SCO v. IBM settlement deal is done, but zombie case shuffles on elsewhere One strand of the ancient and convoluted SCO versus IBM legal mess that sought to determine who owns UNIX – and perhaps has a claim over Linux – may be about to end. The case commenced in 2003, but its roots go even deeper. In 1998 IBM, the Santa Cruz Operation (SCO – a vendor of UNIX for x86 CPUs) and others teamed to create …
Does anyone know which components form this dispute? A summary of the case for those only mildly interested at this stage and who can’t be bothered to wade through all the legalese? I would have thought it should be straightforward to resolve - just show us the source code. If it’s really in Linux we can see it anyway, but if there’s concern that revealing SCOs hand might release proprietary code then we’ll know that the SCO zombies were just having us on.
In any event, as long as it isn’t in the kernel, if the court case is lost then surely the solution is to replace any contentious modules…
There are bits that both parties copied from FreeBSD. It is perfectly legal for both of them to do this.
SCO Unix today is basically a re-badged copy of FreeBSD 10, as they ditched their legacy codebase.
Also, remember the three hurdles they have to overcome:
1. They don't own the copyright to Unix. Novel/SuSE owns it. Not sure which one now with the recent reorganisation, but it isn't SCO.
2. They published their own copy of Linux in the form of Caldera. Therefore they would have granted a licence to any copyright they might own under the terms of the GPL.
3. They need to prove that any code was written by them and not by for example FreeBSD.
SCO Unix today is basically a re-badged copy of FreeBSD 10, as they ditched their legacy codebase.
To be fair that isn't really accurate. It's FreeBSD kernel with a few extras for ABI compatibility. The userland on top of that is the original for compatibility, not FreeBSD. Try migrating that old SCO app to vanilla FreeBSD you will very likely run into snags.
They need to prove that any code was written by them and not by for example FreeBSD
Which they spectacularly failed to do in a presentation made (as I recall) in one of their SCOsource annual conferences in front of a significant portion of their developer base.
They also tried to claim in court that SGI was violating their copyright for a kernel-based memory allocator used in XFS (used in SGI's Irix Unix, and migrated to Linux when XFS was ported to that platform), only to discover the hard way that that code had never been compiled into the kernel, and that the offending source had actually been removed from the Linux kernel codebase years beforehand.
This was never about royalties as it rapidly became apparent that there was no realistic claim to be had.
This was about maintaining a potential Bogeyman for Linux.
Now, can anyone see a smoking gun?
https://www.networkworld.com/article/2331322/microsoft-behind--50-million-sco-investment.html
No but the point is that either the MSFT dealers believed this, or had bene briefed by somebody to say this. And how many of our potential customers believed this, or had lawyers who were prepared to 'raise concerns' about this ?
There is no "Reality" in FUD
There were multiple claims against multiple parties. The main ones against IBM were mainly:
1) That IBM bailed out of the Monterrey partnership.
2) That IBM contributed their JFS file system driver to Linux, and that this driver supposedly contained unspecified Unix "IP" which SCO was suing over, while being very vague on what the actual grounds were.
Claim #1 falls down because the original SCO were the ones who bailed out of the partnership, and sold their interest to Caldera (who renamed themselves The SCO Group, commonly known as SCOG to people following the case). IBM then exercised their right to withdraw from the partnership as specified in the clause covering "change of control".
Claim #2 falls down because the JFS driver that IBM contributed came from OS/2 not AIX, and was an independent implementation containing no Unix copyrights. Not many people use the Linux JFS driver anyway, so it was always a peripheral issue for anyone outside of those two companies.
The Linux community always saw the Monterrey partnership break up as something not really involving them, as it was a joint project to make a more standardized version of proprietary Unix. I'm not entirely sure, but I think this is the main thing being settled here.
Most of SCOG's "Linux lawsuits" were actually SCOG suing customers of their proprietary Unix products (OpenServer) over alledged minor license violations or suing customers on Novell's behalf (and against the objections of Novell) over similar alleged violations on UnixWare licenses, for which SCO/SCOG were the outsourced collecting agency. SCOG then issued press releases implying these lawsuits were over Linux when there was no Linux involved.
One of the classic ones was where SCOG sued Chrysler over UnixWare licensing issues for their supercomputers. Chrysler were required to file annual paperwork listing the CPUs they were running UnixWare on. They sent a demand to Chrysler for such paperwork, and Chrysler responded with "none", because they had scrapped those supercomputers and weren't running UnixWare any more.
SCOG then filed suit on the grounds that "none" is not a CPU serial number, and that therefore Chrysler were in violation of their license agreement. The judge naturally handed SCOG their arse in court. SCOG then none the less issued a press release claiming total victory in their battle against Linux.
This ancient case used to be covered thoroughly on Groklaw - until personal threats against PJ made that too dangerous. The case starts with Caldera/TSG/... doing a thorough job of delaying making any kind of clear accusation followed by demands for an enormous amount of discovery materials then more delays to read through the discovery materials. Eventually Darl had no choice but to make a verifiable/falsifiable complaint. I cannot remember that far back to what the complaint was but I do remember the refutation:
IBM wrote some design documentation for some features.
Some IBM staff used the documentation to code the features into AIX.
Some other IBM staff used the documentation to code the features into Linux.
Darl claimed to own Unix (rubbish - if it belonged to anyone that would have been Novell at the time)
Darl claimed that anything added to IBM's AIX was his because of his fake claim to Unix.
Darl claimed ownership of the completely different code in Linux because of the fake claim to the code that was not in Unix.
Darl was not a completely clueless idiot. He convinced David Boies of Boies Schiller Flexner to fight this through the courts and all the appeals up to SCOTUS for a percentage of the profits. Boies quickly worked out be had been scammed and dumped all the work on a junior colleague. BSF also sent bills for expenses back to TSG. TSG payed those bills by collecting Novell's royalties (as TSG were contracted to do) and not sending any of the money to Novell (contract required sending 100% and Novell would send commission back). Just before an expensive ruling against TSG could land, TSG declared bankruptcy. Effectively, Darl gets replaced by a court appointed trustee with a far better understanding of nuisance litigation. The trustee promptly spent all of Novell's revenue on legal firms to ensure that TSG's creditors never got any money and dragged this rubbish on for years at the expense of IBM and BSF.
There never was any valid claim. TSG had no standing to sue (that would have been Novell and they instructed TSG to end the litigation). With the help of the bankruptcy court this has effectively become Tax Payers vs IBM and even IBM does not have the funding to win that one no matter how daft the accusations.
[IBM were not the only defendant. There were other claims of ownership to code in the Linux kernel including code created by Caldera and contributed by them to the Linux Kernel with a GPL license. The code was removed shortly after this became an issue - no one was using it anyway.]
Ignoring the fact that Novell owned the UNIX copyrights and that both SCO and Novell licensed all Linux code to everyone under GPL on their websites, SCO never presented any code in Linux that wasn't followed up and proven to be independently written.
My theory is that someone at SCO copied chunks of Linux code into SCO's code base, and they didn't realise it was that way around until *after* they started the lawsuit...
They are not claiming they own FOSS, they are claiming FOSS is built from what they own.
They have never succeeded in proving this. In fact, they tend to get all waffley when asked to produce that evidence.
From what I can tell, the lawsuits are basically fishing expeditions. Or, if you prefer, throwing shit at the wall to see if anything sticks.
The funny thing is that, after all this time, and all the money they have thrown at it, they will never turn a profit even if they succeed!
Why not? I'm glad you asked ... it's because the world at large will stop using whatever code some idiot Judge decides belongs to them ... we will simply code around it, as if it never existed in the first place.
They are claiming the FOSS world took the code without consent. Sadly for them, after the whole MaBell vs BSD debacle, that dog don't hunt no more.
Hopefully,the next Judge in this matter will take the litigant's lawyers aside and explain what kind of affect getting smacked with a Rule 11 "frivolous litigation" contempt charge would have on their career ...
That isn't the issue. If I take your code without permission and slap a GPL on it that code still belongs to you. It doesn't become GPL because I didn't have the right to relicence it.
Of course, the onus is then on you to prove ownership of the code. After almost 20 years this is what TSG and successors still have to demonstrate.
First up: the copyright to new source code either belongs to the programmer or to the person/company that hired him/her. In some jurisdictions that initial ownership can be transferred. In others ownership remains with the programmer/client permanently but a permanent license to distribute can be granted.
These days, contributing to a libre software project usually requires a legal document that grants distribution under the project's license and that the person making the contribution has undisputed right to grant that license.
Going back in time to before this litigation, many projects - including the Linux kernel - did not make the effort to ensure that contributions came with a cast iron license. It was theoretically possible to nick code from an employer and offer it to a GPL project without permission from the owner. The distributors of the project would then be in legal trouble.
Something similar actually happened: The solver code from OpenOffice's spread sheet was contributed with a valid license from the author. The author then changed his mind. In theory the code could have been kept because it came with a valid licence but it was removed anyway to avoid any possibility of litigation.
Some projects require transfer of copyright ownership with contributions. This creates a single point capable of changing the license or starting litigation against someone distributing without following the requirements of the license. The Linux kernel does not require such transfer. Some of it belongs to individuals, other bits belong to organisations. Although individual contributions can be made under a different (GPL compatible) license, it would be a nightmare to change the overall license as that would require permission from every contributor - even the dead ones. Individual contributors could in theory sue distributors who do not follow the license requirements - but only for distribution of the contributor's parts of the kernel.
...the upside, or so the suits have convinced themselves, is a huge payday from Linux using companies and the downside is corporate bankruptcy, which in the USA of America is no deterrent[1].
[1] Bankruptcy laws provide extensive protections for the bankrupt entity. Ask the people of Armley Yorkshire etc, etc..
It's hard to believe that this has been running for almost 20 years. I've been through every twist and turn, watching SCO get demolished and going bankrupt. Linux was, and is, just fine. SCO's attempts to claim Linux had stole IP were pathetic - anyone remember the Greek font switch the tried to use to obfuscate the code? Or the claims that it was millions of lines, which then dwindled to....none at all? All that remains is Xinuous...and that's just another sham. If you want to get some tawdry views of this nonsense, there's STILL a stock-chat board for SCOXQ at Investor Village..but beware...it's a cesspool of right-wing lunacy rather than a serious discussion site.
Indeed. Left wing nuts, right wing nuts, what's the difference? They are all power hungry, and a disease on the backs of humanity. Is there a cure? They say laughter is the best medicine ... and you've got to admit it, they are all pretty clownish.
I propose that all wingnuts should be pointed out and laughed into oblivion. If enough people do it, it should work fairly quickly ... there is nothing they hate more than not being taken seriously.
Imagine, if 50 people a day ...
Why is IBM paying the left over SCO this much money?
Wasn't IBM also countersuing SCO for related matters?
From the article :
The trustee recommends the settlement because it is felt that "ultimate success of the Trustee's claims against IBM is uncertain". While the trustee could litigate the claims, it would be expensive to do so – more so if IBM appealed.
So they admit they may not win the case, and yet get paid by IBM? Arn't the IBM lawyers well feared, with the nickname nazgul? So, why are the nazgul agreeing to this when they should be looking at such a victory that the whole world will not dare trying such crap suits again? Or are the nazgul also getting retrenched with the rest of IBM, and so are powerless to carry on to victory?
It's probably a case of IBM figuring that it will cost more than that to win the case (and with no possible recovery of anything from the bankrupt remains of SCOG), so it's a "minor" go away and don't bother us payment.
One may hope that it will result in the claims being dismissed with prejudice (so they can't be re-filed) and a statement from what's left of SCOG that there is no valid claim there.
If you look back, there was a case brought by Red Hat against SCOX effectively for libeling their software business as being tainted (Lanham Act), which got frozen when bankruptcy was declared. With the settlement, which apparently explicitly leaves out Red Hat as a participant, I wonder if there's a chance of restarting that case and clawing some of those millions back?
It's all here under one roof. The dysfunction of USPTO in granting the patents that were originally asserted then thrown out, our IP laws in general allowing not-practicing patent trolls to shake down entire industries. Those same businesses gaming the system against each other by paying millions in settlement fees to SCO to then tie up their competition in court for decades. After the collapse of the patent claims, having the claims fall back to copyright on 20 year old code.Then allowing an endless stream of zombie companies that are a front for the lawyers slowly draining the accumulated extor-- I mean "settlement" money out of the shell company. IBMs settlement is the icing on the cake, as the way the courts work here, the blood sucking ticks can run up costs in the millions stalling things in court for years, but sell all the IP assets to another company the night before a ruling goes against them, declare bankruptcy and shuffle across the street to a new rented conference room across the street.
We need to reform this stuff, or we might not make it to the next chapter of American history.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
