OSI vs SCO 655
the jackol writes "As expected, the OSI's just given the SCO vs IBM case a bite with this position paper. "SCO has never owned the UNIX trademark. IBM neither requested nor required SCO's permission to call their AIX offering a Unix. That decision lies not with the accidental owner of the historical Bell Labs source code, but with the Open Group.""
Re:GPL the best bet (Score:5, Informative)
Re:Confused (Score:1, Informative)
G.N.U.
GNU is Not Unix
In full, read up on the Free Software Foundation site as to when and why Richard Stallman developed the GNU utilities and how Linus Torvalds was around at the right time to supply the missing piece to the system. i.e. the Linux kernel.
Re:Confused (Score:2, Informative)
Re:Confused (Score:4, Informative)
Defining Linux is notoriously tricky, since some people primarily rely on kernel heritage, while others try to build a definition based on the collection of software considered part of the "standard" package. This is complicated by various distributions that may or may not use the same pieces of software for similar tasks. The basics may be all GNU/BSD, but once you get beyond that, things can get ugly fast.
Someone making a chart would therefore likely either decide to stuff Linux off to one side on the justification that the kernel has no direct heritage, or draw a crapload of lines to other Unixen and codebases to really nail down every last relation.
Re:Confused (Score:3, Informative)
Not true. Linux is almost completely written from scratch. There are a few ported BSD drivers, but the core OS is a completely new work that shares no code with "genuine" Unix. Of course, Linux distributions included BSD derived code, but the suit here regards the kernel, not userland, and the kernel wasn't derived from anything...
Re:Please don't take SCO seriously! (Score:4, Informative)
You say:
What has SCO Unix got to do with it? IBM never saw SCO Unix code. They're claiming IBM stole stuff from Project Monterey, i.e. UnixWare 7.If you're claiming that UnixWare can't run on 24 CPU's, doesn't have a journaling FS, doesn't have a LVM, doesn't run on NUMA machines I want to know what planet you're coming from.
Re:More Conspiracy ... (Score:2, Informative)
Re:Please don't take SCO seriously! (Score:5, Informative)
I think this article [theinquirer.net] sums it up nicely.
Sigh...
Re:GPL the best bet (Score:4, Informative)
If I distribute my application in source code form, I still have all copyrights associated with it. The GPL grants others some additional rights to redistribute which they otherwise wouldn't have, even if they had the source code.
Major Nitpick (Score:4, Informative)
VERITAS File System (VxFS) and VERITAS Volume Manager (VxVM) are owned by VERITAS Software Corporation. They are not part of the Bell Labs code. By reading Eric's article one could infer (I did) that he was implying this code is part of the Bell Labs code.
Disclaimer: I work for VERITAS Software Corporation.
Re:GPL the best bet (Score:3, Informative)
Some corrections about the histroy of x86 UNIX (Score:3, Informative)
SCO and Intel UNIX: OSI has it wrong and so does SCO. SCO did port Xenix to the 8086 and 286. Intel/AT&T paid to have Interactve Systems port UNIX 5.2.2 to the 286 and 5.3 to the 386. SCO used the Interactive port for the basis of the later products. Interactive built a packaged UNIX based on V.3 which was eventually bought by Sun which used this as the base of Solaris for Intel.
IBM hired Locus Computing to port UNIX to the PS/2. They used a V.2.2 variant and did not use the same code base that was used for the RISC AIX, which was developed by Interactive for IBM.
The OSF ported the MACH kernel and UNIX layer but still used a variety of the Bell commands. I think the kernel was Bell license free, but I can not remember the exact terms. I know you needed an AT&T 5.2 level license, but I think this was because OSF still used the commands, libs, etc from Bell.
In a way, SCO is correct, in that the Intel ports of the various UNIX'es all derive from some version of Bell Labs UNIX which the vendors had access to via an AT&T license.
Re:Sco, Caldera, WTF (Score:3, Informative)
Re:Please don't take SCO seriously! (Score:5, Informative)
> more than 8 processors has more to do with Intel CPU scaling
> issues than kernel support problems.
One thing most people don't seem to mention is that Unixware has a journalled filesystem and volume manager, but they don't own the copyright on that code. It's just VxFS and VxVM from Veritas repackaged and sold by them.
Matt
Re:Yes, we certainly do use AIX. (Score:3, Informative)
This system as a whole, including the Oracle database and the apps which use that database, are parts of a revenue-generating organization. That organization could not operate as successfully without that system, and before it was purchased 7 years ago, our operation was done solely by manual accounting procedures. The financial benefits (expense reduction plus additional revenue) realized above and over the old manual process during that 7 year period are at least twice what that the system cost to purchase and maintain over that amount of time... so the system has made our operation more efficient and profitable.
Re:a good explanation from.... (Score:3, Informative)
But whose theft?
"The exact terms of final settlement, and much of the judicial record, were sealed at Novell's insistence."
"The University of California then threatened to countersue over license violations by AT&T and USL. It seems that from as far back as before 1985, the historical Bell Labs codebase had been incorporating large amounts of software from the BSD sources. The University's cause of action lay in the fact that AT&T, USL and Novell had routinely violated the terms of the BSD license by removing license attributions and copyrights."
Methinks it gets very interesting if that stuff gets unsealed.
Caldera contributed before purchasing SCO (Score:5, Informative)
Re:Please don't take SCO seriously! (Score:5, Informative)
In a quick review of that brochure, I also can only find such astonishing accomplishments like "relational database", "integrated HTTP server" and "Web browser".
If they have journalizing, LVM or NUMA, they seem to keep it a well guarded secret. Maybe you can come up with some proof about these great accomplishements of SCO?
Anyway, Linux could run on 24 CPUs even before Monterey was started.
Re:sale of property an "accident"? (Score:3, Informative)
But you're still missing the point. He's not debating that SCO own the Bell Labs source: that's a record of historical fact. As the document makes quite clear, the Open Group owns the Unix trademark. So whether or not SCO own the Bell Labs makes no difference: IBM need the permission of the Open Group, not SCO, to call AIX a Unix. SCO's ownership of the source (not the trademark) is entirely accidental to this.
In the same way, I'm not denying that you own your car. However, in any discussion of whether I own my car, your ownership of yours is accidental.
Nobody is saying that anything is 'an accident'. You're still working to another definition of the word 'accidental'.
Re:As an attorney... (Score:1, Informative)
Re:sale of property an "accident"? (Score:5, Informative)
As you pointed out, SCO/Caldera did not have any outstanding claims to the source, so they had no integral or premediated claims to the source. The authors could have said 'adventitious' instead.
Your description of objects being bought and sold actually demonstrates how accidental or adventitious ownership works. If I bought a car, I had no previous ownership-interest in it; I'd be doing it out of legal chance, as accidental ownership -- I just bought the car because I could.
A company or person who built the car could say that they have some interest (may or may not be ownership) in the thing, and therefore have some intrinsic reason to buy it. Their ownership would not be accidental. They bought the car because they already had something at stake in the value of the it, perhaps as a demonstration of their workmanship.
Actually, the fact that SCO/Caldera never obtained *ALL* the rights from *ALL* the licences is one of the main points of the article. They make this clear several times by showing different systems that were licenced and SCO has no right to, and systems that SCO released to the public both freely and under the GPL. A second is that SCO/Caldera profited for several years from the actions, including distributing infringing code under the GNU licence and contributing to the code in a public work, but are only now attempting to assert some rights against another company. A third point is that SCO/Caldera probably does not have those rights that it is trying to assert, through the earlier settlement and licence issues, mutally accepted 'theft' of code [which isn't theft if both parties were aware of it and took no official actions], and other history.I think that in spite of some slightly incorrect dates, omitting the free/open arguments and the GNU/Linux OS vs. the Linux kernel, and the inclusion of anecdotes like 'But that emperor has no clothes', the authors have a very clear and solid attack against several aspects of the suit
frob.
Re:Wow (Score:3, Informative)
1. Read the Position Paper. SCO haven't a leg to stand on in the IBM case.
2. Prove it. I cannot understand why Linus, Alan Cox, et al haven't filed some sort of libel suit against SCO. If they'd called me a thief in public like that you can bet I'd be having my lawyer call them straight away and suggest that they retract their accusations or have their arses sued off.
As to SCO withdrawing their versions of Linux, if they did that to avoid being trapped by the GPL then they should have taken the source of the FTP site before they sent out the letters.
I'm tempted to email Darl and officially accept the terms of the GPL pertaining to the copy of kernel-source-2.4.19.SuSE-152.nosrc.rpm that SCO distributed to me the day after they sent those letters out.
By pulling their Distro they have plainly stated that they know how the GPL works and how it would prejudice their position. By failing to stop distributing they have given away any code they claim had been stolen.
Even so they should still reveal what they claim has been stolen from the Bell Labs code. It should be noted that this code is copyrighted, NOT a trade secret, and has nothing to do with the IBM case and SCO has no excuse for not revealing the proof.
Comment removed (Score:3, Informative)
Re:Please don't take SCO seriously! (Score:1, Informative)
They bundle Veritas's LVM stuff in UnixWare; the original poster doesn't seem to know what is in UnixWare anymore than you do. (I believe SCO calls it Online Data Manager, aka Online storage management with RAID?) And yes, there's NUMA support too, despite the relative lack of NUMA x86 hardware out there; did you check google? Try here [byte.com], here [unisys.com] or here [caldera.com].
Regarding the 24 CPU thing, you are comparing apples to oranges. AT&T (later Novell) was adding SMP to SVR4 when Linus was printing AAABBBB to his tty. Running on 24 CPUs with one big kernel lock (in 1998) has no relation to the finer-grained locking that was being put into SVR4 (aka UnixWare) back in like 1991-1995. UnixWare can boot and run on a 32-way machine too. (See the unisys link above if you doubt me.) SCO isn't aiming their marketing at Unisys/Sequent 64-way customers though. Those are 1-off deals. Now I'm not saying UnixWare scales great at 24-way (Linus/Cox don't say that about Linux either btw.) SCO's aiming their brochure and marketing material at the commodity 8-way that your average business can go out and buy from a bunch of suppliers.