Slashdot Mirror
Did SCO 'Borrow' Linux Code?
An Anonymous Reader writes "Apparently someone inside SCO has stated that SCO(actually Caldera) copied Linux code into System V. They did it to build what they now market as Linux Kernel Personality - the ability to run Linux software on their Unix. Now, the open source community(of course they don't mention who) is jumping on this, because they didn't return the changes to the OS community or give the community credit. Of course, SCO says it's a misunderstanding and, get this 'SCO also never used any of the Linux kernel code.'"
Yes ... And why can't it be the other way around?
... It hasn't been tampered with ... Try to convince and explain that to a judge and jury ;)
That the code inside Linux isn't from SCO but the code inside SCO is from Linux?
How to prove that? If SCO wasn't using any kind of CVS - and
I would really like to see documentation on this "offending code" to prove that it was theirs.
... no answers.
Otherwise, how do we really know who added it and when? What if it was pre Sys V code? What if the code came from Linux in the first place?
I have heard that the comments were the same, but who made the comments? Is there a name? Does he/she work for SCO?
Lot's of questions
Can the open source community sue? If so - can't we just give them what they wanted to give Linus?
-- Ilya
This is the last gurgle of a dying corperation. SCO produces, for all purposes, nothing. Why didn't they pursue these claims when they manufactured Caladra, a form of Linux? because they know they're groundless. There's no way that the Linux kernel isn't in public domain. To try to prevent distribution based on 80 lines of code of a program with thousands of lines is ridiculous. Not only that, but any 80 lines of code could be easily removed, thus SCO can't reveal it. When this goes to trial and they're forced to reveal the lines of code and the lines are promptly removed, no one will care and SCO will fade into obscurity. Good Riddance.
Since when has this country used intellectual elite as a pejorative term?
They may have released it as Caldera Linux, but this instance is in SCO Unix which hasn't had it's code released.
If you belive the merits of this claim....If you belive them, now's your chance to cash out from the pot those greedy bastards at SCO have on the table...it's something like $9.50 / share now! ....grab it... .....SELL SHORT.....first thing when the market opens....
Take all of that money before the greedy bastards grab it off the table!!!!....Monetary damages are the only thing the SCO mgmt. and the speculators funding this operation respect. Take their money before they figure out that the knife cuts both ways...
It's also interesting to note just how easily SCO found their code in Linux; you'd think it'd be too difficult to find such things unless you were looking...or if you already knew they were there...
SCO sues IBM because they use Linux code that SCO says has SCO code in it. Novell says Unix isn't SCO's. SCO says Unix is theirs to exploit. Linux geeks angered by SCO says SCO copied Linux code into its products, ... etc etc ...
:
...
When I was, oh what?, five years old, I remember that kind of talk in the courtyard at school during recess
- Hey, Johnny stole my yellow marble
- No I DID NOT !
- YES YOU DID !
- It's not your marble anyway, it was mine, I just told you to borrow it, I didn't give it to you
- I'll tell my Mom Bruce stole Robert's marble, and you'll be GROUNDED !
- I DID NOT !
- YES YOU DID !
Replace one of these kids by SCO, another by Novell, a third by IBM, a fourth by the Linux community, the one who tells Mom by Microsoft, the courtyard by the computer industry and Mom by the DOJ and there you have it.
*sigh*
"A door is what a dog is perpetually on the wrong side of" - Ogden Nash
No duh. How many people have been saying this? AND how many people are ignoring the fact that SCO group themselves released the "offending" code under Caldera? Everyone but IBM it seems, since they think this lawsuit is frivolous.
The moderator who modded you "insightful" was on crack, because you completely failed to read the article. Timothy is suggesting that SCO may have copied (presumably GPLed) code from Linux into their proprietary Unix[tm]. If true, the repurcussions could be, erm, quite interesting.
As lore would have it, the original USL suit against BSD and Berkely University broke up on the rocks for a similar reason.
Have you got your LWN subscription yet?
Here's what I've been wondering: if linux has code that's the same as code in SCO unix, why assume that linux copied it from SCO and not vice-versa? You can't just point to code that's the same and say "Aha! linux copied!".
It's not about seeing the same code in both places, it's about establishing which was developed first. You can't look at just the current version of either linux or SCO - you have to look at the change history of the common code. In one version, the code should show some evolution over time - across RCS versions, or across versions of kernel releases. In the copied version, a whole bunch of code will have appeared "Poof!" all at once. You can't just look at the surface - you have to look beneath the surface, into the code's history.
Of course, there is the possibility - I consider it unlikely - that large chunks of code appeared in both places all at once. This will mean that the code was developed over time external to whichever version of linux or SCO unix had it first, then copied in as part of a major rev... but somebody, some developer somewhere, will have interim versions, notes, design docs. Code doesn't just spring from the head of Zeus - it evolves, and whoever developed it will have to be found to prove its origin.
Except you're wrong. SCO has evidence. This is a rumor.
To be precise, SCO is spreading rumours that they have evidence.
Have you got your LWN subscription yet?
>A source close to SCO, who spoke on condition of anonymity, told eWEEK that parts of the Linux kernel code were copied into the Unix System V source tree by former or current SCO employees.
Someone outside, but _potentially_ credible said they copied. Not an employee. This is only a _possibility_. Perhaps the source assumed too much or maybe I'm wrong.
The way to prove the code?
Compile it.
SCO *has* shown something to the public, to whit: binaries. Compile the original programs that contain suspected infringing code, compare it against the binaries they shipped, and if the match, the are able to place their code in time.
CVS logs can be altered. The code that is compiled cannot, nor can they change the binaries that have already shipped.
It's easy, it's fast, and it is accurate.
That should place the SCO code in time within about a six-month period. If the Linux code pre-dates this period by a significant amount, the infringing code came from the Linux kernel, and SCO is a smoking crater.
If the Linux code came later, then it is IBM who is curb-stomped, followed by a full-frontal assault on Linux itself.
Microsoft is to software what Budweiser is to beer.
"Did SCO Violate the GPL?"
No. If they had published Linux code as proprietary software, they have violated the copyright law.
Karma: Positive (probably because of superiour intellect)
However, I'd be very surprised if Microsoft used anything from Linux, considering it's actually legal and therefore far more tempting to use something from the BSDs, and there are not many features Linux has but the BSDs lack.
O'DOYLE RULES!
My blog
No, they had to reimplement it because while you can make BSD code GPL you can't take it the other way, at least without copyright holders permission. They were free, of course, to look at the linux code while doing it, making it a relatively easy task. Probably if you looked there are sizeable chunks of identical code and comments there too. I bet the header files are a real treasure trove for those.
Identical chunks of code and comments do not prove copying or copyright infringement. It takes more, in a case like this, because there are plenty of perfectly legitimate reasons for it to occur. To determine if something illegal happened, whether we're talking about Caldera copying from Linux or vice versa, you've got to do a much more fine-grained analysis than just counting lines that match.
=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Friends don't let friends enable ecmascript.
Maybe someone else has standing as well (were those intimidation letters legal?) but I suspect the interesting stuff won't happen until IBM's lawyers start speaking up. They're suspiciously quiet at the moment.
===== Murphy's Law is recursive. =====
and SCO did copy Linux code and as a result will be held accountable for that we should all thank RMS for insisting on what he believes in.
:)
Well we should thank him anyway and often
- Back off man. I am a scientist
Challenging the GPL in court is on a hiding to nothing.
Nerd: Derogatory term typically directed at anybody with a lower Slashdot ID than you.
"During that project [Linux Kernel Personality for SCO] we often came across sections of code that looked very similar, in fact we wondered why even variable names were identical. It looked very much like both codes had the same origin, but that was good as the implementation of 95 percent of all Linux system calls on the Unix kernel turned out to be literally 'one-liners'," the source said.
I don't quite understand this. If the guy was working on the LKP project and they discovered similarity between SCO UNIX and Linux during that work, then SCO did not copy that code as part of the LKP project (although they may have copied it before). Or did he join the LKP project late and alleges that other people on the same project copied the code before he joined? Or is he saying that SCO had copied Linux source code for other reasons and they were just discovering that fact during the LKP project at SCO?
I think you are underestimating IBM's interest in Linux. IBM is doing well largely due to Linux. What do customers want? Freedom from lockin. The only way to give that to them is by selling services around non-proprietary software. I don't think IBM would settle this for a dollar and let SCO take control of Linux. It doesn't make sense.
Besides, I have never heard of any public domain, BSD, or GPL copyright being slurped up into a propriety product because of the proprietary product's "overriding interest", or what have you.
Timeline wise the origins of the SCO code predates the Linux code. thus if you discover two things are similar one should probably assume that the older one is more original.
Only if one assumes that both codebases grow and gain functionality at the same rate. As ESR showed in his whitepaper, Linux has grown and gained functionality substantially faster than SCO, and is ahead in many areas.
why would SCO want to alter any existing functionality in their code by changing it to linux code?
Perhaps the Linux implementation was more featureful. Perhaps it was more efficient. Perhaps the copied linux code delivered some functionality that was not present at all in the UnixWare kernel. Indeed, ESR shows in his whitepaper that Linux is substantiall more advanced than UnixWare in many areas, and has been for some time.
would this not be a recipe for destroyng backward compatibility?
Not as a rule. Programmers add features to all kinds of software without destroying backward-compatibility.
At a minumim any cribbed code would have to be patched for backward compatiblity. the only cases where this would not be true is if the new code was for completely novel functionality that would not have any influence on or need to call existing SCO code. not likely.
I don't follow at all. Why is that unlikely?
The more likely explanation is that they either both got the code from a source older than SCO's code (e.g. BSD) or Linux copied SCOs code as alledged.
Actually, if we know that the code was copied either from Linux to UnixWare or vice versa (i.e., we know that it didn't come from BSD), then the far likelier explanation is that SCO copied Linux code. Rememeber that UnixWare is proprietary; an SCO coder could easily copy a chunk of the Linux kernel and no one would ever know. However, anyone who copied in the other direction would be putting himself at risk, since the Linux kernel is open source, and the evidence of the copying would therefore be plainly visible.
TheFrood
If you say "I'll probably get modded down for this..." then I will mod you down.
That's what happens with future versions, yes. However past versions that have this code are now, whether they like it or not, GPLed, too. This means that anyone that currently has a version of UnixWare with Linux support can demand the source code for it.
...and...
So... when they distributed their UNIX with the LKPM included (their "work") and that contained GPLed code, they accepted the terms of the GPL. But they have not distributed, or offered to distributed, the source to their (now GPLed, since the accepted the terms) "work".
This means that either they violated the GPL after agreeing to it. The owners of the copied code will band together and sue them for $2G, I hope, and settle for costs plus distribution of the full source of UnixWare 7 distributed as per the licence agreement SCO acceded to. Just to labour the point, they have already distributed derivative code, so halting distribution does not undo their requirement to distribute full source.
Do I need to make it simpler for you?
Got time? Spend some of it coding or testing
So they claim, but install the sources from an old Cladera linux distro. Grep for Caldera and see the code they contributed. Infact they even say it's GPLed in there comments. Is it cut and pasted from Unix? I don't know I don't have the source to Unix (I don't know anyone who does? do you?). Is it the same lines that they're claiming people stole from them? I don't know that either, IANAA
Fast forward to the present and you have SCO suing IBM about getting chocolate in their peanut butter. SCO would have a much better chance of winning if they sued themselves.
Nope. All this means is that SCO is guilty of a license violation.
IANAL - but I'm betting that you would have to take them to court and convince a judge that the violation was intentional ("Hey, let's use this GPL code!"), willful ("Yah. We can just ignore the license."), and pervasive ("Sure, why not - the VP of development and legal already said that's fine.") Otherwise, SCO can just claim that the inclusion of the GPL'd code was a "misunderstanding" between a long-gone developer and a long-gone manager, neither of whom had the authority to make this kind of decision on behalf of the company.
Even if you got past that hurdle, I expect that you'd have to explicitly request that the code for the past versions be placed under the GPL as part of the settlement, and SCO would probably value the code so highly ("One billion dollars, your honor!") that any order to GPL the code would give SCO a good chance to get that aspect of the ruling either thrown out or reduced on appeal.
"Great men are not always wise: neither do the aged understand judgement." Job 32:9
Another possibility, at least for some of the code in question, is that someone at Caldera authored code that was contributed to Linux under GPL and was placed into Unixware to be released under Unixware's more restrictive licensing.
Dual licensing is legal as long as the copyright holder agrees to it. If the code in question originiated with Caldera/SCO (and thus Caldera/SCO own the copyright) and was provided to Linux as GPL code and inserted into Unixware as non-free code, there's no lawsuit in either direction. So, if we find that the code came into Linux from Caldera and/or SCO, it means we're in the clear, and so are they.
Another possibility is that a third party (eg. IBM) authored the code and effectively dual-licensed it--licensed it to SCO for proprietary use, and licensed it to Linux under GPL. Again, that's most likely fine.
The only way SCO might have a lawsuit is if the code originated in UNIX/Unixware, and was contributed into Linux by someone other than the copyright holder.
--JoeProgram Intellivision!
If the file containing the code you stole has prominent text in it referring you to the GPL (as recommended by the FSF), then you are deemed to have been responsible for reading the GPL before using the code.
Got time? Spend some of it coding or testing
The GPL doesn't count because violating it essentially voids it leaving you with copyright law for distributing terms. (You can't, but if you do, you have to pay thousands per copy.)
If SCO had not been trying to screw over the Linux community, this would probably be, small cash settlement, an apology and stop using the code, as it is the copyright holders are probably not going to want to be nice.
Work bio at MMWD
Apparently someone on the inside knows a lot
and really hates the way SCO is trampling on
the community.
Isn't it nice how they can look out our hard work
and no one is allowed to look over their shoulders?
Kudos to this brave soul who took a stand but I suggest
watching your back.
To mod you down somehow each and every time I can.
We're not children, and most of us know that crashing SCO's site intentionally does nothing but demonstrate that we too can be cocks.
I am not saying this is what has happened but it is at least a possibility. Big companies can be incredibly incompetent at times.
As lore would have it, the original USL suit against BSD and Berkely University broke up on the rocks for a similar reason.
As lore would have it, the proper spelling of Berkeley is B-E-R-K-E-L-E-Y, and the proper usage is "University of California, Berkeley," being that Berkeley is the University of California; the other UC schools (UCLA, UCSC, et al) are merely extensions of UC Berkeley, which was founded in 1868.
So no, it's not spelled "Berkly," Berkely," Berkley," or any combination of the three, and it most certainly has no connection to the Berklee College of Music.
I'm amazed that any self-respecting geek can misspell "Berkeley", given the advances made there. Where the hell do you think Berkelium and Californium were discovered? If it weren't for Berkeley, which runs LANL and LBNL, the DOD would be up shit creek, and GWB wouldn't have any of those "nuke-u-ler" weapons he likes to talk so much about. For the love of god, the guy who won a Nobel prize for inventing the frickin LASER is a professor there.
Without Berkeley, there'd be no BSD; it's the Berkeley Software Distribution. It's in the name of the operating system. If you can't even properly spell the name of the operating system to which you're referring, why even bother to make any comment at all?
The theory of relativity doesn't work right in Arkansas.
"...[SCO] basically re-implemented the Linux kernel with functions available in the Unix kernel to build what is now known as the Linux Kernel Personality (LKP) in SCO Unix."
and
"The LKP is a feature that allows users to run standard Linux applications along with standard Unix applications on a single system using the UnixWare kernel."
and you said
"so I think it is safe to say that these functions were in UnixWare first"
So you are saying UnixWare was compatable with Linux binaries before Linux came to being? Just asking.
You are right, thank you. I'm trying to add every one of them to my "freak" list, but it's not easy, when I'm still being insulted by some new ones... Thank you for good word, though.
Karma: Positive (probably because of superiour intellect)