Basically what the guy wants is nationalization of all telcos, so that your taxes pay for everything.
I don't think he would ask "Do we still need telco's" if he saw a role for them being run by the government. You miss the point entirely. What this guy wants is commoditization of bandwith equipment to the point that oridinary citizens can beome self sufficent. He wants the exact opposite of what you suggest: he wants maximally decentralized private ownership.
Here the contract is between SCO as a distributor of the linux kernel and the owner of that copyright, Linus Torvalds et al (call this collective kernel.org).
If you accept SCO's version of the facts, the the GPL on the unix-linux mix is rescinded because of a "mutual mistake". SCO then bears the burden of mitigating damage to kernel.org. They are clearly failing to do so when they refuse to simply inform kernel.org of the specifics of which code is infringing.
This refusal to mitigate damage harms kernel.org and also harms the entire Linux market, including IBM who profits from its Linux support and consulting services. I would argue that whatever marginal harm is caused by the misappropriation of a few hundred lines of Unixware code is dwarfed by the attempt by SCO to keep the million-line Linux whole in a state of unfixable dubious licencing.
I just reviewed the licences between AT&T and IBM that are posted on SCO's site as Exhibit A and Exhibit B.
In section 3.03 of exhibit B it clearly states that "AT&T" may revoke the licence for non-compliance. Moreover paragraph 4 of the first page contains a standard "no alterations unless signed in writing" clause. I see nothing that allows AT&T to sell this termination right without IBM's approval.
Article 78 is a New York state law, and would not apply in Utah.
Actually, if you read the original contracts between IBM and AT&T posted on SCO's site as Exhibit A and B, the forum selection clauses provide that the state law of New York applies. See for example, section 7.13 of Exhibit A and section 6.05 of Exhibit B.
SCO isn't obligated to tell anyone other than IBM what code it is basing its suit on;
I'm not sure that is correct. SCO here is arguing that the licence on the linux kernel is void because of a "mutual mistake" (the inclusion of non-GPL'd code). When a contract is breached, there is a duty to mitigate damage to the other side. It seems rather straightforward here what that duty implies: SCO should tell kernel.org what the wrongly incorporated code is so that it can be removed. Obviously, attaching an NDA would preclude an open source release. It is not acceptable for SCO to say: our code got mixed with yours, so you are screwed because we won't tell you about it.
Read Exhibit A, Exhibit B, and Exhibit C, in particular.
SCO can revoke the license for breach of contract. The procedure for doing this is not at all clear.
I'm reading exhibit A. Section 6.03 (page 4) authorizes "AT&T" to "terminate all rights granted". Section 4 is a standard boilerplate to the effect of 'no modifications to this contract unless agreed to in writing by both parties'. In particular, I don't see any language that allows AT&T to sell the termination right over IBM to SCO without IBM's approval, nor do I see anything indicating IBM approved of this.
If parties A and B enter into a contract and party A wrongly accuses B of breaching the contract and then takes actions that itself breach the contract, then B would countersue A for breach of contract and ask not just for the reinstatement of the contract but for damages caused by the *SCO's* breach. Moreover, when a contract is breached, there is a duty to mitigate damage to the other side.
SCO is really treading dangerously here. Even if IBM did misappropriate their trade secret, it is very unlikely that a court would order IBM to cease and desist its entire AIX operations. The worst case would be a large monetary fine on IBM and an order to remove the offending code from Linux. For SCO to attempt to impose a death penalty before a court issues a judgement is, even if SCO is completely correct in what they allege, is an extreme move that probably harms their legal case more than it helps.
If SCO eventually loses on their trade secret claim, then IBM will have a very serious counterclaim for unfair competition, breach of contract, defamation, etc...
Don't buy into the "software isn't sold, it's licenced" mentality just because the big corps claim that is the way it works. The reality is that courts are divided on the issue and it has never been settled definitively. It is precicely for this reason that UCITA was persued so agressively, but remember that only two states passed it and some even passed UCITA non-enforcement laws.
I believe that eventually the law will conclude that you need a licence in order to make a copy, but that once made, that copy is sold normally through the transactions that move it through retail distribution. The recent case Adobe vs Softman adopted this view (to the minimal extent needed to decide the case).
n a nutshell, they aren't really taking it seriously - at least not in their initial response to SCO's allegations...
No. Rest assured, they are taking it very seriously. IBM is WAY more disciplined than SCO and their army of lawyers will, at a time of their choosing, descend upon SCO with great fury. I would expect countersuits for unfair competition and probably copyright infringement of IBM's GPL'd kernel contributions, and I wouldn't be surprised if they lauch them in multiple jurisdictions simultaneously.
On the merits, you have to remember that this is ultimately a trade secret claim in which the plaintiff ran a business republishing the secret under terms that on their face abandon trade secret claims. What SCO alleges IBM did is no worse that what SCO itself must admit it did itself.
All copyright law contains questions of whether an enforcable agreement was entered into by the parties. That does not make it "contract law". The difference is jurisdiction and remedy. It would be settled in Federal, not state court, for example, and would rely heavily on Title 17 of the US Code.
The case I'm imaginging would proceed like this: Linus would file in Federal court for copyright infringement, claiming SCO has reproduced and distributed his work without a licence. He would be arguing that SCO did not comply with the GPL on portions of the code he authored and thus that SCO's reproduction and distribution of these violates 17 USC 106 (1) and (3). The problem for SCO is that they have to fully comply with ALL enforcable aspects of the licence. Their only hope is to try to argue that the parts they did not comply with are not enforcable and are severable. The critical parts that SCO would have to attack are section 2b, 4 and 6. None of those seem assailable in the slightest.
There are two very easy ways SCO could have avoided liability. (1) Not distribute Linux in the first place (2) Distribute Linux and give up all these rediculous claims.
I just don't agree with your analysis of Section 7. In fact, I think Section 7 is essentually vacuous because that is the law anyway. It states the obvious: if you distribute a mix of code derived from several sources, you must be able to simultaneously comply with the distribution licences from all of those sources or you must not distribute.
From Linus perspective, it restates the obvious: you may not distribute his code mixed with anything with an incompatible licence, regardless of who did the mixing. In most cases, mass reproduction and distribution is consider much worse than just creating the infringing item, because it generates large amounts of revenue. SCO has no rights to distribute his parts of the code outside of what Linus grants them and he never granted them the right to distribute his stuff with proprietary stuff (from any source) and his licence even explictly forbids it, though this is unneeded. Congress requires an explicit grant with the burden of proof on the distributor.
That's because it is copyright law, not contract law.
The unreasonable part is that they are simultaneously claiming that the code they received was not properly licenced, so that the terms therein don't apply to them, at the same time they are pretending this isn't a problem for them as distributors. Which is it? If the code was improperly licenced, they commited copyright infringement, if the code was properly licenced, they gave up their patent and trade secret claims as part of exercising that licence.
In the sense that anyone can sue anyone for anything with or without grounds, he can be sued. Up until the announcement that they might sue Linus (which was the reaction under discussion), he was probably underestimating the incredible stupidity and desperation of SCO.
You have this exactly right. It was the right approach until that point, too.
At the same time, one might ask why doesn't he round up his legal staff and throw some of his billions.....OH! He doesn't have that!
You need to learn the first rule of politics: if you can't afford it, use somebody else's money. Plaintiffs lawyers often work on a contingency basis that doesn't require the Plaintiff to pay anything -- and then they get a cut 30-40% of the winnings. But more importantly, such a suit would be in IBM's interest. It might also be in his employer's interest. Or Red Hat's interest. Or Novell's interest. etc...
How were they to know that their proprietary code was present in, say, a file that they didn't even modify?
It doesn't matter if they knew. Knowledge is not an element of copyright infringement. In fact, most cases of copyright infringement that go to trial do not involve wilfull infringement.
They certainly knew that they were under an obligation to give up any further ristrictions on what they were shipping. If they didn't care to look deeper, that was their choice. Negligence is never a defense.
The fact that they did eventually know dismantles any argument they try to make. They only cared when it was their IP that was being harmed, but they were negligent on their duty to assure that they didn't ship contrary to the licence.
They eventually proved that they could in fact identify the mixed code (I'm assuming they have the proof -- otherwise their screwed). Why can't they do that before they ship, instead of after?
SCO redistributed that product, with no reasonable way of knowing that they were distributing their own code.
No, they have a duty to the copyright holders of all rights that they licence to assure they abide by the terms they agreed to. As a practical matter, this product is open source, so all they have to do is examine what they ship. If they failed to examine the contents of what they shipped, they chose to blindly accept the contents and if those contents are infringing, then their neglicence to their duty to respect the licence terms is no defense.
But even you would agree that since they began preparing this case, that at some point they did know. At that point, their violation of the GPL became wilfull, since they continued to ship it. Do you disagree?
The fact that they did discover the problem only proves that an argument that they had "no reasonable way of knowing" what they were shipping is untenable. All they had to do to protect their own interest -- all they had a duty to do to protect Linus's interest -- was exactly what they did do, just BEFORE they shipped.
If a judge determines that it was not possible for SCO to uphold certain aspects of the contract no matter what, and that SCO could not have known this, he/she may rule that certain aspects of the contract cannot be enforced.
Is SCO really going to suggest that reading the code to the product they are shipping is impossible? Since when is "I didn't read what I was agreeing to" a defense? Why can't they use diff to examine changes, like the rest of the world does? If SCO can't do this, why do they think IBM or anyone else can? If somebody embedded DeCSS or a Metallica song in the code, do you think Jack Valenti or Hilary Rosen would accept "we can't be expected to know what we are distributing" as an excuse?
After all, IBM is (allegedly) the party at fault in both cases.
My argument is that whether or not IBM committed infringement against SCO (and it would be against Linux as well), that SCO also committed separate infringement against Linux by trying to enforce additional restrictions to elements of what they shipped mixed with the real deal GPL'd kernel code. SCO has no right to say "Oh great, this licence is screwed -- let's ship it anyway".
But it is the GPL that exposes you to potentially being non-compliant, even when you've made all reasonable efforts to comply.
No different from any other copyright licence in this regard. You better get it right. They have NOT made a reasonable effort to comply with the GPL. They want Linus's code, but they don't want to abandon rights to things they ship with it.
If someone breaks into your company and steals proprietary code, all they have to do is insert it into a GPLed software package that you also contribute to and distribute. At that point, you're screwed.
I think you are getting at a fair question: if SCO legitimately owned proprietary code and discovered that IBM had mixed with GPL'd code, what should they have done?
The answer is very simple: they should have immediately sent IBM and the Linux community a C&D letter stating what parts of the code were proprietary. They should have refused to ship that code themselves under the GPL.
Does this place a burden on them to actually understand everything they ship? Yes, it does, but only if they want to keep their proprietary stuff cleanly separated from their GPL stuff.
Any company that relies on proprietary code and also works on GPLed code could be at risk.
Not quite. Only if the company wants to ship two such code bases that are candidates for mixing. Consider Corel from before: there was little risk that WorkPerfect code and linux code would intermix.
Furthermore, it would pretty much confirm Microsoft's earlier ramblings about the GPL putting a company's IP at risk.
They were overdoing it. The risk is only there if you mix code and distribute the result. Most USERS of GPL code aren't going to distribute modifications of it to others, and if they do, they'll have looked at the diff. Even if you modify GPL code, your extensions don't become GPL until you distribute them.
Consider Oracle. They released a clustered file system under the GPL. What proprietary rights have they lost -- just the ones contained in what they GPL'd. All they have to do is make sure they know exactly what's in the code they ship. If somebody else sends them a patch that adds large chunks of their proprietary RDBMS code in, should they blindly smile and ship it out? Um, no. They should be checking every patch submission they get.
Ok, if you don't believe me, then read the damn law for yourself. 17 USC 504 (c)(2) explicitly contemplates damages not less than $200 when the infringer proves that he was "not aware and had no reason to believe that his or her acts constituted an infringement of copyright"
SCO became the "offender" as soon as they new what was going on. They gave away their own property. They did so intentionally.
I really don't know why you are insulting me, you are claiming slightly less than what I am claiming and acting as if I am disagreeing with you. In fact, they gave away more than just their own property. They also copied and distributed code written by Linus, which is an act of infringement (knowingly or not) unless they satisfy the terms of his licence.
SCO became the "offender" as soon as they took an action that violated the GPL copyright licence on the sections of the kernel legitimately authored by Linux, which they distributed within their product. It's no defense to say IBM infringed first. SCO's acts are separately accountable to any copyright holder whose rights are in play. The first such violation was the attempt to enforce additional restrictions against IBM, (trade secret rights), because the GPL requires SCO,as a distributor, to grant rights to "all third parties" (including right back to IBM) and not to impose "further restrictions" on the GPL licencing of the mixed unix/linux work "as a whole", which is required because it contains GPL parts. SCO would be able to wage their claim against IBM if they hadn't shipped code written by Linus that was mixed with it. But since they did and the GPL is viral, they did more than just give away their own property -- they put it under the GPL and attached their own infringement liability to Linus to any attempt by them to enforce any other rights to it.
When SCO realized that the GPL of the product they shipped was voided (no later than the filing of the lawsuit), they became a **worse** offender by continuing to distribute it, knowing and even claiming openly that the GPL had already been violated. They failed to explain what authorized them to copy and distribute code, partially written by Linus, without any licence at all.
All I have been saying here is that Linus should ask SCO "hey, you reproduced and distributed some of my code -- did you comply with my license which requires anything mixed with my code to be GPL'd and thus released from trade secret and patent protection". They say "No we didn't comply with your licence because our stuff was in there against our will". He says "I'm sorry IBM screwed you, but it doesn't give you any right to screw me -- I've been infringed as soon as you even attempted your claims because my licence requires you to licence everything you distributed with it to the whole world, including IBM. If you blow that off, you have no right at all to distribute my stuff."
Who in god's name would ever deal with GPLed code if there's even the slightest chance that it could expose their business to this sort of legal immolation?
It isn't the GPL that exposes them to legal immolation, it is non-compliance with the GPL that exposes them. And non-compliance with the GPL is also called copyright infringement. It really doesn't matter WHAT licence you are using, if you don't comply with it and base your business on it, then you are exposing yourself to legal immolation.
Please explain to me any reason why Red Hat (say) would change its business if SCO was called to the carpet for copyright infringement here.
There is nothing "good faith" about trying to retain trade secret status for something that you yourself distributed for-profit as part of a GPL'd product, when that licence forbids you from adding any "further restrictions". If SCO want to call "mulligan" then it better pony up all the profits it got along the way.
Once they realized that they had essentially been tricked into distributing their own trade secret, they cannot try to shift "who gets screwed" to Linus. The legitimate parts of the kernel that are truly authored by Linus & freinds are GPL and SCO has an independent duty to Linus, as a distributor to assure that his licence is respected. That licence requires them to not impose additional restrictions, such as trade secret protections.
The bottom line is that the secret is out not just because IBM divulged it, but also because SCO re-divulged it with a statement to "all third parties" (which back-covers IBM) that it was free software. If SCO only did that because they are too stupid or too negligent to inspect their own product carefully, then they deserve what they get. After all, they are PROFITING from shipping this code.
Talk about unclean hands -- they accuse IBM of mishandling this code, when they did the exact same thing themselves in such a way that, on its face, back-authorized IBM to do what it did.
Had SCO not been distributing Linux, they would have a much stronger claim. But they did, and this comes with some very far reaching requirements to abandon various IP protections for all code within what they distribute.
Most plaintiffs attourneys will work for a 30-40% contingency fee. Under 17 section 504 statutory damages for willful, knowing infringement are up to $150,000 per work. There are a whole lot of plaintifs attourneys in the world that will go to work for a good shot at their share of a six digit number. Actual damages include profits from infringement and are likely higher.
I count five separate versions of linux that Caldera currently ships, so each one infringes separately. I'm guessing that only ones that they distribute after they filed would be deemed willful. So we are talking about a case where you don't need to prove much at all to ask for $750,000 in damages.
SCO's allegations themselves breach the GPL, because they are , per section 4, an "attempt" to "otherwise... sublicense or distribute" portions of the code that they MUST distribute solely under the GPL because they contains stuff authored by Linus Torvalds, even if it contains other stuff which they own the "contract" rights to.
It doesn't even matter if SCO is actually correct. The "attempt" to retain other rights that they distributed in a GPL mixture is what violates the GPL. If those rights are non-existent, it is still an "attempt". Read section 4.
It depends whether or not SCO continued to distribute after they'd verified that their proprietary IP had entered the Linux codebase, and for how long.
No!! You can go back in time (up to the statute of limitations) and say "I just discovered that way back in 1998 you distributed your product FOO which included unlicensed portions of my work. Pay up. Oh, and I want your profits from the infringing sales, too." Just ask Dr. Dre who just lost 1.5 million in such a verdict.
Wilfulness is an element of statuory damages, which can be quite large, but actual damages do not depend in any way on state of mind.
If they stopped distribution, or removed the offending code after discovering it, then it could be said that they'd made a good-faith effort to obey the terms of the GPL.
The point here is that if SCO admits it EVER shipped any version of Linux under the GPL, then they have no case for patent infringement or trade secret violations on any protected element within the code they shipped. Conversely, if they claim to retain patent of trade secret rights to something, it cannot be in anything they shipped unless that shipment was unlicensed piracy.
In fact, section 7 of the GPL offers some clarifications on this. It says that any company that knows it cannot legally redistribute must cease distribution.
It has no such knowledge requirement. Section 7 is phrased in absolutist terms: "If you cannot distribute so as to satisfy simultaneously your obligations under this License and any other pertinent obligations, then as a consequence you may not distribute the Program at all."
It does not necessarily hold that a company may be penalized for previous distributions where it acted in good faith, but was undermined by the actions of some third party.
SCO did not act in good faith if it attempts to retain rights to portions of code that are intermixed with GPL'd code, because the work as a whole has to be GPL'd. As for what happens downstream from the infringer, The GPL is very explicit on this in section 4. It states that if you violate the licence, for example if you "attempt" to "sublicence or distribute" under any other terms than the GPL, then your rights are terminated. It goes on to say those who received their licence from you "will not have their licenses terminated so long as such parties remain in full compliance." Thus SCO can't get improperly licenced code from IBM and have any GPL rights unless it's action are 100% consistent with the GPL.
If they weren't the ones to insert it into the kernel, then the GPL on their code is void.
Well, YES. That is the whole point. They are distributing unlicenced code, a large chunk of which is derived from GPL code written by Linus and friends.
The fact that they distribute it is irrellevant.
Alas, you lose during the bonus round. Read your previous statement until you understand. Their licence is VOID. So what gives them the right to distribute something they don't have a license to? Or to make the additional copies in the first place? How would Jack Valenti respond if you operated a business to sell copies of the Matrix that were unlicensed? That is how Linus should respond to SCO. Jack wouldn't care if you could show that what you were distributing was actually some third party mix of the Matrix plus some parts that were copyrighted by you. Why should Linus?
Here, I'll help you:
Q: True or False -- Under standard copyright laws of most nations, such as those in the US, in order to distribute a derivitive work, you must have a valid licence from all of the copyright owners for all of the works from which the work in question is derived.
A: True. For more information, including citations to caselaw start here in section c.
Please answer this: by what authority does SCO distribute code which includes among it code authored by Linux Torvalds? If you get confused, go back an read your first sentence.
Slashdot Mirror
Basically what the guy wants is nationalization of all telcos, so that your taxes pay for everything.
I don't think he would ask "Do we still need telco's" if he saw a role for them being run by the government. You miss the point entirely. What this guy wants is commoditization of bandwith equipment to the point that oridinary citizens can beome self sufficent. He wants the exact opposite of what you suggest: he wants maximally decentralized private ownership.
Well, I think IBM has already filed to move it to federal court, which is very wise.
Here the contract is between SCO as a distributor of the linux kernel and the owner of that copyright, Linus Torvalds et al (call this collective kernel.org).
If you accept SCO's version of the facts, the the GPL on the unix-linux mix is rescinded because of a "mutual mistake". SCO then bears the burden of mitigating damage to kernel.org. They are clearly failing to do so when they refuse to simply inform kernel.org of the specifics of which code is infringing.
This refusal to mitigate damage harms kernel.org and also harms the entire Linux market, including IBM who profits from its Linux support and consulting services. I would argue that whatever marginal harm is caused by the misappropriation of a few hundred lines of Unixware code is dwarfed by the attempt by SCO to keep the million-line Linux whole in a state of unfixable dubious licencing.
I just reviewed the licences between AT&T and IBM that are posted on SCO's site as Exhibit A and Exhibit B.
In section 3.03 of exhibit B it clearly states that "AT&T" may revoke the licence for non-compliance. Moreover paragraph 4 of the first page contains a standard "no alterations unless signed in writing" clause. I see nothing that allows AT&T to sell this termination right without IBM's approval.
Article 78 is a New York state law, and would not apply in Utah.
Actually, if you read the original contracts between IBM and AT&T posted on SCO's site as Exhibit A and B, the forum selection clauses provide that the state law of New York applies. See for example, section 7.13 of Exhibit A and section 6.05 of Exhibit B.
SCO isn't obligated to tell anyone other than IBM what code it is basing its suit on;
I'm not sure that is correct. SCO here is arguing that the licence on the linux kernel is void because of a "mutual mistake" (the inclusion of non-GPL'd code). When a contract is breached, there is a duty to mitigate damage to the other side. It seems rather straightforward here what that duty implies: SCO should tell kernel.org what the wrongly incorporated code is so that it can be removed. Obviously, attaching an NDA would preclude an open source release. It is not acceptable for SCO to say: our code got mixed with yours, so you are screwed because we won't tell you about it.
Read Exhibit A, Exhibit B, and Exhibit C, in particular.
SCO can revoke the license for breach of contract. The procedure for doing this is not at all clear.
I'm reading exhibit A. Section 6.03 (page 4) authorizes "AT&T" to "terminate all rights granted". Section 4 is a standard boilerplate to the effect of 'no modifications to this contract unless agreed to in writing by both parties'. In particular, I don't see any language that allows AT&T to sell the termination right over IBM to SCO without IBM's approval, nor do I see anything indicating IBM approved of this.
If parties A and B enter into a contract and party A wrongly accuses B of breaching the contract and then takes actions that itself breach the contract, then B would countersue A for breach of contract and ask not just for the reinstatement of the contract but for damages caused by the *SCO's* breach. Moreover, when a contract is breached, there is a duty to mitigate damage to the other side.
SCO is really treading dangerously here. Even if IBM did misappropriate their trade secret, it is very unlikely that a court would order IBM to cease and desist its entire AIX operations. The worst case would be a large monetary fine on IBM and an order to remove the offending code from Linux. For SCO to attempt to impose a death penalty before a court issues a judgement is, even if SCO is completely correct in what they allege, is an extreme move that probably harms their legal case more than it helps.
If SCO eventually loses on their trade secret claim, then IBM will have a very serious counterclaim for unfair competition, breach of contract, defamation, etc...
Don't buy into the "software isn't sold, it's licenced" mentality just because the big corps claim that is the way it works. The reality is that courts are divided on the issue and it has never been settled definitively. It is precicely for this reason that UCITA was persued so agressively, but remember that only two states passed it and some even passed UCITA non-enforcement laws.
I believe that eventually the law will conclude that you need a licence in order to make a copy, but that once made, that copy is sold normally through the transactions that move it through retail distribution. The recent case Adobe vs Softman adopted this view (to the minimal extent needed to decide the case).
n a nutshell, they aren't really taking it seriously - at least not in their initial response to SCO's allegations...
No. Rest assured, they are taking it very seriously. IBM is WAY more disciplined than SCO and their army of lawyers will, at a time of their choosing, descend upon SCO with great fury. I would expect countersuits for unfair competition and probably copyright infringement of IBM's GPL'd kernel contributions, and I wouldn't be surprised if they lauch them in multiple jurisdictions simultaneously.
On the merits, you have to remember that this is ultimately a trade secret claim in which the plaintiff ran a business republishing the secret under terms that on their face abandon trade secret claims. What SCO alleges IBM did is no worse that what SCO itself must admit it did itself.
All copyright law contains questions of whether an enforcable agreement was entered into by the parties. That does not make it "contract law". The difference is jurisdiction and remedy. It would be settled in Federal, not state court, for example, and would rely heavily on Title 17 of the US Code.
The case I'm imaginging would proceed like this: Linus would file in Federal court for copyright infringement, claiming SCO has reproduced and distributed his work without a licence. He would be arguing that SCO did not comply with the GPL on portions of the code he authored and thus that SCO's reproduction and distribution of these violates 17 USC 106 (1) and (3). The problem for SCO is that they have to fully comply with ALL enforcable aspects of the licence. Their only hope is to try to argue that the parts they did not comply with are not enforcable and are severable. The critical parts that SCO would have to attack are section 2b, 4 and 6. None of those seem assailable in the slightest.
There are two very easy ways SCO could have avoided liability. (1) Not distribute Linux in the first place (2) Distribute Linux and give up all these rediculous claims.
I just don't agree with your analysis of Section 7. In fact, I think Section 7 is essentually vacuous because that is the law anyway. It states the obvious: if you distribute a mix of code derived from several sources, you must be able to simultaneously comply with the distribution licences from all of those sources or you must not distribute.
From Linus perspective, it restates the obvious: you may not distribute his code mixed with anything with an incompatible licence, regardless of who did the mixing. In most cases, mass reproduction and distribution is consider much worse than just creating the infringing item, because it generates large amounts of revenue. SCO has no rights to distribute his parts of the code outside of what Linus grants them and he never granted them the right to distribute his stuff with proprietary stuff (from any source) and his licence even explictly forbids it, though this is unneeded. Congress requires an explicit grant with the burden of proof on the distributor.
That's because it is copyright law, not contract law.
The unreasonable part is that they are simultaneously claiming that the code they received was not properly licenced, so that the terms therein don't apply to them, at the same time they are pretending this isn't a problem for them as distributors. Which is it? If the code was improperly licenced, they commited copyright infringement, if the code was properly licenced, they gave up their patent and trade secret claims as part of exercising that licence.
In the sense that anyone can sue anyone for anything with or without grounds, he can be sued. Up until the announcement that they might sue Linus (which was the reaction under discussion), he was probably underestimating the incredible stupidity and desperation of SCO.
You have this exactly right. It was the right approach until that point, too.
At the same time, one might ask why doesn't he round up his legal staff and throw some of his billions.....OH! He doesn't have that!
You need to learn the first rule of politics: if you can't afford it, use somebody else's money. Plaintiffs lawyers often work on a contingency basis that doesn't require the Plaintiff to pay anything -- and then they get a cut 30-40% of the winnings. But more importantly, such a suit would be in IBM's interest. It might also be in his employer's interest. Or Red Hat's interest. Or Novell's interest. etc...
Your "life's work" is also being used to bomb civilians in foreign nations.
To bad you aren't among them.
How were they to know that their proprietary code was present in, say, a file that they didn't even modify?
It doesn't matter if they knew. Knowledge is not an element of copyright infringement. In fact, most cases of copyright infringement that go to trial do not involve wilfull infringement.
They certainly knew that they were under an obligation to give up any further ristrictions on what they were shipping. If they didn't care to look deeper, that was their choice. Negligence is never a defense.
The fact that they did eventually know dismantles any argument they try to make. They only cared when it was their IP that was being harmed, but they were negligent on their duty to assure that they didn't ship contrary to the licence.
They eventually proved that they could in fact identify the mixed code (I'm assuming they have the proof -- otherwise their screwed). Why can't they do that before they ship, instead of after?
SCO redistributed that product, with no reasonable way of knowing that they were distributing their own code.
No, they have a duty to the copyright holders of all rights that they licence to assure they abide by the terms they agreed to. As a practical matter, this product is open source, so all they have to do is examine what they ship. If they failed to examine the contents of what they shipped, they chose to blindly accept the contents and if those contents are infringing, then their neglicence to their duty to respect the licence terms is no defense.
But even you would agree that since they began preparing this case, that at some point they did know. At that point, their violation of the GPL became wilfull, since they continued to ship it. Do you disagree?
The fact that they did discover the problem only proves that an argument that they had "no reasonable way of knowing" what they were shipping is untenable. All they had to do to protect their own interest -- all they had a duty to do to protect Linus's interest -- was exactly what they did do, just BEFORE they shipped.
If a judge determines that it was not possible for SCO to uphold certain aspects of the contract no matter what, and that SCO could not have known this, he/she may rule that certain aspects of the contract cannot be enforced.
Is SCO really going to suggest that reading the code to the product they are shipping is impossible? Since when is "I didn't read what I was agreeing to" a defense? Why can't they use diff to examine changes, like the rest of the world does? If SCO can't do this, why do they think IBM or anyone else can? If somebody embedded DeCSS or a Metallica song in the code, do you think Jack Valenti or Hilary Rosen would accept "we can't be expected to know what we are distributing" as an excuse?
After all, IBM is (allegedly) the party at fault in both cases.
My argument is that whether or not IBM committed infringement against SCO (and it would be against Linux as well), that SCO also committed separate infringement against Linux by trying to enforce additional restrictions to elements of what they shipped mixed with the real deal GPL'd kernel code. SCO has no right to say "Oh great, this licence is screwed -- let's ship it anyway".
But it is the GPL that exposes you to potentially being non-compliant, even when you've made all reasonable efforts to comply.
No different from any other copyright licence in this regard. You better get it right. They have NOT made a reasonable effort to comply with the GPL. They want Linus's code, but they don't want to abandon rights to things they ship with it.
If someone breaks into your company and steals proprietary code, all they have to do is insert it into a GPLed software package that you also contribute to and distribute. At that point, you're screwed.
I think you are getting at a fair question: if SCO legitimately owned proprietary code and discovered that IBM had mixed with GPL'd code, what should they have done?
The answer is very simple: they should have immediately sent IBM and the Linux community a C&D letter stating what parts of the code were proprietary. They should have refused to ship that code themselves under the GPL.
Does this place a burden on them to actually understand everything they ship? Yes, it does, but only if they want to keep their proprietary stuff cleanly separated from their GPL stuff.
Any company that relies on proprietary code and also works on GPLed code could be at risk.
Not quite. Only if the company wants to ship two such code bases that are candidates for mixing. Consider Corel from before: there was little risk that WorkPerfect code and linux code would intermix.
Furthermore, it would pretty much confirm Microsoft's earlier ramblings about the GPL putting a company's IP at risk.
They were overdoing it. The risk is only there if you mix code and distribute the result. Most USERS of GPL code aren't going to distribute modifications of it to others, and if they do, they'll have looked at the diff. Even if you modify GPL code, your extensions don't become GPL until you distribute them.
Consider Oracle. They released a clustered file system under the GPL. What proprietary rights have they lost -- just the ones contained in what they GPL'd. All they have to do is make sure they know exactly what's in the code they ship. If somebody else sends them a patch that adds large chunks of their proprietary RDBMS code in, should they blindly smile and ship it out? Um, no. They should be checking every patch submission they get.
You're simply a moron.
Ok, if you don't believe me, then read the damn law for yourself. 17 USC 504 (c)(2) explicitly contemplates damages not less than $200 when the infringer proves that he was "not aware and had no reason to believe that his or her acts constituted an infringement of copyright"
SCO became the "offender" as soon as they new what was going on. They gave away their own property. They did so intentionally.
I really don't know why you are insulting me, you are claiming slightly less than what I am claiming and acting as if I am disagreeing with you. In fact, they gave away more than just their own property. They also copied and distributed code written by Linus, which is an act of infringement (knowingly or not) unless they satisfy the terms of his licence.
SCO became the "offender" as soon as they took an action that violated the GPL copyright licence on the sections of the kernel legitimately authored by Linux, which they distributed within their product. It's no defense to say IBM infringed first. SCO's acts are separately accountable to any copyright holder whose rights are in play. The first such violation was the attempt to enforce additional restrictions against IBM, (trade secret rights), because the GPL requires SCO,as a distributor, to grant rights to "all third parties" (including right back to IBM) and not to impose "further restrictions" on the GPL licencing of the mixed unix/linux work "as a whole", which is required because it contains GPL parts. SCO would be able to wage their claim against IBM if they hadn't shipped code written by Linus that was mixed with it. But since they did and the GPL is viral, they did more than just give away their own property -- they put it under the GPL and attached their own infringement liability to Linus to any attempt by them to enforce any other rights to it.
When SCO realized that the GPL of the product they shipped was voided (no later than the filing of the lawsuit), they became a **worse** offender by continuing to distribute it, knowing and even claiming openly that the GPL had already been violated. They failed to explain what authorized them to copy and distribute code, partially written by Linus, without any licence at all.
All I have been saying here is that Linus should ask SCO "hey, you reproduced and distributed some of my code -- did you comply with my license which requires anything mixed with my code to be GPL'd and thus released from trade secret and patent protection". They say "No we didn't comply with your licence because our stuff was in there against our will". He says "I'm sorry IBM screwed you, but it doesn't give you any right to screw me -- I've been infringed as soon as you even attempted your claims because my licence requires you to licence everything you distributed with it to the whole world, including IBM. If you blow that off, you have no right at all to distribute my stuff."
Who in god's name would ever deal with GPLed code if there's even the slightest chance that it could expose their business to this sort of legal immolation?
It isn't the GPL that exposes them to legal immolation, it is non-compliance with the GPL that exposes them. And non-compliance with the GPL is also called copyright infringement. It really doesn't matter WHAT licence you are using, if you don't comply with it and base your business on it, then you are exposing yourself to legal immolation.
Please explain to me any reason why Red Hat (say) would change its business if SCO was called to the carpet for copyright infringement here.
There is nothing "good faith" about trying to retain trade secret status for something that you yourself distributed for-profit as part of a GPL'd product, when that licence forbids you from adding any "further restrictions". If SCO want to call "mulligan" then it better pony up all the profits it got along the way.
Once they realized that they had essentially been tricked into distributing their own trade secret, they cannot try to shift "who gets screwed" to Linus. The legitimate parts of the kernel that are truly authored by Linus & freinds are GPL and SCO has an independent duty to Linus, as a distributor to assure that his licence is respected. That licence requires them to not impose additional restrictions, such as trade secret protections.
The bottom line is that the secret is out not just because IBM divulged it, but also because SCO re-divulged it with a statement to "all third parties" (which back-covers IBM) that it was free software. If SCO only did that because they are too stupid or too negligent to inspect their own product carefully, then they deserve what they get. After all, they are PROFITING from shipping this code.
Talk about unclean hands -- they accuse IBM of mishandling this code, when they did the exact same thing themselves in such a way that, on its face, back-authorized IBM to do what it did.
Had SCO not been distributing Linux, they would have a much stronger claim. But they did, and this comes with some very far reaching requirements to abandon various IP protections for all code within what they distribute.
Most plaintiffs attourneys will work for a 30-40% contingency fee. Under 17 section 504 statutory damages for willful, knowing infringement are up to $150,000 per work. There are a whole lot of plaintifs attourneys in the world that will go to work for a good shot at their share of a six digit number. Actual damages include profits from infringement and are likely higher.
I count five separate versions of linux that Caldera currently ships, so each one infringes separately. I'm guessing that only ones that they distribute after they filed would be deemed willful. So we are talking about a case where you don't need to prove much at all to ask for $750,000 in damages.
You didn't understand what I said.
... sublicense or distribute" portions of the code that they MUST distribute solely under the GPL because they contains stuff authored by Linus Torvalds, even if it contains other stuff which they own the "contract" rights to.
SCO's allegations themselves breach the GPL, because they are , per section 4, an "attempt" to "otherwise
It doesn't even matter if SCO is actually correct. The "attempt" to retain other rights that they distributed in a GPL mixture is what violates the GPL. If those rights are non-existent, it is still an "attempt". Read section 4.
It depends whether or not SCO continued to distribute after they'd verified that their proprietary IP had entered the Linux codebase, and for how long.
No!! You can go back in time (up to the statute of limitations) and say "I just discovered that way back in 1998 you distributed your product FOO which included unlicensed portions of my work. Pay up. Oh, and I want your profits from the infringing sales, too." Just ask Dr. Dre who just lost 1.5 million in such a verdict.
Wilfulness is an element of statuory damages, which can be quite large, but actual damages do not depend in any way on state of mind.
If they stopped distribution, or removed the offending code after discovering it, then it could be said that they'd made a good-faith effort to obey the terms of the GPL.
The point here is that if SCO admits it EVER shipped any version of Linux under the GPL, then they have no case for patent infringement or trade secret violations on any protected element within the code they shipped. Conversely, if they claim to retain patent of trade secret rights to something, it cannot be in anything they shipped unless that shipment was unlicensed piracy.
In fact, section 7 of the GPL offers some clarifications on this. It says that any company that knows it cannot legally redistribute must cease distribution.
It has no such knowledge requirement. Section 7 is phrased in absolutist terms: "If you cannot distribute so as to satisfy simultaneously your obligations under this License and any other pertinent obligations, then as a consequence you may not distribute the Program at all."
It does not necessarily hold that a company may be penalized for previous distributions where it acted in good faith, but was undermined by the actions of some third party.
SCO did not act in good faith if it attempts to retain rights to portions of code that are intermixed with GPL'd code, because the work as a whole has to be GPL'd. As for what happens downstream from the infringer, The GPL is very explicit on this in section 4. It states that if you violate the licence, for example if you "attempt" to "sublicence or distribute" under any other terms than the GPL, then your rights are terminated. It goes on to say those who received their licence from you "will not have their licenses terminated so long as such parties remain in full compliance." Thus SCO can't get improperly licenced code from IBM and have any GPL rights unless it's action are 100% consistent with the GPL.
Actually, awareness, knowledge, or wilfullness doesn't matter as far as finding infringement has occurred, it only affects damages and remedy.
If they weren't the ones to insert it into the kernel, then the GPL on their code is void.
Well, YES. That is the whole point. They are distributing unlicenced code, a large chunk of which is derived from GPL code written by Linus and friends.
The fact that they distribute it is irrellevant.
Alas, you lose during the bonus round. Read your previous statement until you understand. Their licence is VOID. So what gives them the right to distribute something they don't have a license to? Or to make the additional copies in the first place? How would Jack Valenti respond if you operated a business to sell copies of the Matrix that were unlicensed? That is how Linus should respond to SCO. Jack wouldn't care if you could show that what you were distributing was actually some third party mix of the Matrix plus some parts that were copyrighted by you. Why should Linus?
Here, I'll help you:
Q: True or False -- Under standard copyright laws of most nations, such as those in the US, in order to distribute a derivitive work, you must have a valid licence from all of the copyright owners for all of the works from which the work in question is derived.
A: True. For more information, including citations to caselaw start here in section c.
Please answer this: by what authority does SCO distribute code which includes among it code authored by Linux Torvalds? If you get confused, go back an read your first sentence.