4. You may not copy, modify, sublicense, or distribute the
  Program except as expressly provided under this License.
  Any attempt otherwise to copy, modify, sublicense or
  distribute the Program is void, and will automatically
  terminate your rights under this License...

This is false for GPLv2, sadly, despite people wanting it to be so.

GPLv2 says "6. Each time you redistribute the Program (or any work based on the Program), the recipient automatically receives a license from the original licensor to copy, distribute or modify the Program subject to these terms and conditions. You may not impose any further restrictions on the recipients' exercise of the rights granted herein. You are not responsible for enforcing compliance by third parties to this License."

So every time vmware downloads the kernel, they get a license. Even if the prior time, it was terminated.

(THis is pretty well accepted, but not happily, by most open source lawyers).

GPLv3 is different here.

In particular, take a look at Footnote 4 on that page, which documents state of this situation succinctly.

"Each time you redistribute the Program"

In this sentence "You" would be VMWare, not the authors of Linux.

"the recipient automatically receives a license from the original licensor to copy"

the Linux license is between Linux authors and recipients of VMWare software.

So section 6 does not say Vmware gets a new license each time they get a copy of Linux.

Section 6 just makes sure that even if someone along the chain of distribution of the GPL software fails to comply, not all receiving parties down the path are automatically in failure to comply.

So if redhat redistributes (through their website) to vmware, redhat is you, and vmware is recipient.

You don't get to pin You to be a specific person just because it comes out the way you like it. The license says what it says. (and you can't anyway since this would ruin the rest of the GPL ...)

That's not how it works. "You" is, consistently throughout the GPL, whoever gets the program under the GPL license, and that license is between "You" and the copyright holders on the program (i.e. the licensors).

The person who is "you" can, in another situation, be a licensor or mere redistributor, but that's irrelevant to the legal relationship at stake here (i.e the relationship between the copyright holders and a person who failed to comply with the GPL).

Section 6 is amazingly clear, and this is completely and totally unambiguous.

I don't remember if Eben disagrees (I actually don't think he does, last I remembers), but it's generally accepted that this is the likely outcome.

Sure, it's an interpretation.

[edit] the point is: nowhere as clear as the original comment expresses.

But this is a policy argument, not a legal one, and it will absolutely never win in court over very clear and unambiguous text.

This is, as i said, why GPLv3 was revised here.

I can't even see how you would come to the conclusion you did, given it says "6. Each time you redistribute the Program (or any work based on the Program), the recipient automatically receives a license from the original licensor to copy, distribute or modify the Program subject to these terms and conditions. "

IE it very specifically says they get the rights you claim they don't.

You are reading who is giving who a license wrong.

Redhat, or whoever they download from, is the "you" there. Not the person who license was terminated.

So let's fill in the blanks:

6. Each time Redhat redistribute the Program (or any work based on the Program), the company who violated the license automatically receives a license from the original licensor to copy, distribute or modify the Program subject to these terms and conditions.

All i did was replace the word "you" and "recipient" with the right people. This is clear and unambiguous. You may not like the result. I actually don't like the result. It does not change the legal meaning, or whether you'd win.

But you don't have to take my word for it, ask another open source lawyer, they'll tell you the same thing i just did.

The only lawyers I find who agree with your position (and, BTW, they agree with it for completely different reasons than you state, and I suspect they wouldn't agree with your reading of v2 Section 6) are a few lawyers in Germany.

I thus find myself in the odd situation of having to defend your trolling a bit, because Till Jaeger, Christoph's lawyer in the VMware case, has indeed stated publicly that he believes you can regain a license under GPLv2 by coming into compliance and redownloading in Germany. However, this is likely specific to Germany because no other legal expert I've ever met who has studied this issue has argued it works anywhere else in the world.

This point is therefore salient insofar as this thread is discussing a copyright case in Germany. But, as I wrote in the copyleft.org footnote I referred to earlier, the issue is just an esoteric legal detail except in the case of proprietary relicensing business models. Community-oriented GPL Enforcement Organizations always restore rights anyway once the violator achieves compliance, so the final impact of both interpretations ends up the same in most enforcement actions, unless of course your goal is to abhorrently use strict termination to extort gobs of money, in which case, this detail matters a lot.

It looks like a shitty, evidence free insinuation of dishonesty used as a rhetorical tactic, but that's a big assumption to make.

Even then, a copyright-holder can still permanently terminate your rights if they so choose and notify you of such termination.

That said, (and continuing with my possibly incorrect assumption that the GPL hasn't been litigated) if it does get decided that this remedy can be enforced by the license holders through the court, I expect it will create the largest software effort since Y2K to get rid of GPL'd software in any enterprise larger than about 10 people. The business risk would become too great that they might be litigated by a rival into inoperation. (I know the author's aren't out to kill companies, they are out to enforce their rights, but there are lawyers that like nothing more than extorting companies for large sums of money and those lawyers would write "As you know VMWARE was damaged to the tune of several billion dollars because they didn't use the GPL correctly, and we know that you aren't using GPL correctly either, why take the risk and get a writ of compliance for a mere $<large sum>." Or like the guys who are coming out of the wood work to "test your house for formaldahyde" by basically clipping some PH strips to a stick, waiting for them to turn blue, and them selling the poor home owner a multi-thousand dollar "remediation" service.

I really respect what the GPL tries to do, but especially early in its life it was way too ambiguous in its interpretation (consider the difference of opinion between FSF and Linus on loadable kernel modules for example)

http://www.timelex.eu/en/blog/detail/gpl-license-reaffirmed-...

Second, most businesses already avoid the GPL, but among those that can't, such as Android device makers - well, some of them seem to be fine with avoiding the edge cases by just publishing the source to their kernel modules. Considering this, I suspect that for the other half, it would not be the end of the world to have to reveal some minor secrets about the hardware or whatever; they just neglect to publish source because they are unlikely to be penalized any significant amount for doing so.

Of course, there is the risk of GPL making it into codebases by accident, but really only if you spend zero effort keeping track of it. There are companies like Black Duck that provide tools to scan a codebase for known open source code, and honestly a quick grep should often be sufficient as well. In VMware's case, I'm not sure about the executable loading or kernel module issues, but if "vmkernel" itself has code copied and pasted from Linux, as is alleged - well, they should have known better! No need for the largest software effort since Y2K when you could just ensure someone in the company understands the GPL and how it relates to the company's code, before said code is highly developed.

If a company want to remove this risk, they can't license someone else work. That mean no third-party engine, no font types, no xml parsing library. The cost vs risk is so insanely on the cost side that it would be fools and fools only who rather make everything inhouse in fear of follow simple industry practices that has existed for 20 years or more.

If you link together GPL'ed work with something you distribute, release the source code for that work. That is not a more risky proposition that say, giving 5% for using unreal engine.

Commercial engines generally have a clause in there that you are not allowed to bring any GPL code in contact with the engine, no exceptions.

Note this problem only applies (in practice) to GPL licenses which are explicitly disallowed in all gaming environments I have encountered. As such the problem is moot. There are lots of Open Source code in computer games, just not GPL. It's a non issue.

In practice, license conditions are balanced to the interest of the company. Some might be perfectly fine with 5% revenue losses in order to pay for a engine, where others would strongly disallow it. If this lawsuit succeed, my guess is that any damages will be far less than 5% total revenue from the years of Linux usage.

* Most of the time commercial libraries like unreal make a distinction between interface and implementation. In the worst case you could swap out their implementation and reimplement the methods you use, which will be expensive but at least at the end of it you'll have a clean codebase that you own. With GPLed code even that might not be enough - you could replace the GPLed implementation and your codebase might still be a derivative work of the GPLed library. * Epic is a profit-making business with obligations to their shareholders; ultimately resolving any issue with them is going to be "only" a matter of money, it's not in their interests for your company to go out of business. An individual GPL copyrightholder is more likely to be idealistic/spiteful about it. And GPLed libraries often have several copyrightholders, multiplying the risk.

Because not many other licenses like GPL "spread" to the entirety of the project. Sure, you can draft another license like this, but out of the popular ones there really is only the (A)GPL which does that.

If the GPL is litigated to the point of a published decision, doesn't that give the legal team of the targeted company the ability to say "actually, we know we ARE using it correctly, so buzz off"?

According to the lwn link 'JoshTriplett provided upthread, corporate general counsels often contact the Conservancy to request GPL enforcement: "Their interest, instead, is in a demonstration that the GPL has teeth so that they can be taken seriously when they tell management that the company must comply with the license terms of the code it ships. "

Only the owners of a copyright have standing to sue, so I think the complete version of your concept fails unless, like in this case, the copyrights are held by many people instead of being assigned to one person or organization. I also can't imagine that a lot of companies haven't considered this risk, it hardly matters to such companies if this hasn't reached the level of a contested (vs. default) decision if they believe it could if pushed. E.g. see this: https://news.ycombinator.com/item?id=9153278

EDIT: Maybe not true for copyleft software, reported not to be true in France (see below).

This is somewhat moot if the litigation happens in Germany of course.

In contrast to camperman's comment, the copyright issue is first determined as a matter of law (its instructive to see the Groklaw archives on this with regards to SCO) and then if the copyright is valid then the question of the license comes up. And party A will say "That wasn't what I meant" and party B will say "But this is what it says" and for each of the clauses and sentences in the GPL a discussion will be had on what it says versus what it means. And finally if what it means is something that can be legally enforced[1]. A good example of that was the AT&T versus UC Regents suit over BSD where the BSD license and the AT&T Unix licenses were extensively examined.

[1] You could write a shrink wrap contract that said "By opening this package you agree to give me all your future income." which is clear what it says and what it means, but you would be hard pressed find a court willing to enforce those terms with that contract.

A better one, that I'm a lot more familiar with, is Stephens Media, their property the Las Vegas Review-Journal, and https://en.wikipedia.org/wiki/Righthaven which produced a Federal District Court decision that in order to sue, one must actually own the copyright.

But if you're just worried about the prospect of being sued, none of this matters as you point out. New SCO dragged our corner of the world through a huge, long and expensive drama about code that they never had the copyright to, and knew in advance they didn't. Which means that risk exists for any code you don't write yourself....

I'm not sure the SCO case shows the progression you cite. As far as I can remember, everyone assumed SCO had the copyright, there was lots of litigation based on that, and it took years for the case(s) to wander around until in rather a big surprise we learned they didn't. Of course, to continue the theme of "(too) easy to get sued", there should have been severe sanctions levied on the parties responsible for that fraud, which somehow never seems to happen in the US.

That's not true, at least in some countries. For example, see this case in France from 2009:

http://arstechnica.com/information-technology/2009/09/big-gp...

> "We've long said the GNU GPL is enforceable, and of course we're pleased to see another court reaffirm that fact," said FSF France president Loic Dachary in a statement. "It's a commonly held belief that only the copyright holder of a work can enforce the license's terms—but that's not true in France. People who received software under the GNU GPL can also request compliance, since the license grants them rights from the authors."

However, I do think that parties receiving GPL software also have standing to sue, not just authors.

And this may also be true in the US: see for instance one of the decisions in Versata v. Ameriprise where the court considered GPL to create contractual obligations towards third party beneficiaries -- i.e. towards people receiving the code. That whole case has been settled out of courts but that's still an interesting decision (and to my knowledge, this decision had not been further contradicted by upper courts so it is still interesting).

What about people who receive the compiled software (customers) and would like the source code as well? Don't they have standing?

In the US, IIRC (and IANAL and this is not legal advice) license are analyzed under contract principles to the extent applicable, which might extend to standing, and, under contract principles, you could argue that those receiving the software are intended third-party beneficiaries with standing to sue.

So, its possible.

Using GPL SW in a product sold to clients is a different issue but in my (big) enterprise the GPL is a well-known topic: there are internal guidelines about what you can do and what you can't do with GPL SW and who you have to contact when you want to use GPL SW for clients products so that he can check that you're doing the right thing.

Maybe I worked only in surprisingly enlightened enterprises, but the worst fear I was informed about regarded lack of support by a "real" company.

http://sfconservancy.org/linux-compliance/vmware-lawsuit-app...

Disclaimer - I'm on the Board of Directors of Conservancy.

The law is a tool that can be used to achieve certain goals. Similarly, software is a tool that can achieve certain goals. We wouldn't suggest that people shouldn't use software (which, in turn, requires employing programmers) because "in the end, you're only helping programmers", would we?

I'm no fan of the legal profession: the only sticker I have on my laptop reads "Keep your lawyers off my computer". However, I work with lawyers a lot because they are experts on the law, and I (or my organization) can hire them to utilize the law to promote good in the world.

I find proprietary software abhorrent and those who write in my mind are doing something harmful. I find Free Software wonderful and those who write it in my mind are doing something wonderful. Similarly, some lawyers do good in the world and some do bad, just like programmers.