Slashdot Mirror
SCO Claims Linux Lifted ELF
fymidos writes "SCO has finally spoken. According to this linuxworld article, they claim that linux illegally uses the ELF binary format, the JFS filesystem, the init code and some more 'copyrighted Unix header and interfaces'. Finally SCO makes its move. The JFS part was expected of course, but according to the article, as far as the ELF format is concerned 'the Tool Interface Standard Committee (TISC) came up with a ELF 1.2 standard' and 'granted users a "non-exclusive, worldwide, royalty-free license" to the stuff'. Oh, and of course 'both Novell and the old SCO - as well as Microsoft, IBM and Intel - were on the committee'."
I'm wondering how this accusation would effect linux compatibility mode in FreeBSD.
-SD
: [i]The JFS part was expected of course[/i] Am I right when I guess JFS means Journaling File System? Why would it "of course" be mentioned by SCO? Could someone give us its history?
Well, knowing what I know about SCO (only what I read here and on groklaw), I'm willing to believe that there is substantial similarity. However, I'd bet dollars to donuts that the code came from Linux (or more properly, IBM, perfectly within its rights) into SCO, and not the other way 'round.
Their lawyers have fudged so many facts, fumbled so many easy misses, that I wouldn't doubt this for a moment. "Please stay the RHAT case, because it will be decided by the IBM case," and "please dismiss the IBM counter-claim; it will be decided by the RHAT case."
-paul
Pistol caliber is like religion: everyone has their favourite, and theirs is the only right choice.
Ahh, that's funny, I've been been keeping an eye on SCOX on my Yahoo stock watchlist, and I did notice they've been going down the past few days. Even down 3.6% today on an otherwise up market.
So I was expecting some "good news" from them.
Come on SCO - as part of my operating systems course in college we loaded ELF binaries (which we also had to create) and RAN them. They have got to be stretching a long way on that one - that is for sure. Not to mention many other OSs (such as the BSDs) can use ELF binaries....
.... boy. Here comes this crap again. What on earth are they sticking in headers to copyright? #define ONE 1;??? Sounds like I might have a case myself.
The JFS claims, that seems like an awful stretch as well, It does make more sense in targetting IBM though as I believe they had heavy involvement in JFS. Honestly I am not nearly as familiar with the ins and outs of it, but unless they are claiming something ridiculous like memcopy() or something...
Which brings us to number three, 'copyrighted unix headers and interfaces'
Now the interfaces, which could perhaps be interpreted as API... there is some chance that could have some fuzzy ground I imagine. But how on earth can the judge/court even take them seriously at this point?
These guys make absolutely no sense. They're flailing their arms in the air and saying "You owe us money!" Let's pick apart each claim:
1. Sontag also says that any entities that ignore SCO's ELF copyrights are infringing."
Hello? Copyrights are for "copy" protection. Not "patent" protection. If you invent something new and fail to patent it, I can re-implement it all I want, copyright or no.
2. SCO also claims "substantial similarity" between the Read-Copy-Update (RCU) routine in Linux 2.6.5 and Linux patches and SCO's copyrighted work, specifically SVR4.2 MP.
Again, do they have a patent? If they don't, then they should STFU.
3. It thinks that Unix SMP 4.2 System V initialization (init) code was copied into Linux 2.6, that there's "substantial similarity" between the user level synchronization (ULS) routines in Linux and Unix, that its Unix System V IPC code was copied into Linux 2.4.20 and that copyrighted Unix header and interfaces were copied into Linux.
Oooo... it's similar! Well, that just cinches it, doesn't it. It's not like anyone would ever re-implement a design they liked. I better go take the wheels off my Chevy. It's too similar to a cart and horse.
4. It also says the journaled file system (JFS) module from later versions of AIX, which SCO believes may derive from the JFS Unix, is in Linux 2.6. - MOG
Just to reiterate, SCO is not reading their own contracts. What IBM develops is IBMs. Plain and simple.
These guys ought to be tried and convicted for barratry, regardless of whether their legal counsel is this stupid or not.
Javascript + Nintendo DSi = DSiCade
They are pulling ELF in to court because they are desperate to avoid a ruling on IBMs tenth counterclaim scheduled to be heard on Aug 4th.
This may be a hard sell to the court since they were not a party to the original effort and they can't seem to find any of the documents transferring any rights from old SCO to Caldera/TSG.
We'll see if it staves off the inevitable for a while longer at the aug 4th hearing. The best part is that the more deperate they act the more chance the principals will end up in jail when all of this fraud is exposed.
Cisco IOS post 11.2 is ELF. So is nearly every format out there except HPUX. This includes BSDs, embedded systems, so on so fourth.
So they have just got themselves into the aiming calculations of the entire computer industry including the other big Blue, not just IBM.
Anyway, do not see a problem even if they win this one. While I want to puke just at the thought of ECOFF, it is if IIRC (C) intel and HP and all it will take to get linux to use them will be one big rebuild and a rewrite of libdl. That is if Intel and HP do not decide to put the dl for ECOFF into the public domain.
In, btw, this is something on which Cisco can buy them just to shut them up (if everyone agrees to go home and stop the lawsuits).
Baker's Law: Misery no longer loves company. Nowadays it insists on it
http://www.sigsegv.cx/
Since the seller sold the copyright with the explicit understanding that ELF was in the public domain, NewSCO can not claim anything.
If someone "exceeded" their authority it's a matter for the parties involved at the time.
Help fight continental drift.
Gee, we're using the ELF format for ARM embedded firmware development... so I guess we're infringing on SCO's intellectual property too! Oh, and by the way "we" is Intel Corporation...
Readable Description of the ELF standard
Help from SCO on ELF compatability in SCO OpenServer
Summary of a.out, COFF and ELF from a FreeBSD pov
Nice descriptions of various BFFs (binary file format) including ELF
I admit to not knowing much about the GPL, but if it does turn out that the code went from Linux -> SCO, doesn't that mean that the SCO property would be subject to the GPL (what with it being "viral" and all, right?)
Maybe this is why they are doing all this. They've realized that their code is vulnerable, and so to protect it, they have to attack the GPL, lest they go down in flames.
The only thing I hate more than hypocrites are people who hate hypocrites.
I admit to not knowing much about the GPL, but if it does turn out that the code went from Linux -> SCO, doesn't that mean that the SCO property would be subject to the GPL (what with it being "viral" and all, right?)
The GPL is very clear that if someone tries to use the GPL on software that they have no right to use the GPL on, then they can't use the GPL for it. This means, if someone GPL's someone elses code, it is considered NOT GPLed at all, and it never was, regardless of any claim. In order to license any software, you must first possess the RIGHT to do so.
Only the legal owner of any code can choose the license his/her software is under. No one can take that away by illegally releasing it under a different license. It is no different than if I stole the source for Windows 98, then "released in under the GPL". That doesn't make it GPL because I never had the rights to do so.
Tequila: It's not just for breakfast anymore!
IANAL, but I don't belive you can copyright a format- you could patent it, but that's not what's alleged here. Note that the FAT implementation that Microsoft is waving around like a big stick is due to their patent, not a copyright claim.
Unless the code to implement ELF is identical in Linux, I don't see anything like a case here. And if SCOX had a patent on ELF, we'd know it by now.
What a strange bird is the pelican, his beak can hold more than his belly can.
So, I'm reading the shaky claims from SCO, and I'm remembering that Linux used a.out format.
I think that the worst case scenario is that SCO get a tempory injunction blocking the use of ELF (nevermind that no judge would actually agree that to not issue on would cause irreprable harm, given that it's 9 years old). Therefore, there is some merit in ensureing that it is possible to build a modern Linux system on a.out format binaries. Additionally, it torpeados thier claim that ELF is the 'magic pixie dust that they stole to make Linux work'.
This sounds like a case for a source based distribution. Gentoo, being the most used source based distro sounds like a place that might be able to do that. Alas, I'm not familiar enough with Gentoo to comment - so can anyone who is give a definiative answear on whether Gentoo can be used to give a system without ELF binaries. I undertand that there will likely be a bootstrap step (I remember the a.out -> elf shift in the first place, after all).
It's not just (not even, I should say) fear of SCO, but there is significant meaning in being able to switch executable formats - that's a lot of flexability for the future.
So, a.out Gentoo, anyone?
I suggest we make a fund with a goal of around $80 million. Then with the funds buy SCO, fire the executives, desolve the company, sell the assets, settle the debts, then change the source to GPL.
The exact price if you buy all the stocks is less then $70 million. However, the debts may be a cause for concern.
http://ir.sco.com/stock.cfm
So any takers?
Oh, another little point that the article messes up, and that is somewhat amusing.
:-)
ELF is not "similar to Microsoft DLLs", or that's badly worded. It's similar to Microsoft PE format.
ELF is derived from COFF. It was mostly a rewrite of COFF with some bad assumptions and nn-portabilities fixed. It so happens that Microsoft's PE (PE-COFF) format is also derived from COFF and is very similar to ELF. If the format was somehow "protected" (which wouldn't be via copyright as pointed out elsewhere), then Microsoft are also guilty of copying. If SCO really own the copyrights to Unix (they don't), and if copyright applied here (it doesn't), then MS are in the same boat with everybody else. Lucky for them that SCO don't have a leg to stand on
Tim
The problem we see here and throughout the SCO case is that copyright was never designed with software in mind. The nature of software licensing is such that there's frequent cases of derrived works from different sources, which is rarely the case in books.
Actually I feel this is a good analogy; whoever wrote the first "Whodunnit" story (probably that dude who wrote the Bible) could claim "Murder She Wrote", "Columbo" and "Hong Kong Phoowee" are all derivative works.
Stephen King (deceased at 54?) would be dragging Dean Koontz through the courts by his lank, greasy hair, and there'd be no Dallas ripoffs... hmm, actually this is making a compelling argument...
Code, Hardware, stuff like that.
Best of all it affects SCO. If ELF support has to be dropped from gcc, what will they use for a compiler?
TIS, UNIX International, and ELF 1.1
That's a handy reference document for SCO, the version 1.2 of the specification. Too bad it wasn't the first published version:
Google for "pfmt11.pdf"; here is the most interesting and damning excerpt:
-
"Some of the major reasons for selecting this format are the public nature of the specification and the fact that the PLSIG and ABICC standardization committees can enhance its formats."
-
This version came out of USL and UNIX International, who are jointly credited with the creation of the ELF 1.1 standard. Even if USL could argue rights, the current SCO can't. This standard, along with the DWARF standard, TET, ETET, and the last draft of Specification 1170 (the original Single UNIX Specification) were published on the UNIX International FTP server. UNIX International was a legal agent for USL at the time of publication.
If you want to check into anything, check into the contractual agreements between USL and UI with regard to what rights UI did or did not have. You will find that they had full rights to publish the standard on behalf of their member USL.
In the interests of full disclosure, I was a Novell/USG ("UNIX Systems Group") employee at the time. Novell/USG was comprised of the NWU ("NetWare for UNIX"), NUC ("NetWare UNIX Client"), and the former USL. I'm one of the people who rescued the public content of the UNIX International FTP server and found it a new home at various other corporate sites when UNIX International effectively disolved in 1994. One of the documents rescued was this very document.
-- Terry
What happened with TISC in 1995 is irrelevant. As the Usenet post from 1993 shows, the ELF support in Linux does not come from any code that TISC released into the public domain in 1995 with the support of SCO. It was a clean room implementation based on documented interfaces from "SYSTEM V RELEASE 4 Programmer's Guide: >ANSI C and Programming Support Tools" (ISBN 0-13-933706-7).
There is no copyright issue here SCO, but have fun taking the US Navy to court over it.
The fact that this had to happen is scary enough.
u me nts/dojinterrogationmemo20020801.pdf
The torture policy from August 2001 can be found here:
http://www.washingtonpost.com/wp-srv/nation/doc
LedgerSMB: Open source Accounting/ERP
My thoughts also. The microkernel crowd in general seems to agree that Mach is not a good microkernel. (L4 is much better.) And writing a monolithic kernel on top of a microkernel, rather than taking advantage of the full microkernel flexibility...I don't see the point.
Can anyone explain to me why Apple did this with MkLinux, why NeXT did this with NeXTSTEP, and why Apple continued it with OS X? I love OS X's user interface and blend of Unix and old-school Apple stuff...but this one decision, it just seems so dumb.
just because something isn't patented doesn't mean it has no protection under copyright
Umm, what? Copyrights and patents are orthogonal. They cover two completely different things.
According to the USPTO, if something (like software) can be copyrighted, then it *cannot* be patented - although aspects of it (like an algorithm) might qualify for a patent, the code itself would not. The general rule is that patents protect ideas, copyright protects expressions or implementations.
This is essentially what SCO is claiming with their copyright arguments, as I understand it.
In order to determine what SCO is claiming, one must ask which day of the week it is.
which is not to say those formats cannot be patented, just that it's not what is going on here
Nobody is claiming that - they're claiming that the format *is not* patented, and that (by definition) you *cannot* copyright a format, so therefore SCO's claims are baseless.
Unless they can show that the kernel's ELF loader contains code copyrighted by them, they have no case.
Just as the share price is slipping below they release this. I cant imagine this being a coincidence. In court they have said nothing about this at all. They have been ordered to provide proof of what specifically linux infringes on, not only the IBM stuff, and has brought nothing but "the dog ate my homework". This despite two court orders to bring forward specifications like this.
The board is surely behind this and i suspect the SCO laywers are furious. Now they have to explain to the court why this was so damn hard to bring forward in a court with two orders hanging around their neck while it was this easy to tell the media.
Share price up tomorrow and then as theese unfounded allegations get shot down it will fall again.
HTTP/1.1 400
What is the *real* motivation?
I used to think that way too, before "I made $10 million bucks and all I got was this crummy t-shirt."
Seriously. Built an early ISP. Sold it for a ton of money. Raised an additional $20 million to chase UUNET and PSI's coat-tails. Had a neutral broker in the whole deal steal the money and the company. His attorneys stalled everyone out long enough to get money to offshore banks. We all won the suits, and never saw a dime. Company was long dead by the time the court system got done.
What I didn't understand at the time was that people like this know how to make money going down. Study what they call penny stock companies (ala bulletin board stocks/pink sheets). You can run one of these that does absolutely nothing, never files SEC documents, and make a nice income. Here's the middle item of the underwear gnomes strategy:
1. Incorporate a dummy privately trading C corporation which you and your friends own.
2. Use your public bulletin board company to acquire the dummy company. Claim the dummy company has some mystical super secret technology, or better yet, a patent claim against someone like Amazon.com, Priceline, Ebay, etc. When you buy it, it not only loads you and your friends up with more shares of stock (you buy it with the public company stock, and if you're smart, you do much of this w/ registered stock you can immediately dump), but it puts a bunch of press releases out about the event and gets half of the stock buying market going in a feeding frenzy buying your public stock.
3. Dump a bunch of your newly accumulated shares. Ride the stock price down to $0.001.
Repeat steps 1-3 over and over. Oh, when creditors come, make sure you drive the price to zero again (this makes it worthless for them to come after you and even keeps them from pushing involuntary bankruptcy which would take over your company). Claim you were an innocent investor who lost millions too. If pressed on where financial documents are, make sure you had hired some stupid kids and promoted them to CEO by promosing them multi million dollar paychecks. Have them sign all the IRS and state revenue department documents, but don't ever let them actually have control of the bank accounts. This sends the IRS and SEC after them and will stick them with tax obligations for decades. During this lying low time, acquire a few of your dummy companies and load back up on stock.
Sound far fetched? It happens every single day. The SEC told us it could not do anything about these and could only worry about the big Worldcom sized matters because they were "underfunded and overworked." The rest of the story is that it takes financial and political clout to get the SEC to investigate. Common folks don't matter.
So is there any wonder that the names behind SCO include power hitters from both parties and much of Utah's power base? Just another way of taking money from the lower classes...
I was working with CBIS (Cincinnati Bell Information Systems) in the early '80s when Sequent S81 was running Dynix/ptx with SVR and BSD interface flavours.
The technical sales reps were very hot on RCU and how it worked with NUMA to scale the system. 32 way SMP of 386's might not seem much now, but it was significant back then.
Their plans were to have multiple OS flavours under Dynix, relying on RCU to provide scalability to AIX, SunOS/Solaris, and other operating systems that would sit on top of the monolithic ptx core.
From day one, RCU and NUMA were marketted by Sequent as seperate products targetted at many operating systems and vendors, not as just a part of Dynix.
I'm tired of watching SCO's insanity. The industry did not develop in a vaccuum -- there are hundreds if not thousands of us who are witness to how the systems were really developed, despite some IP vulture corp's claims otherwise.
It's a disgrace that SCO's legal team and IP trolls haven't been arrested for fraud and locked up by now. They have no legal case, they never did, and they never will. It's time for them to be spanked and sent to sit in a jail-cell corner for 20 years each.
I do not fail; I succeed at finding out what does not work.
Oh dear, my new kitchen which I am slowly installing, might just resemble some small corner of Darl's kitchen, the washing machine, tumble drier and dishwasher might be arranged in the same order for example...... I think I need a good schyster!
I am still waiting for MS to let SCO completely self destruct, then buy their "IP" at a bargain
I've been wondering about this as well, though really see the SCO deal as more of a trial balloon and not the actual mother of all battles for open source.
It really doesn't seem like Microsoft's approach (remember how many billions they have and the pride that comes with the wealth and dominance). A proxy battle is more interesting as a means of testing the waters and seeing how the open source world responds. Where are they resiliant, and where are they weak? Toss pathetic little SCO and their delusionary CEO at them (much in the manner Hillary Clinton is 'endorsing' Kerry/Edwards). How often have they acquired a weak product and threw it in the competitive pool prematurely, fully knowing it wouldn't swim - only to come back in version 3 or 4 and destroy the competition (after having learned all they needed).
In this respect, the more interesting game is the one not being played - the game of omissions, unmoved pieces, etc. Where is F/OSS weak and scrambling? You can bet Microsoft is designating many millions to carefully observing this battle - their dominance depends on it.
Microsoft does not play to lose. They won't let SCO hit the real IP issues - rather have them throw items like ELF, JFS and header similarities that you know will fail. This is textbook Art of War.
*scoove*
You can't talk about the ELF format without mentioning this gem. You learn a lot about the format when you push it to its limits.
/me stands back and waits for Sony to crush SCO like a bullant that's just bitten them.
That's not a dream. When this all comes to light, if SCO can't actually act on any of its myriad threats, the remaining shareholders will sue the company because it will be deader than disco.
They aren't. In none of their claims in any of their lawsuits do they ever accuse anyone of infringing their copyrights by putting their stuff into Linux.
According to the article, these ELF charges are part of two new documents filed in the IBM case.
One can only conclude that this is a SCO shenanigan aimed at blocking IBM's motion for summary judgement regarding noninfringment of SCO's copyrights. IANAL, but Autozone's lawyers (and the folks over at Groklaw) seem to think that if IBM gets the declaration, SCO will instantly lose its case against Autozone (and probably DaimlerChrysler and Red Hat, too). So SCO really needs to delay that judgement as long as possible.
On the other hand, if SCO was serious about this ELF stuff, you'd think they should have brought it up a year ago when it filed against IBM, or at some point fairly early on in discovery.
SCO's real problem is that the lawsuits are starting to reach the point where SCO loses. AutoZone got a stay. Damlier-Chrysler has moved for summary judgement, and they have a procedural hearing tomorrow at which SCO probably won't look good. Early next month, the big one, IBM's motion for summary judgement on the copyright claims, goes before the judge.
In the IBM case, SCO is frantically filing motions in all directions, desperately trying to stall a ruling on summary judgement or to raise issues that will convince the judge not to dismiss the copyright claims. If they lose on that one, it's the beginning of the end.