel/0305.0/0237.html) suggests that that would be the case again - I suspect from comments elsewhere he is referring to ext3. Of course if K&R edn 1 really includes the malloc code or the nsys code that Dennis Ritchie released really has no copyright claimed on it then Linux is off the hook for now.

Someone suggested in another thread that Multics, the system Unix was supposed to be a cutdown version of never saw the light of day. Having used a Honeywell Multics system I can confirm that it definitely was released - the Unix relationship was visible and some of the features that got dropped to make Unix a viable 2 man project were actually quite nice.


Adam Baker

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 02:33 PM EDT
I think you meant a <br>

right?

Sorry about that...


Jonathan Williams

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 02:41 PM EDT
In respect of my previous comment about the GPL, re-reading I think I was
speaking rubbish. Sanjeev's arguments in particular form a convincing (to me)
case.
Dr Drake

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 02:57 PM EDT
MathFox - point taken

The noticeless distribution seems to have been a real killer for 32V and earlier. After reading that I'm surprised that SCO even tried to claim a BSD copyright when they released them.


Adam Baker

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 03:12 PM EDT
Jonathan Williams, how'd you get that to show up? I enclosed it in blockquotes and PJ's system chewed it up.

Dr. Drake, for a while I was also convinced that the pregnant cow defense applied legitimately to SCO (I think my comments on that are still on the twiki), and further reading convinced me otherwise.
For a while near the beginning of this whole thing I thought "let's find SCO's code and get it out ASAP."

Now my reaction is a bemused, befuddled, scratching-my-head " ?? ??? WHAT IS DARL TALKING ABOUT? ??? ??!!!! ???? ???? "
Sometimes I get the feeling SCO is a haven for lost USENET trolls who've stumbled onto a bigger audience.


Sanjeev

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 03:23 PM EDT
(This is mostly in reply to Sanjeev)

I don't think we should be surprised that SCO's comments are confusing to us. After all, they aren't intended to enlighted - they are intended to spread FUD, look plausible enough to PHBs, and generally throw our community off track.

The more I read of McBride, the more I want to ignore his statements as diversionary and take a closer look at SCOs legal filings. Sadly, they haven't really filed very much as, as far as I know, they haven't replied to RedHat's suit and they didn't respond to the injunction (or whatever it was) in Germany. And with me not even being a paralegal, I can't understand the filings anyway (notwithstanding PJ's excellent help).

What I don't get is why RedHat didn't file for a preliminary injunction (I haven't verified that they didn't myself, but I heard that they didn't), or why the FSF haven't. Surely the most important thing is to stop the FUD right now - there's no question that it's damaging the business image of Linux? Can PJ comment on the usual requirements for a preliminary injunction to be awarded, and whether there would be a case to answer if one of the Linux copyright holders did so?

One important difference I do understand is that (under UK law anyway) libel of a business has to have malicious intent (not something needed, under UK law, for libel of a person). But that's clear enough here. And I thought it was fairly easy to get an injunction to stop a libel. Maybe UK law is different - I know the burden of proof for the truth defence of libel is very different. Perhaps some Linux users over here ought to consider a suit of their own.

Finally, I'd like to reiterate that it's vital for the Open Source community to get a solid argument together. A weak case is worse than saying nothing. I'm still not convinced that Bruce's rebuttal (now referred to on at least one tech news website apart from the OS community) is solid.


Dr Drake

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 03:46 PM EDT
Dr Drake

I've read comments here on what's required for a prelim (not in one long commentary that I've come across here, but scattered stuff.)

The concensus seemed to be when the RedHat suit came out that RH could not afford the bond that's commonly required in case you turn out to be wrong in getting the prelim.

pj did discuss the declaratory judgement that RH is requesting long before RH filed for it, it was also discussed when RH filed. browse around this site for a while, RH's filing was around Aug 5 - you'll run into the declaratory judgement article sooner or later.

groklaw seems hidden from GOOGLE.

http://lwn.net/Articles/42457/

>>> Surely the most important thing is to stop the FUD right now -

I don't know. The believers are galvanized by it, the trolls are behaving in their usual way, various technical press types are showing their true colors (or at least the colors they're required to show by the guy who signs the paycheck) .... It looks bad from some sides but overall it's not really that bad. In the end we'll come out stronger.

I think if business was hurting IBM would have filed for immediate preliminaries/immediates, and they can afford the required bonds.


Sanjeev

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 03:53 PM EDT
Thanks for the links - I'll take a look at them tomorrow. I was aware of the discussion on the declaratory judgement but not the prelim.

I do think that stopping the FUD is vital, for the long term image of Linux. (e.g. Microsoft is going to bring up "IP problems" forever more). Whatever the outcome of various lawsuits, it will be very difficult for any judgement to get across to the public quite how outrageously SCO is behaving right now and how wrong they (presumably) are - better to stop the behaviour soon. Some people argue that the relative dominance of Linux over BSD (at least in publicity terms) stems from the lawsuit which left BSD under a cloud for some time.


Dr Drake

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 04:13 PM EDT
It seems you no longer need a pacer acoount to get some of the court filings as this has been declared a high profile case http://www.utd.uscour ts.gov/documents/profpage.html

It looks from there as though nothing has happened since IBM filed their counter claim 2 weeks ago. Some initial discovery exchange seems to have occured before then but discovery on matters of fact can continue until 4 Aug 04, until 22 Oct 04 for expert witnesses.

I believe it may be necessary to put up a bond of some sort for a preliminary injunction to cover the damages if you lose - I can imagine IBM might not want to do this for the patent infringment if they aren't yet sure they have made the strongest possible claims in that area. Even RedHat could potentially put SCO out of business by forcing them to shut up as they now have no other business model so the bond req'd might be too big for them.

I agree that Bruce's paper looks weak - The arguments that USL vs BSDI says that code (as it appears in 32V) is in the public domain and the other publications if they can be verified would make it watertight. In a court of law you can come out with another reply whenever the opponent reveals new information, in the court of public opinion you've got to make the opponents reply look weak as you may not get your response printed.


Adam Baker

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 04:17 PM EDT
In regard to the "pregnant cow" defense, if SCO could show that it was tricked into including valuable proprietary code in its Linux products and simply wanted to demand the removal of that code within a reasonable amount of time, I could buy that as a legitimate position. In such a case, the (figurative) cow has a huge number of joint purchasers (the entire Linux community), and SCO was was relying on information from one or a handful of those purchasers in thinking the cow was barren. Under such assumptions, the fact that SCO was deceived by a person or persons with a vested interest in obtaining the cow would call for giving SCO more latitude in getting the cow back than would make sense if SCO, or even a third party with no vested interest, originated the mistake. That's not to say that SCO would be justified in demanding their cow back so suddenly that babies go hungry because no other cow is readily available, but their desire to get their cow back within a reasonable amount of time would be both reasonable and compatible with the interest of justice.

But look at SCO's actual behavior. The Linux community keeps saying, "Tell us which cow in our herd is yours and we'll give it back." And SCO is saying, "No. We'll only tell people which cow is ours if they promise not to give it back or identify it to anyone else who might give it back. In the meantime, we demand that everyone who buys milk from you (or even that gets it for free) pay us five dollars a gallon because a few drops of that milk come from our cow." The idea of using a "pregnant cow" defense in such a manner is completely absurd, and I seriously doubt that SCO could find a precedent that could even begin to justify such behavior.


Nathan Barclay

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 04:38 PM EDT
Sanjeev, I entered the code like this:

&lt;br&gt;


Jonathan Williams

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 04:39 PM EDT
Ah, well. Let me try that again...

&lt;br&gt;


Jonathan Williams

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 04:40 PM EDT
There we go.
Jonathan Williams

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 04:47 PM EDT
PJ,

If you were a programmer, you would find the whole thing more hilarious than you already do ;)

With this slide demo, one can assume SCO wanted to reassure there resellers by the best example they can find. So, "pattern analysts" found some code from the 70's.

I imagine this code was chosen due to the dense commentary (comment are not code, and it' what SCO called DNA). Comments are valid proof that the code was basically cut and paste, straight out plagiarism, unless there was permission granted.

For the code in the slide, there needs be no permission. As that code falls under the Berkley vs. Novell law suit.

I haven't gone through the exercise yet, but here's a draft of a judge will eventually have to go through, in deciding if there is any illegally copied code in Linux:

For any code similarity to be deemed illegal, or even worth investigating, SCO needs to present the following source code:

1. The code was written after the 80's law suit between Berkeley and Novell/AT&T, as that code is explicitly deemed not a trade secret.

2. The code must not be part of ancient Unix, V7 or pre-V7, as SCO explicitly released all that code in 2002 under a BSD-like license.

3. The code must not be part on UnitedLinux or Monterey, as SCO/Caldera explicitly contributed such code to an Open source Linux distribution under GPL.

4. (Maybe) The code should not be available through SCO public ftp site, neither now, nor in the past. Any such code should be restricted to authorised customers/licensees

5. The code must not be part of "grandfathered" code released or deemed none trade secret by a court settlement by SCO or any corporation that SCO grandfathered, or grandfathered some of their code property. (Similar to #1)

6. The code must not be developed by SCO/Caldera engineers hired by SCO/Caldera to code for Linux. There were many such developments easily found in Google groups (USENET).

7. The code must be developed by SCO, or ancestral companies, outside the above excluded code.

8. The code must not be the intellectual property of another company or copyright holder that was not grandfathered by SCO/Caldera (such as Sequent etc.)

9. The code must be proven by a CVS system to be logged prior to the entry log in Linux's source tree. i.e. SCO must submit, or provide access to their source tree and CVS revisions to show the code was submitted into their tree, without benefiting from Linux's publicly accessible source tree. In simple terms, since SCO's products are closed source, the burden is on them to show their code preceded the equivalent/similar Linux code portion in question.

10. Only SCO/Caldera source code that matches the above criteria may be submitted for comparison to identify any code plagiarism.

Notes: 1) The above criteria is extremely lenient in favour of SCO/Caldera, as it allows them to cling to the argument that what they released under GPL was by mistake, and their copyrights should not be lost just because they released an entire Linux distribution before that inspected the code line by line.


tamarian

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 04:47 PM EDT
Browsing around the web on this issue (but I could have missed it), I haven't noticed anyone else wondering: <devil's advocate mode on>

What if SCO knew exactly the history and current status of the code on the Heise site?

1. It shows that UNIX code even from 1973 and thus their IP has been copied to Linux.

2. Did they ever say this was infringing code? If not, they are still abiding by the confidentiality clauses they they say will not allow them to show Linus et al the code for removal, and no need for the NDA in the speech.

3. It is publicly available licensed code, so they aren't giving away any new IP.

4. The copyright statement says SGI but it was submitted by HP and the original says AT&T. Clearly Linus does not check origins and copyrights of code before accepting patches into Linux.

5. The code was accepted in 2.4 then rejected in 2.6. It becomes a non-issue in the future releases and the law suits.

6. The Linux community has expended an awful amount of hot air and possibly injudicious words over the code.

7. Add some trivial encryption and typoes to confuse the issue and get people thinking about the DMCA too.

8. Chess players will have heard of the Greek (!) Gift sacrifice (Bxh2 or h7).

Is Drool McBeth (that was another thought) laughing his head off at us right now?


Chris Priest

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 04:52 PM EDT
I think the only leg remaining for SCO to stand on is this: The terms and conditions of their contract with IBM regarding derivative works, and how those derivative works are defined in that contract.

I find it extremely negligent, if IBM did sign all their copyrights and patents on their own work and intellectual propert to SCO, if IBM put those technologies in AIX.

Seeing that IBM pulled an irrevocable license and such favourable terms through their lawyers, I can't imagine them selling all their rights in the same contract. :)


tamarian

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 05:19 PM EDT
Tamarian,
I think that's what this has always been about, and I think it's the only leg
they've had to stand on since they started distributing Linux under the GPL.
That they believe they can make it about anything else, even going so far as to
make public boasts about absurd legal arguments, simply shows that we're dealing
with some rather foolish people.
Jonathan Williams

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 05:47 PM EDT
Jonathan, I agree.

Now I wonder, since all the suits and counter suits should be public records, why can't we access those submitted exhibits (SCO-IBM contract re: derivitive work)? They were referenced in the filed documents by both parties.

PJ, any legal explanations for us non legal programmers (re: exhibits)? :)


tamarian

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 06:15 PM EDT
Tamarian, everything that is available online at the courthouse is now here:

http://www.utd.uscour ts.gov/documents/ibm_hist.html

The case is now a "high-profile" case, so you don't need a Pacer account any more.

Some exhibits are there, some are on SCO's web site, and some are not in electronic form. If someone who lived in Utah went to the courthouse, they might be able to get access to them. I don't know the policy of that court house, but normally you can. That's the only way to get a copy of the paper exhibits, other than to ask one of the parties to the suit if they'd give you a copy, that I know of. One of the derivative works side agreements is attached as an exhibit to SCO's complaint and is available there on their site or on Twiki. Links in my Legal Links page. But there are some 400 side agreements, or something like that, I've heard, so for some things, we'll just have to wait.


pj

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 07:24 PM EDT
I checked K&R 1st ed (at least I think so...its copyrighted 1978) and I didn't see that code in either Chapter 5 (pointers and arrays) or Chapter 8 (Unix System Interface). Both have examples of some form of alloc but none like calloc (which actually is an exercise left for the student).

Now, I'm fairly certain I wasn't programming in 1978 but I've had that book forever...heh the comment in the copyright section indicates that the book was typeset in Times Roman and Courier 12 by the authors on a PDP-11/70.


vinea

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 07:29 PM EDT
Chris Priest, the theory that this is decoy code is interesting.
First, this code fits the descriptions of several of the NDA signers.
Ian Lance Taylor said he found the code on the Internet in places other than Linux, that it was "basic" and "esily replaced"
Trevor Marshall (someone correct me) (Byte) said "the code would be more interesting to those using SGI Altix servers "...
I can't remember exactly what Bill Claybrook said but it's been reported that the Vegas code matches what he described.

So this is (some of) the code that SCO has been showing in their NDA code show. On to your questions

1. "their IP " from 1973 has been severely impaired in many, many ways.
1a SCO's own buddy, HP (via Linux Alpha/IA64 port master David Mossberger) put that code into Linux, NOT IBM. 2. yes. 3. it's publicly available but has not been singled out until now by SCO as "infringing SCO's IP". They have been unwilling to show it until now because they've claimed it's a trade secret. Since they made such a huge deal out of their trade secret NDA show, this little drama shoots a lot of holes in their credibility. 4. Clearly you don't read every book, newspaper article, trade journal, patent application in existence from every country throughout history to determine if your next work will infringe on someone's IP. Why should Linus? Why should anybody? It's up to a book's author to come to you and say "you stole that specific thing, stop stealing it." Always has been.
Why should SCO get special treatment that nobody else gets?
Have Terry Pratchett, Harlan Ellison, Stephen Pinker, John Gresham ... textbook writers, novelists, short story writers, trademark consultants, and so on, demanded that every other writer read their works and make sure all those other writers don't infringe?
THAT'S BEYOUND RIDICULOUS

5. If this is the best they have to offer it will impact their stock price and their credibility, during an exercise meant to do the opposite. 6. Par for the course. We've been insulted and our voluntary contributions are about to be pilfered. The response was appropriate. 7. ? this is the 1st i've read that anyone is concerned about DMCA issues. 8. What has SCO gained from this that they did not already know? What did the sacrifice gain them, and what did it cost?


Sanjeev

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 07:36 PM EDT
Sorry, let me try again.
1. "their IP " from 1973 has been severely impaired in many, many ways.
1a SCO's own buddy, HP (via Linux Alpha/IA64 port master David Mossberger) put that code into Linux, NOT IBM.
2. yes.
3. it's publicly available but has not been singled out until now by SCO as "infringing SCO's IP". They have been unwilling to show it until now because they've claimed it's a trade secret. Since they made such a huge deal out of their trade secret NDA show, this little drama shoots a lot of holes in their credibility.
4. Clearly you don't read every book, newspaper article, trade journal, patent application in existence from every country throughout history to determine if your next work will infringe on someone's IP. Why should Linus? Why should anybody? It's up to a book's author to come to you and say "you stole that specific thing, stop stealing it." Always has been.
Why should SCO get special treatment that nobody else gets?
Have Terry Pratchett, Harlan Ellison, Stephen Pinker, John Gresham ... textbook writers, novelists, short story writers, trademark consultants, and so on, demanded that every other writer read their works and make sure all those other writers don't infringe?
TO expect that is BEYOND RIDICULOUS

5. If this is the best they have to offer it will impact their stock price and their credibility, during an exercise meant to do the opposite.
6. Par for the course. We've been insulted and our voluntary contributions are about to be pilfered. The response was appropriate.
7. ? this is the 1st i've read that anyone is concerned about DMCA issues.
8. What has SCO gained from this that they did not already know? What did the sacrifice gain them, and what did it cost?

Addendum for point 6. - Anyone who's had anything to do with Linux should know this by now. Attack Linux and you will be attacked back. You may not have been around for the Bill Machrone fiasco.

Bill was the 1st "mainstream" guy with lots of credibility to support Linux. He made a few critical remarks along with a generally advocative stance. The Linux wolves tore him apart. It was hilarious and pathetic to watch.

Machrone himself was great and it was sad to watch but like I said, ANYONE in the Linux world should know this already.


Sanjeev

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Tuesday, August 19 2003 @ 08:55 PM EDT
Oh My! I've been out all day and just dropped by to catch up. malloc.c - A heuristic first fit algorithm! Almost everything that can be done with this algorithm, had been done (with bin packing or scheduling) long before there even were computers. It's small wonder that DMR would use it as a quick and dirty snippet of "applied science". "Common Sense" is now an intellectual property of the SCO Group!

This well known algorithm can differ from an optimal packing by as much at 70%. It was fully discussed in the literature shortly after the birth of Unix. For example, M. Garey, R. Graham, J. Ullman, "Worst-case analysis of memory allocation algorithms, " Proc. 4th ACM Symposium on Theory of Computing, 1972.

It's hardly a Unix trade secret, and not the code we need to go from a bicycle to a luxury car at the dawn of the 21st century!

TITLE 17 > CHAPTER 1 > Sec. 102. In no case does copyright protection for an original work of authorship extend to any idea, procedure, process, system, method of operation, concept, principle, or discovery, regardless of the form in which it is described, explained, illustrated, or embodied in such work.

if (!(year%400) || (!(year%4) && (year%100))) leap_year();

can't be copyrighted, can first fit?


Harlan

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Wednesday, August 20 2003 @ 04:20 AM EDT
http://www.zdnet.com.au/newstech/enterprise/story/0,2000048640,20277500,00 .htm

Torvalds: "Hey, that was what we claimed was the most likely source of common code from the very first time," he wrote.

Perens: "It strikes me that SCO would show their best example. This is it?!?!? Hoary old code from 1973 that's been all over the net for three decades and is released under a license that allows the Linux developers to use it with impunity? If this is their best example, they are bound to lose."


Steve Martin

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Wednesday, August 20 2003 @ 05:51 AM EDT
For those of you asking about IBM's signing derivative works over to the owner of UNIX, the answer to your question lies in paragraph 2 of the 1985 side-letter between AT&T and IBM (Exhibit C of SCO's Amended complaint, provided at http://www.sco.com/scosourc e/ExhibitC.qxd.pdf). This states that IBM owns any changes they make, and they do not become derivative code owned by AT&T (Novell, SCO...)

This does have some possible implications that I haven't seen discussed elsewhere. Since IBM's UNIX license is unusual in that IBM owns any changes they make (instead of the changes becoming property of AT&T/Novell/SCO), what are the chances that IBM made some changes, AT&T/Novell/SCO got access to the changes but forgot about the license differences, and incorporated those IBM changes into their main code base? Does SCO UNIX incorporate unlicensed IBM intellectual property? Hmmmm....

While looking at the IBM side letter, you might also want to check out paragraph 9, which grants IBM the right to reverse-engineer anything in UNIX as long as they don't look at the manuals and code while they do so. That pretty much invalidates every place in SCO's amended complaint where they talk about UNIX architecture, know-how and techniques being stolen trade secrets.

Chris (my first time posting, so forgive me if my HTML goes awry.)


Chris Beckenbach

[ Reply to This | # ]

radiocomment
Authored by: Anonymous on Wednesday, August 20 2003 @ 09:24 AM EDT
Chris Beckenbach, >>> While looking at the IBM side letter, you might also want to check out paragraph 9, which grants IBM the right to reverse-engineer anything in UNIX as long as they don't look at the manuals an <<<

Check out exhibit D, paragraph 6 (on page 4) which removes this restriction as well, as of January 2001


Sanjeev

[ Reply to This | # ]

Groklaw © Copyright 2003-2013 Pamela Jones.
All trademarks and copyrights on this page are owned by their respective owners.
Comments are owned by the individual posters.

PJ's articles are licensed under a Creative Commons License. ( Details )