SCO Lists Specific Code-Infringement Claims

mugnyte writes "Those tireless folks at groklaw have transcribed and published the documents from the latest IBM/SCO hearing. In it, the exact lines of the supposed Dynix / AIX / Linux logic are given. SCO claimed that Linux's read copy update, journaling file system, enterprise volume management system, AIO (Asynchronous I/O), and "scatter gather" I/O code had been derived from either AIX or Dynix/ptx. Now we can take a look at what SCO thinks makes Linux an enterprise-ready platform started at 2.4, stealing away their market share. However, IBM released these things under the GPL ... so what license did IBM really have from SCO to do this? Which raises the question, What license did SCO have from Novell to disallow this?"

Comments

Computerworld article seems to have this sorted ou

[jenkin sear]

This computerworld story seems to have it sorted out:

1) AT&T licenses SysV to IBM and Sequent
2) IBM writes a bunch of cool enterprise level stuff for their flavor of SysV, and acquires Sequent
3) AT&T writes a letter to their newsletter ($echo) saying their license doesn't cover the derivative stuff, just the basic system
3) IBM eventually kicks their stuff into Linux
6) SCO buys up all the old licenses
7) SCO says the work is derivative after all, and they ownzors it

At some point the chewbacca defense starts to look a lot more rational.

Re:So now we have it

[damiam]

Is the code SCO's? Yeah, probably.

Bull. Not even SCO is claiming it's their code (IANAL, but I'm fairly sure about that). They're claiming that it's IBM's code, but the terms of IBM's UNIX license agreement didn't allow them to release it under GPL.

Re:So now we have it

[zjbs14]

Is the code SCO's? Yeah, probably.

Um, no. This is all code that IBM and/or Sequent wrote and incorporated into their own UNIX versions. It did not exist in any code originally from AT&T->Novell->SCO. SCO is attempting to claim that since it was written to work with SysV UNIX, they have the rights to it. Even AT&T specified almost 20 years ago that such code did not fall under the UNIX license.

Not to mention that the JFS code was originally written for OS/2 before it was ported to AIX.

Be careful!

[pb]

DON'T LOOK!

This publicly available Caldera documentation could contain items including but not limited to proprietary, unpublished SCO code, copyrights, trade secrets, and/or patents!

Re:SCO Waiting for 2.6.0 before submitting this?

[Curtman]

2.6.0 is now in freeze mode and it will be really hard to remove all the lines that SCO alleges are infringement

I don't know where you got that idea. Remember when the entire VM was replaced in the midst of a stable kernel?

Re:WTF is 2.4 1-01 ?

[jadel]

According to those busy beavers at groklaw the kernel version they are referring to is *not* a vanilla kernel. It's vanilla-2.4.1 with rclock-2.4.2-01.patch from http://lse.sourceforge.net/locking/rclock.html applied.
Looking at the patch, IBM is listed as the copyright holder. The code also acknowledges that it is based on the Dynix implementation.

A joke, right?

[GreatBallsOfFire]

I scanned through the lists, and found quite a few single line entires in the table. I decided to look one up, specifically net/bridge/br_stp.c. Here's the offending line:

p = br->port_list;

If it wasn't for the history behind this, I'd have to laugh out loud. The only conclusion I can come to is that the majority of the claim comes from structure/union definitions in header files, and they extend it to C files wherever a structure/union member is referenced.

If I get some time, I may do an analysis of the list and determine exactly how little this case actually has to do with IP theft.