In light of the SCO allegations, SGI’s code comparison was done during September using the Comparator software created by open source advocate Eric Raymond, as well as some other internally developed tools, according to SGI. It compared source code from the Unix System V release 4.1 software that SGI has licensed from SCO with a version of the Linux kernel released this June, SGI said.
SGI Compares Linux, Unix Source Code
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
You have to read through so much to get to the point: what was the result of the comparison?! The result, it seems, is about 200 lines of code, most or all of which has already been put into the public domain.
that SCO is pulling this crap on SGI. So what that XFS is in Linux. SGI legally put it there. The worst part of this is that SGI offered to pull the code to make amends to SCO and SCO says “Sorry, you have already done irrevocable amounts of damage to us by putting XFS in Linux”. I hope the SGI fights all out on this one. Sure, SGI is suffering finacially but I really hope they stick it to SCO on this one.
A more important point, that still seems to be lost on the blind and the zealous, is that the kernel is NOT lily white and pure. If anyone reads the article with an open mind they can clearly see that SGI states that there IS System V code in the Linux kernel. It goes on to state that some of it IS line for line copied. Still further it says that there may be more but, they either didn’t find it or weren’t interested in it because they were not responsible or liable for those parts. This is a far cry from the “there is NO SCO code in the kernel” that some people keep screaming.
I and most reasonable people will grant that the actual quantity of copied code is indeed small and it may have been an inadvertent mistake that put such infringing code into the Linux kernel but, it can no longer be denied that such infringing code exists. The fact is that there is infringing code of one kind or another in the Linux kernel and effort msut be made to remove it. But, denying its existence or arguing that the small amount of copied code makes the case irrelavent is as irresponsible and insane as SCO’s rantings about $699 licenses. Everyday, it seems more likely that SCO does have some sort of case, even if it isn;t the multi-billion dollar case they are dreaming about.
You automatically assume that someone outside SCO put Unix code into Linux but that has to date never been proven. I want SCO to prove that someone inside their cozy headquarters did not do it.
By the time SCO gets torn apart in court for damages, their wouldn’t be much left to collect. It really is very flagrant for them to even state that XFS is a “derivative” work of UNIX System V code. Every OS out there is influenced by UNIX code that has been out in the wild for decades.
If you look at this logically, it was just a bunch of grumpy, disillusioned shareholders losing money with little chance of profitibility in the near future. They knew that they could bring up this bogus wrap against IBM (and now SGI) and gratuitously inflate their stock value. No one even half sane would make a statement like this if they planned on their company surviving its outcome. We can thank Wall Street for eating it up like morons. This is a case where a dying star imploded on itself and suddenly went supernova.
SGI is trying to do the right thing here, and some people (who obviously want to remain nameless,) are making a big stink about it.
I will reiterate what I have said before: SCO has become a valueless, evil company which seems to be unwilling to do anything good. To carp about a few mistakenly-copied, or public domain code, when the people at SGI are making a sincere effort to insure things are on the level, is about as low as I’ve seen anyone go.
If you are going to rant about the misbehavior (or perceived misbehavior) of Linux developers, who are sincerely against doing these kinds of wrongs, then at least put your name on your statements.
With regard to SCO… they acquired ownership of the aforementioned UNIX code… while, at the SAME time, they were also producing, shipping, and selling their own Distribution of linux that obviously included the licensed code.
If a company owns code and distributes it through GNU Linux… Regardless of who put it in there, they are basically condoning and validating its legal distribution and openness.
If there was a problem, it should’ve been addressed more promptly and/or they should’ve eliminated their Linux products as soon as they acquired the UNIX license.
If you read the article, it says that the code in question was removed in 2.4.22. Also, if you read a previous article on the topic, the code was things like atoi() functions, that are rather basic and easy to reimplement, as well as trivial in general. Also, it appears that most of the 200 lines was already in the public domain, thus there can be no claim on it.
If anyone reads the article with an open mind they can clearly see that SGI states that there IS System V code in the Linux kernel.
…
This is a far cry from the “there is NO SCO code in the kernel” that some people keep screaming.
Let’s get that straight. There is no SCO code in the kernel.
If there’s any System V code, it was most likely implemented by AT&T or Berkley and legaly distributed under the BSD license. The only problem is that some companies removed BSD copyright notices when copying the code.
Anonymous wrote:
“I and most reasonable people will grant that the actual quantity of copied code is indeed small and it may have been an inadvertent mistake that put such infringing code into the Linux kernel but, it can no longer be denied that such infringing code exists.”
The above statement is incorrect. It has not been established that any code in Linux infringes on any SCO code. There are several possibilities that would negate any claimed infringement even in the case of identical code fragments being found in Linux and whatever SCO owns:
1. SCO (or its predecessors, such as Caldera) contributed the code to Linux.
2. The code came from BSD, and is therefore free of any SCO claims, regardless of how it came to be in Linux and SCO code.
3. The SCO code identical to Unix 32V is not entitled to copyright protection at all because Unix 32V code entered the public domain due to the actions of AT&T. See:
http://cm.bell-labs.com/cm/cs/who/dmr/bsdi/930303.ruling.txt
4. SCO has licensed the code under the GPL (when it had a Linux distribution) and has lost the right to claim infringement
In addition, on the derivative works claims, items 2. and 4. above negate such claims, unless the creator of an alleged derivative work had a separate contractual agreement with SCO that provided otherwise. SCO would still have to establish that what they claim are derivative works are in fact derivative works and that their contracts did not permit the creator of the derivative works from dealing with such derivative works independently of the original work. In other words, I think SCO’s derivative works claims are bogus.
Now if SCO (or anyone) can identify copied code and show that it is not BSD based or otherwise in the public domain, then there might be an infringement and then the question of remedy would be raised.
Regards,
Mark Wilson
Read the article carefully:
“All together, these three small code fragments comprised no more than 200 lines (of code),” wrote Altmaier. “It appears that most or all of the System V fragments we found had previously been placed in the public domain, meaning it is very doubtful that the SCO Group has any proprietary claim to these code fragments,” he added.
200 lines of code is a trivial amount of code. An experienced programmer could reimplement all 200 lines in a matter of hours. Further, most of that code was already in the public domain. The SysV code is so old and so widely distributed (in so many formats, from presentations and class lectures to books and source listings), it would be highly likely that many source trees have code derived from it. However, this amount is so minimal, its tantamount to perfect purity. Its like a surgeon’s scapel. You can never remove 100% of germs from the instrument, but you can clean it to a degree to which you can treat it as being perfectly pure.
Read your own posts and listen to yourselves.
However, this amount is so minimal, its tantamount to perfect purity.
I’m sorry but a minimal amount of copied code is an amount of copied code. It is NOT tantamount to perfect purity. It is copied code!
it says that the code in question was removed in 2.4.22. Also, if you read a previous article on the topic, the code was things like atoi() functions, that are rather basic and easy to reimplement, as well as trivial in general. Also, it appears that most of the 200 lines was already in the public domain, thus there can be no claim on it.
The fact that the code was removed does not mean that it never existed. Also, the fact that the code is trivial to reimplement does not decrease the value of the code or alter the fact that it was copied.
This is precisely what I was referring to in the first post. People seem to be so bloinded by their zealotry that they refuse to see the facts or think for themselves. The facts are that SGI has CONFIRMED that there indeed is or was System V code that was inappropriately COPIED into the Linux kernel. The code is being removed and that is a good thing. But, that does not mean that the kernel is as pure as you would like it to be and indeed continue to claim. I am not favoring SCO, nor am I attacking the kernel developers. I am attacking the posters that allow their zealotry to blind them of the fact that there is indeed copied System V code in the Linux kernel. These people sound like the Iraqi Information Minister ranting that there are no US troops in Iraq.
gee, do you still not get it? SCO was and is dead. It’s right now about how to get as much money out of the corpse as possible. and it works fine if they play the stockmarket right.
This is one of those legal situations where a simple leap of logic proves one side to be false, but said leap is inadmissible in court.
If those 200 lines of code belonged to SCO, and SCO attempted to charge license fees based on that, then the code would have been reimplemented overnight, and SCO would not have had any code to charge for. I find it puzzling that SCO can say with a straight face that they believe that charging licensing fees for tiny snippets of code that any hacker with two hours of free time can replace would have been a winning business model all this time.
Honestly, if they tried charging for Linux, we can say with 100% certainty that they wouldn’t have made a cent. But that’s still a “what if,” and can’t be truly proven.
Isn’t this frustrating?
Anonymous wrote:
“The facts are that SGI has CONFIRMED that there indeed is or was System V code that was inappropriately COPIED into the Linux kernel.”
Not exactly. SGI said the copied code was public domain. SGI did not say that it was inappropriately copied.
Regards,
Mark Wilson
You are correct. The use of the word inappropriately was my own and not SGI’s. My use of the word was inappropriate.
I think It is not the removal of code what SCO is after, it is the “compensaton” for the money they “lost” (or didn’t made) because other OSes aledgedly took advantange of “their” code.
If a court of people with unsufficient knowledge on the matter, and/or personal interests on it, decides that this is true, that would certanly mean a big lot of finacial troubles for companies that are investing real money in linux development.
You don’t seem to understand what people here are saying. SysV code is *everywhere*. There are tons of articles, papers, books, classes, etc, that expose programmers to SysV code. Under those conditons, it cannot be expected that a code base will be 100% free of SysV code, and the state at which a code-base would be like that is just an intellectual curiosity. According to SGI, the code that they did find is mostly in the public domain, which makes perfect sense because so much of this code is out there. At that point, it isn’t illegitimately copied code, but perfectly legitimately copied code. What tiny fraction that there is left (if there is any) that is not in the public domain is probably too little to form the basis for a valid copyright infringement case. Remember, copyright law does not consider a fragment of the work of minor length to be an infringement. For a code-base comprising millions of lines, a dozen lines here or there might very well be too little to file a suit over.
My point was that, in that case, removing the offending code would be of no use.
people are giving any credence to SCO’s argument at all. Their case is baseless and they’re facing a mounting number of countersuits. Does anyone forsee any outcome to this besides SCO declaring bankrupcy after they are countersued into oblivion?
Folks, this is very good news, but not unexpected. It is time to forget about SCO and push on to make Linux the dominant operating system in the world.
. . .and place the UNIX source code into public domain so that everyone can live happily ever-after?!?!
Seems to me the morons at SCO, and yes, morons, do not have a business plan. According to some sources at Automation Access, the Canopy Group has a good history of instigating lawsuits rather than executing a well thought out business plan.
I am praying, if not hoping, that IBM and RH squash the SCO Group into the ground. Its a shame that the workers at SCO don’t know what their own bosses are doing. TTFN.
Pretty doubtful. From what I read, the source code of the UNIX (remember, UNIX is a standard, not a single OS) we all refer to here is a horrible mess of copyrights.
“Seems to me the morons at SCO, and yes, morons, do not have a business plan. According to some sources at Automation Access, the Canopy Group has a good history of instigating lawsuits rather than executing a well thought out business plan”
…their business plan is to bug any company worth anything, until one of them buys them out to shut them up.
It’s kind of like that annoying fly…buzzing around everybody’s head at the picnic…. sooner or later, it’ll fly into that bright, blue *BUG ZAPPER*… (in this case, Big Blue!)
…*HOT*
So what if a tiny amount of ‘stolen’ code is present. We are talking hundreds of lines out of millions of lines of code. That is like saying a single sentence in a book is identical to a single sentence in the Encyclopedia Britannica.
people are giving any credence to SCO’s argument at all. <<
I am equally amazed. Scox has been caught in dozens of outright lies. Yet, the stock price keeps going up.
>>Does anyone forsee any outcome to this besides SCO declaring bankrupcy after they are countersued into oblivion? <<
Scox transfers all assets, unix license etc, to another canopy company. Then scox goes out of business.
I could type “/ignore SCO_Lawsuits_Issues” and be perfectly happy.
Can anyone truely see a way in which this will matter to any of us who are not massively invested in any of these companies? I have a massive “so what” reaction brewing inside and it wouldn’t take much more of this stuff will make it bubble to the surface in a huge belch of “Shaddup already!”
And yet you care enough to post in the comment sections of a SCO story? 🙂
if only I had some stock… and a company… and some shady IP… and a community of zealots to outrange.
aaaaah.
actually, just getting a job again would be real nice.
sigh.
SCO claiming that because SGI releases code THEY (SGI) produced, some how, through some act of god, it becomes the property of SCO. That would be like me writing a driver for my brand new widget and SCO then claiming that because I have created a driver for UnixWare, it instantly becomes the property of SCO.
If I could go back in time, the one thing I would stop is Caldera purchasing SCO. Had SCO remained what it was, and old battle axe in the Unix industry, it would still have customers and probably would have jumped back into profitability.
Instead of SCO bashing the crap out of Linux, they realised that customers wanted the flexibility of opensource, so they started making opensource tools available for their operating systems.
Also, one has to stress, when Caldera bought SCO, they only lost a small amount of money and were actuall getting contract. Instead of being bought by a stronger company, they got bought by a company leaking money like water through a sieve.
Oh, and here is SCO CEO wooing people to buy SCO before it goes belly up:
http://www.computerworld.com/softwaretopics/os/linux/story/0,10801,…
Nothing more than the Canopy Group pumping up the stock price so that they can selvage some of the lost money they pumped into Caldera over the years.
I’ll just add to that the fact that there is code contributed by IBM to Linux from AIX, however, the issue isn’t copying but contractual. SCO claims that IBM’s contract doesn’t allow them to do that, IBM has replied saying that they’re well within their rights.
Lets wait until the contract is disclosed so that “canyon pissers” such as I will find out whether it goes down or whether a big gust of wind will come and wet me with great embarrassment.
IMHO, with IBM’s legal team, I am sure they went through that contract with 400 fine tooth combs to ensure that there is no one got-cha which SCO could exploit if a legal challenge is launched by them (SCO).
As confirmed by my trusty book “The Magic Garden Explained : The Internals of UNIX (R) System V Release 4 : An Open Standard”, at least the AT&T Unix File System – not the 7th Ed. file system however, aka the s5 file system – was the BSD Fast File System, ffs, slightly mutated via Sun Microsoystems and then renamed by AT&T as the Unix File System or ufs.
So much for any of SCO’s proprietary interest in that part of UNIX System V Release 4!
It’s a pity the source hasn’t been opened – I’d like to see just how much of it is actually BSD in disguise.
well SVR4 you got to remember was a product of SUN and AT&T attempt to merge the two lines of UNIX (BSD and System V) so there’s quite abit of BSD code in there that would have come via SunOS
Point well taken! I would guess that if you can find it in one of the mirror sites of The Unix Heritage Society:
http://www.tuhs.org/
either as part and parcel of the original AT&T code –
http://minnie.tuhs.org/PUPS/
or as part of the BSD lineage,
http://www.tuhs.org/archive_sites.html
then I would argue that it is for all intents and purposes unencumbered and/or under the BSD/MIT-style license.
The problem with their SCOnesses is that SVRx is by now archaic – I mean, the book I refer to is nice to have – and indeed, in the light of the accusations of the said SCOnesses, I’m glad I got it when I did – it’s a very useful book – and AT&T lost the research plot way back in the late eighties, leaving all the research – for all practial intents and purposes – up to the University of California @ Berkeley, and letting Thompson, Kernighan, Ritchie, etc, very much out on a loose lease. I don’t think there’s a scrap of original AT&T Research Unix (8th to 10th Edition) in SVRx. SVR4 represents the best they could do with proprietary code in 1989.
That was 4.3BSD and SVR3.x. Now we’ve got FreeBSD, NetBSD, OpenBSD, Dragonfly BSD, and a host of others along the side, all considerably more advanced than SVR4.
SVR4 has nothing to bring to the table this time.
“If I could go back in time, the one thing I would stop is Caldera purchasing SCO.”
You must REALLY care about this. Think that you could use your one-shot time machine to kill infant Hitler or bring a vaccine to plague ridden Medieval Europe, saving millions of lives and probably altering history enough to preclude the purchase of SCO by Caldera anyways. Or you could go back to kill and eat a dinosaur, which would probably do nothing beneficial to humanity except confuse a few anthropologists, but would be really cool.
Seriously, I am inclined to side against SCO in this debate, but I have some nagging doubts. Nobody questions that SCO’s motives are twisted and anything they say is a lie. This is probably true – in my experience this is what companies that seek money do – although they probably aren’t as bad as what some people seem to believe. In any rate, the skepticism about SCO’s claims is valid; no one should believe anything without proof. But why accept SGI’s word at face value just because it is convenient to your argument? They are also a profit-seeking company, with the same motives to lie as SCO. Anything they say should be treated with the same skepticism.
Related to this, they identified the shared code with (among other things) a program written by Eric Raymond. He is not an disinterested party in this situation, and even if he were I would need to see some proof that this program does what it supposed to do and does it well. I am not familiar with it and the proof I want may very well exist, in which case please instruct me to its whereabouts.
Along these lines I think there are some arguments regarding the SCO debate that do not deserve to be heard. Blindly asserting what you want to be true does not make it so; it is better to assume that anything you can’t prove is false. Here are some statements from today’s thread that probably deserve the heave-ho:
1.) “There is no SCO code in the kernel.”
An appropriate replacement would be: “It has not been established that any code in Linux infringes on any SCO code.” (acceptable version from post by Mark Wilson)
2.) “It appears that most of the 200 lines was already in the public domain, thus there can be no claim on it.”
Even with the softener “appears,” it is implied that there is some superficial evidence that this is true. But what we really have is no more than SGI’s word.
3.) “I will reiterate what I have said before: SCO has become a valueless, evil company which seems to be unwilling to do anything good. To carp about a few mistakenly-copied, or public domain code, when the people at SGI are making a sincere effort to insure things are on the level, is about as low as I’ve seen anyone go.” There are many many problems with this one. Better just chuck the whole thing.
It would strengthen their arguments immensely if Linux supporters looked at them with the same mistrust as they do SCO’s.
>>It would strengthen their arguments immensely if Linux supporters looked at them with the same mistrust as they do SCO’s.<<
Scox has earned our mistrust. Lie after lie, scam after scam. A few examples:
– scox claimed they would cancle ibm’s unix license – they didn’t.
– scox claimed they would audit aix customers – nope.
– scox claimed they would provide proof of infringing code – nope.
– scox said they would send invoices to linux users – they flip-flopped.
– scox claimed they “bought” vultus – another lie.
– scox claimed a team of mit rocket-scientists discovered all this infringing code – yet another lie.
– scox claimed to own c++ and all the BSDs – a lie.
Any easier solution would be for SCO to disclose the contracts they have between SCO and SGI, SCO and IBM, and for SCO to point out *EXACTLY* what parts SGI and IBM have failed to live up to.
The code that has been found, RCU IS a line by line copy from AIX, *BUT* the crucial thing is that SCO is claiming that *EVEN THOUGH* it was written by IBM, it suddenly becomes SCO IP. That doesn’t make any sense. It would be like Microsoft claiming that all drivers made for Windows is now owned by Microsoft.
As I said previously, IBM would have gone through the code and contract with 400 fine tooth combs to find any loop wholes which SCO could exploit.
was planned. The XFS file system wasn’t magically transplanted into Linux. An announcement was made, changes both to support Linux’s way of doing things and to satisfy
legal requirements were performed and numerous beta versions were released.
SCO/Caldera had plenty of time to object and, in fact, they were part of the effort to improve Linux; this hindsight approach is tantamount to guiding a truck that’s backing up, waving to the driver that’s it’s safe to proceed, watching him crush a pedestrian and then telling the cops “Yup, I saw the whole thing, he just flattened that poor fella like he didn’t care”
was planned. The XFS file system wasn’t magically transplanted into Linux. An announcement was made, changes both to support Linux’s way of doing things and to satisfy
legal requirements were performed and numerous beta versions were released.
SCO/Caldera had plenty of time to object and, in fact, they were part of the effort to improve Linux; this hindsight approach is tantamount to guiding a truck that’s backing up, waving to the driver that’s it’s safe to proceed, watching him crush a pedestrian and then telling the cops “Yup, I saw the whole thing, he just flattened that poor fella like he didn’t care”
RCU was incoporated right back in the early 2.4.0-test series, almost 2 1/2 to 3 years ago, SCO had their chance and they blew it. They can’t honstly expect the court to believe that for 3 YEARS they have been shipping Linux and suddenly, out of the blue they finally realise that the code they have been WORKING ON FOR 3 YEARS has now magically infrindged on their IP.
IMHO, they had 3 years to do something about it. They screwed in, they’re running out of money and now they’re expecting sorrow from the courts and stock market. Let SCO die, the IP BSD’ed and patents disolved. The sooner UNIX is owned by no one, the better.
I would like to point out the old Unix vs. BSD case. In that case 3 whole files were found to be infringing but it was found to be through innocent copying not malicious copying. When a judge favored BSD in a peliminary rulling AT&T settled with the Univeristy. So 200 lines of code that slipped in is not infringment. Keeping them in once they were found would be. So taking them out does clear SGI. The bigger part of the case (and the only one being tried right now) is the contractual agreements between IBM and SCO, not copyright infringment. SCO claims ownership on code they did not write, and that is malicious.
Im not a programmer but it seems to me that ppl are ignoring the possibility that Unix stole from Linux. Linux code can be found for free and if your a big software company and you have this code out there just ripe for the taking why not take it? it would be hard to prove since not many ppl have access to Unix code. From the way SCO is acting i wouldn’t put it beyond them to do somthing like this.