Slashdot Mirror
SCO Complaint Filed -- Including Code Samples
btempleton writes "The folks at Groklaw have posted a story including a preliminary copy of Caldera/SCO's amended complaint, including lines of code they allege were improperly included in Linux. The PDF can be found at this story The file lists unix filenames with line numbers and filenames and line numbers from the Linux 2.2 and 2.4 kernels, so folks can now go into real depth."
I thought 2.2 was safe. Then again, we thought they were not going after copyright infringement. I'm guessing that is a typo.
Since SCO has still not actually complied with previous discovery motions, submitted millions of lines of code to IBM in paper form (real class act, they are) and keeps changing their case, my guess is we will see the end of this case, perhaps this year.
UNLESS, of course, the Novell vs. SCO suit sidetracks the IBM suit until we can figure out who actually owns Unix...
Tequila: It's not just for breakfast anymore!
Patches for every single Linux distribution by the end of the week.
And it will include commented lines "*uck you, SCO"
The surprise isn't how often we make bad choices; the surprise is how seldom they defeat us.
SCO still doesn't have the right to subsume all copyrights to the work that everyone has done on Linux to date, If this is correct, I don't know what the ramifications would be, but linux would survive. Isn't this how BSD ended up? All proprietary code was systematically replaced over time, and the result is still free.
I'm not sending anyone a check for $699.
Very interesting. According to SCO:
- Linux is derived from System V. (75)
- IBM has endeavored to control the open source community. (76)
- IBM plans to destroy UNIX. (77)
- Linus Torvalds can't say who contributed what to Linux. (78)
- A significant amount of UNIX source code is present in Linux 2.4-2.6 kernels. (79)
- Linux developers are incapable of developing enterprise-grade software without stealing from SCO. (80, 81)
- Only IBM's involvement in Linux made Linux viable for enterprise use, and because IBM had access to System V (82), if follows that
- if follows that Linux is a clone of UNIX. (83)
The claimed lines of code appear to be in jfs (which is from AIX, not Sconix), evms (once again from AIX), and RCU. Total number of lines is about 600, plus a few complete files claimed to have be contributed illegally by sequent. I fail to see how IBM is prevented by their contract from contributing their own enhancements (or hell, compatible implementations of their filesystems). The rest of the document seems to just be complaining that with IBM's help, linux is going to wipe a lot of proprietary unixes off of the map. Which I believe fails under the legal term "toughus-fucking-luckus."
====
Crudely Drawn Games
They are still flogging JFS, in spite of the widely known reports that both the current AIX and Linux versions were developed from the IBM OS/2 version of JFS.
Any code in common is probably easily found in the OS/2 sources.
The above text was blatently stolen from a groklaw comment.
87. By making the Linux operating system free to end users, IBM could undermine and destroy the ability of any of its competitors to charge a fee for distribution of UNIX software in the enterprise market. Thus, IBM, with its army of Global Services integrators who earn money by selling services, would gain a tremendous advantage over all its competitors who earn money by selling UNIX licenses.
Seems like someone's sore because IBM has a better business model.
"Break out the gin, and the small violin, I'm a raging success as a failure." --Firewater
Even if we ignore what the term IBM PC means, even if we ignore iRMS, even if we ignore OS2 this still leaves AIX/386 which as far as I recall used to run a considerable part of NATO radar infrastructure. OK, IBM insisted on it being useable only on boards with 1M L2 cache, but I it happily ran on much less then that.
Baker's Law: Misery no longer loves company. Nowadays it insists on it
http://www.sigsegv.cx/
I just noticed that the first lawyer listed on this SCO document is "Brent O. Hatch", the son of Utah Senator Orrin Hatch.
Such a move wouldn't be very smart, even if it was technically possible. SCO could easily argue that "those evil linux people removed it because they knew it was infringing code".
It's much better to leave it in, and show a little patience for the legal system. Believe it or not, the Linux community really isn't getting hurt all that much in the corporate marketplace according to surveys I've seen...and the non-corporate linux user base certainly couldn't give a hoot.
I still say this would be a whole lot easier if kernel developers stood up for their work and reputations by doing whatever they can in their respective legal systems(imagine, lawsuits in 30 countries. The RIAA would be proud). So far, all we've seen is a lot of (amusing, but pointless in a court of law) hot air from Linus.
Please help metamoderate.
In this document SCO identify three groups of what they regard as "infringing" code. They readily admit that the number of examples they can provide is somewhat limited due to the fact that they don't possess themselves enough evidence to prove it. "Damn it, I know they stole it from us - if only I could get the evidence to prove it!".
...
The first set (Table A) is JVS code. As we all know JFS is an IBM/AIX creation, so with this SCO will be focussing on their "derived work" argument.
The second set (Table B) is EVMS code. Again, this is a less-mentioned contribution from IBM AIX into Linux. Again, this will be SCO's "derived work" according to their skewed worldview.
The third and fourth sets are the most interesting (Table C and D). They identify stuff allegedly lifted from Dynix (Sequent) code. I could not find the rclock.* or the kmemdef.* files in 2.4.18 or in 2.4.1 (the version they've named), I presume they were removed at some point - Torvalds or someone else could probably identify when.
In Table D, the code they've highlighted in the 2.4.1 apic.c file consists of #include lines, some comments, and a very basic "if" statement in the middle of some SMP related code for handling timer interrupts, it seems. It's the same in timer.c, they're also complaining about lines which refer to Alpha or IA64, rather odd since they never wrote code concerning those CPUs.
The entry.S reference they've made, going by the code comments, refers to code which switches an i386 back into user mode following a system call. (guess mode - the set of assembly mnemonics to put an i386 into user mode is likely to be very standard; the default code was probably provided free of charge by Intel years and years ago, and probably found it's way into every i386 memory-protected OS written!).
The same seems true of traps.c. The main.c lines are just some includes and some static declarations.
I also did some casual Google searches to see if any of the alleged infringing lines of code showed up anywhere. In all the cases I checked, the lines show only in Linux kernels, and not anywhere else. If this code did appear elsewhere then it isn't immediately obvious where it came from.
So I really don't think that SCO has much of a leg to stand on here
IBM Develops some technology for OS/2
IBM adds it to AIX
SCO (claiming to own copyrights to Un*x) says anything derivative of Unix (AIX in this case) becomes their IP
SCO Sues IBM for copyright infringement
IBM demonstrates this technology existed prior and was given to both operatings systems as an add-on
SCO loses
A feeling of having made the same mistake before: Deja Foobar
SCO seem to be saying IBM knowingly contributed code to Linux knowing it would help linux succeed over traditional Unix flavours...
So let me get this right IBM gave code from their expensive product to a free product...
Hmmm.... why???
SCO should be made to demonstrate what motivation IBM had in acting this way -- other than IBM knowing they could sell services better than OS's...
--- Users are like bacteria -> Each one causing a thousand tiny crises until the host finally gives up and dies.
the story goes something like this:
1) dave cutler's VAX/VMS team gets bored (funding cut)
2) microsoft hires dave and his team (6 people).
3) they code 200,000 lines per year EACH for 5 years, exclusively in c.
4) paul leach (of apollo and NCS 1.0 which became DCE/RPC fame) recommends DCE/RPC for NT Domain services.
5) bill gates orders from-on-high that NT must have a windows interface.
6) dave's team add a windows subsystem to placate bill: they have to port the win16 subsystem to 32-bit (hence the win32 subsystem).
[7) ibm somehow gets involved: nt also has an OS/2 subsystem. someone gets terribly embarrassed that NT uses ibm's OS/2 "HPFS" and orders that NT must have its own file system (esp. because HPFS doesn't support VAX/VMS security model) hence NTFS.]
8) DEC cottons on to what dave cutler is up to, especially when the VAX/VMS security model's interface turns up pretty much function-for-function in NT, and gets integrated properly into the NT Domain Services.
9) DEC gets paid $50m and mysteriously NT 3.51 gets ported to the DEC Alpha.
that's why NT runs on those lovely RISC processors: it was written in c and so was dead-easy to add other OSes.
not bad doing 2 weeks work and getting paid $50m.
Can someone explain why SCO sends people like Weiss & Dal's little brother to the plate when they're clinging to life? These people are wet farts. I would rather be defended by my grandmother in court, and she thinks she's a horse. Really. They just gave Boies millions & a stake in the company - where the uckfa is he?
And yes, I now this is not 100% on-topic. However I think the disappearance of a key figure is noteworthy (I would argue more so that SCO claiming ownership of IBMs work, as they are here).
Groklaw has already been cited as a source in at least one of IBM's briefs to Judge Wells. :)
So yeah, they're definately getting direct credit, and due
Don't even think about looking for this code in linux-2.4.x. Not a single line of any of the mentionened patches has been merged into linux-2.4.
You can find the code in these patches provided by IBM:
EVMS evms_aix.h
JFS ref/jfs_inode.h
RCU-2.4.1-01
Something remotely similar to the rcu patch was eventually merged into 2.5.43 and into United Linux. The EVMS header is used by the compatibility module for AIX partitions and is also in United Linux but nowhere in an official linux. The jfs inode header is not used anywhere, because it is the OS/2 file and was provided only for reference.
However, they have thereby limited their current claims to these sections. And five beeeellion dollars.
Got time? Spend some of it coding or testing
i have no access to the aix or dynix code, but let's take a look at some of these lines. i don't have the ibm patches and am too lazy to download them, but they are all header files of things you would expect to be exactly the same (like inode structures and such). but looking at the code below (and others i didn't bother to paste), i am hardly impressed with sco's claims.
[ cpu]);
sco claim:
dynix kernel/os/kern_clock.c 2028-2059
linux 2.4.1 arch/i386/kernel/apic.c 25-28, 662-664, 676-684
25-28:
#include <asm/smp.h>
#include <asm/mtrr.h>
#include <asm/mpspec.h>
662-664:
* useful with a profiling multiplier != 1
*/
if (!user)
676-684:
prof_counter[cpu] = prof_multiplier[cpu];
if (prof_counter[cpu] != prof_old_multiplier[cpu]) {
__setup_APIC_LVTT(calibration_result/prof_counter
prof_old_multiplier[cpu] = prof_counter[cpu];
}
#ifdef CONFIG_SMP
update_process_times(user);
#endif
SCOX stock price, meet Mr Floor?
I dunno; SCOX buyers seem to have their heads in an invincible reality-distortion field. The next Slashdot poll should be "What will SCOX close at on Monday?":
* $31.89
* $13.75
* $10.00
* $5.00
* $1.00
* -$666.00
The commenter closest without going over wins 30 karma points.
You say that in jest, but Daniel Lyons of Forbes has been known to quote random people from blogs as sources for his stories.
Honestly, were I his editor, I would have fired him after that. Investigative journalism it was not. This gives me very serious misgivings about trusting anything Forbes says, because I cannot imagine how that story could have slipped past even the most minimal editorial review...
It seemed rather apropos, yet disturbing, that that article was meant to be an attack on the credibility of Groklaw, after PJ of Groklaw chided him for accepting SCO's statements without any apparent research, as he had not done even the most minimal fact-checking.
I would be willing to bet that he is glad that I am not his boss... To anyone from Forbes reading this: I value research more than oppinion. And yes, I do mention your failures to anyone I know who might even think of subscribing.
Worst reporter ever. (Maybe seeing his face on Google image search for that would make his day?)
Is SCO guilty of knowingly makinging false claims (ie. that they were being violated by IBM) in order to prop up their stock price?
Martha Stewart's case is based exactly on this (that she made false claims that she didn't engage in insider trading so that her own company's stock, Martha Stewart, wouldn't go down).
Wow, its amazing to see what total crap people can spout off and get modded up
If you are trying to argue to a judge that your code does not contain any stolen code, then SCO points out the alleged lines of stolen code and then as a result, you go changing it, you will have some serious explaining to do to the judge.
I don't mean to be arrogant here, but I'm a lawyer and you do not appear to be, so please don't go around giving legal advice to people who might think its correct.
Don't I remember some company lending a multi-processor system to Alan Cox a few years back so that he could work on scaling SMP code?
Ah yes, I think it was a company called "Caldera" who did that (http://www.linux.org.uk/SMP/title.html). You know, they're called "SCO" now...
If he explores all forms and substances Straight homeward to their symbol-essences; He shall not die.
Don't forget Daniel Lyons' (Forbes Magazine) big bold prediction (from here):
SCO Group will settle its lawsuit against IBM. Both sides will declare victory. The Linux community will turn on IBM.
The only surprise here is that someone (Forbes) pays Dan to write this stuff (apparently the New York Times's fantasy journalism team was all filled up). Dan goes on to point out that Linux and VoIP technologies are no different than Pets.com and other marketing fluff:
technologies like Linux and voiceover-IP still involves this crazy notion that companies can make money by giving things away.
This is like Dan calling electricity a "crazy notion" because he listened to some fly-by-night business scheme involving electrical current. Looks like Forbes is got some fat to trim.
*scoove*