NuSphere vs. MySQL AB Hearing 184
An anonymous submitter sent in: "The hearing is over, and Adam Kessel posted a report about what happened in the court room. The judge is probably not going to issue a temp. injunction against NuSphere on the GPL violation, but probably WILL issue an injunction on trademark issues." Politech has another report on the hearing.
More on the NuSphere case (Score:5, Informative)
Re:More on the NuSphere case (Score:1)
way about source code.
more like: "hey, good luck cleaning up the mess man, I didn't have much time to desgin it. Oh by the way,
don't rearrange the order of inistantion of those
two global classes, they depend on each other, what
can I say, messy object model."
--
the short of it.... (Score:3, Informative)
Also the judge didn't want to get into the technicalities as they were way beyond her at this point.
MySQL wins on the trademark count (for now at least), but the motion on the GPL violation is being pushed to August.
Link to the GPL (Score:2, Informative)
Re:Link to the GPL (Score:5, Funny)
Did you just link to the GPL? Dude, I have a cron
job the searches my harddisk for the GPL and deletes it, incase I have
installed new software.
Infact, the perl script invoked by cron not only
does look for the name "COPYING", it also does an
MD5 checksum, and knows about the finger prints of
both GPL versions.
The next release will be network enabled, and we
are aiming for a full blown web crawler, of google
proportions, that does nothing but exploit servers
and delete GPLs.
The world needs only ONE GPL, in Stallman's box, and
the rest of us can symlink to it.
--
Re:Link to the GPL (Score:1)
immunition for a "viral" outbreak
--
Nice thought, but... (Score:2)
Are YOU going to be the one that pays to buy RMS a box that can take that many hits?
Oh yeah, a P.S. just for those that moderate this kind of crap up because they think Stallman is an egomaniac, a religious nut that wants everything to run on Gnu/Linux and all that rot.
Stallman.org [netcraft.com] is running Apache/1.3.6 (Unix) on FreeBSD.
So there. Get over it.
Nusphere fixed the issue (Score:5, Informative)
Nusphere fixed the basic issue here; they no longer distribute without source code.
MySQLs actions make little sense, they are acting like spoiled children who didn't get their way. (Incomprehensible to me; I think an ego clash occured)
MySQL can't claim damages with regard to the GPL because there aren't any. If you look at the information released the other day you'll note near the end (EFF statement) That Nusphere fixed the problem; That EFF encourages compliance, not court action; and the short acknolwedgement that they believe Nusphere violated the license.
You would note near the beginning they state the purpose of the GPL and their general intent to reach a compliance.
Re:Nusphere fixed the issue (Score:1)
I think the issue is that they STILL have proprietary code (shipping without source code) linking to the mySQL Libraries (thus violating the GPL).
Re:Nusphere fixed the issue (Score:5, Informative)
Assuming this is true (I have read conflicting reports), the GPL does not actually allow you to 'fix the basic issue' and then continue on as you wish.
Section four of the GPL [gnu.org] states:
Once Nusphere initially released without source code (they claimed source would be available later), they were, strictly speaking, in violation of the GPL and by section 4, automatically terminated any claims to a license that they had.
Now, most reasonable companies would allow them to get into compliance with the GPL, then drop the matter. MySQL decided not to. While I do not agree with this action, it is, strictly legally speaking, the right thing for MySQL to do in this situation.
Re:Nusphere fixed the issue (Score:2)
Normally, yes. However, it's important to note that the offended party has no requirement to forgive such violations, and that is intentional.
If this were a simple, onetime mistake I would agree with you. However NuSphere has been making an arse of themselves and being a thorn in MySQLs side for years. There is a lot of history here, and NuSphere's part is VERY blamewourthy. Given the history, I don't blame MySQL in any way for this - if I were them I'd do the same damn thing. Make no mistake, NuSphere picked this fight, not MySQL. MySQL just happens to have section 4 on their side.
Re:Nusphere fixed the issue (Score:2)
"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." (my emphasis added).
All NuSphere has to do is prove that MySQL AB distributed the program again and their licence was reinstated. The concept that they've lost their licence to redistribute because of past transgressions is baloney.
Re:Nusphere fixed the issue (Score:1)
Like it or not, that's one of the reasons of why the license works. If you violate it, you lose all rights to use/redistribute.
Re:Nusphere fixed the issue (Score:2)
Like it or not, if you violate the GPL (IMHO) it's quite clear that you regain your rights to distribute once you fall within compliance again.
Re:Nusphere fixed the issue (Score:2)
Using your logic, one is encouraged to break the law as there is no penalty which was CLEARLY not the intension. Rather, it's intent (which is equally clear) is to restrict your freedom once you've decided to spit on the GPL. This only makes sense as it's quid pro quo. That is, since the violator has decided illegally to restrict everyone else's rights to what should be freely available, which flies in the face of the intent of GPL, it in turn, ensures the violator has no rights which forces you back to the table greatly weakened.
Re:Nusphere fixed the issue (Score:2)
You seem to forget that the discussion isn't about entities that are CURRENTLY in violation of the license. There is a clear penalty for breaking the GPL still - you cannot distribute the code you produce until you are in compliance with the license. The whole need for "forgiveness" from the copyright holder is a rather weak interpretation which comes from people who see a need to perpeptually punish individuals or organisations that step over the line rather than the clear reading of section 6 which gives the intent to allow any company that is in compliance with the GPL's statutes to distribute further.
You also forget that even under some wild interpretation of the GPL which bars an entity forever from using a certain piece of software that it is a trivial exercise to create a separate entity which gains all the IP of the previous entity and yet is not encumbered by the "need for forgiveness" that you are clutching at.
In the end the "intent" of the GPL is to make software free (as professed ad infitium by the FSF). Chaining down companies for past mistakes is NOT making software free - it is a pursuit of vendettas. The FSF has also stated many times that their goal is to bring companies into compliance with the GPL rather than force them out of business. I'm sorry, but there is no intent (clear or obscure) to hold back companies which are currently in compliance with the license's terms.
Re:Nusphere cannot fix the issue on their own (Score:4, Informative)
Re:Nusphere cannot fix the issue on their own (Score:1)
ITYM "The copyright owner or anyone else who has licensed the code from the copyright owner under the GPL."
Re:Nusphere cannot fix the issue on their own (Score:1)
Re:Nusphere cannot fix the issue on their own (Score:1)
Now if you want to argue that this is not the _intent_ of the GPL, then we're just violently agreeing. But it's not the intent that matters - it's the text. The FSF will have to fix this loophole in GPL v2.0.
Nusphere can fix the issue on their own (Score:2)
"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." (my emphasis added).
In other words, they are automatically relicensed by the terms of the GPL as soon as they got a new version from MySQL-AB.
Re:Nusphere can fix the issue on their own (Score:2)
Subject to these terms and conditions - so the new licence does not shrive the sins comitted.
Re:Nusphere can fix the issue on their own (Score:2)
A new license does indeed shrive the sins comitted - in fact it's the only way to shrive them.
Re:Nusphere fixed the issue (Score:1)
Do we let anyone release binary modifications, and only punish them after they get caught?
The GPL becomes mostly toothless, if you can abuse as much as you like, but when you get caught, you "quick, release the source..."
So, the point can be made that Nusphere, it seems needed to be "taught a lesson" and used to make an example.
I don't know that I agree with the result, I don't really know enough about the case. But I'm sure that this is something to think about.
It's more about obeying the spirit of the law, rather than the technical license of the law. The GPL can't be too loose on enforcement, otherwise everyone will learn that you can fudge, but fix it if and when you get caught and think it might cost you...
Just some thoughts.
Cheers!
that's not the whole story (Score:2)
So, I find your assertion that the MySQL people are acting like "spoiled children" unfounded--there is more going on here. In fact, it isn't even clear to me that NuSphere fixed the problem. Even if they did, it isn't their's to fix: once they fell out of compliance with the GPL, they can't fix it--they have lost all rights to using the code.
As for the issue of reaching compliance, if anybody can violate the GPL with impunity and then just fix it up without consequences, that will simply encourage a lot of companies to do that. Open source is at a serious disadvantage here because most GPL violations are very hard to detect. It makes sense to pursue some GPL violations seriously even if the violators eventually come into compliance.
Finally, as for damages, that's bogus. Open source companies live off things like consulting and documentation. If some other company comes out with an enhanced closed source version that ends up getting used widely, the open source companies lose business. The argument isn't as simple (or simplistic) as the MPAA's "1 million people copied Rocky XVII illegally", but there are still damages. And there are damages to the public as well (although they may be harder to enforce).
Re:that's not the whole story (Score:2)
It's rather hard to violate the GPL without involving copyright infringment and a contractual dispute anyway. Since the GPL is all about contract and copyright law in the first place.
The tradmark bit is is a separate issue. Maybe related to something like NuSphere passing off their product as MySQL or similar.
Re:that's not the whole story (Score:2)
Re:Nusphere fixed the issue (Score:2)
At the time, the /. crown seemed pretty upset that RMS was being so pedantic about license complience, though I found that he was just pointing out a technicality.
So, it would appear that RMS, at least, intends for GPL violators to be forever prohibited from redistrubution of the relevent work unless forgiven. This is not inconsistent with also suggesting that returning to complience result in forgiveness - it just gives the GPL some teeth.
Good summary (Score:2, Interesting)
IMO, this is a very nasty detail that's left hanging in the license, and should be fixed in the next version. If nothing else, the license should explicitly say that you lose all rights to use the software "until you are granted explicit permission, in writing, by the original author", or something similar.
Re:Good summary (Score:1)
No one is saying NuSphere can't *use* the software, what they're saying is that NuSphere intentionally violated the GPL and that in addition they're being fookin' annoying to MySQL AB so the AB folks are not going to just let them start *distributing* the software again, compliant or no.
As to some of the issues you mention, I think it's fairly obvious that "you" in the license means the person or entity involved. So conceivably NuSphere could engage in some old shell game and maybe become a different legal entity and then be able to distribute again-- but that's not easy, cheap, or likely to make much sense to their customers.
But the whole thing about sharing the code doesn't make sense at all. If you share the code, you are following the GPL. If you share binaries, eventually you're going to want to recompile (if you're a distributor) -- not possible without the code. Which throws that whole hypothetical paradox on its ear if you ask me.
No original author... (Score:2)
The GNU license is open-ended in such a way that you could set your "command structure" any way you want it. By its very nature, the code/document it protects is public-domain, and no "admin power" is give to anybody.
Perhaps a time of about a year might work, but then again, anybody trying to screw over the licence like that deserves to never touch the project again. Future generations will learn to look over their licenses very carefully if they don't want to get burned.
I'm sure this will give people a message to never fuck with the GNU license, the projects it protects, and never EVER fuck with the FSF!
Re:No original author... (Score:1)
FSF: All your code are belong to us... (Score:1)
Technology and Judges (Score:3, Insightful)
The current system seems kinda like slashdotters passing judgement on the latest fashions from Paris.
Comment removed (Score:5, Insightful)
Re:Technology and Judges (Score:3, Interesting)
What was wrong with MPAA v. 2600? Or the Felten case? Or the Sklyarov/ElcomSoft case (though that has yet to come to trial)?
In case you don't remember, in the 2600 case, the DMCA came up, and free speech got reamed.
Comment removed (Score:5, Interesting)
Re:Technology and Judges (Score:2)
The RIAA and the MPAA fight dirty, they only go after cases where they know they will win and public opinion will be with them. They have deep pockets and lots of lawyers. But there is a chink in thier armor, when they threatened Felton, the EFF reacted by immeaditly going public, when public opinion turned against the RIAA they backed down, Felton gave his speech and then the EFF lost in court by winning. What should have happened is when Felton recieved the threatening letter he should have forced the RIAA to file the suit by writing the following message back.
TO: RIAA
From: Prof. Felton
Subject: DMCA
Body: BLOW ME!
This almost certainly would have produced the desired effect, once the lawsuit was filed, and the rest would have been history. Of course 20-20 hindsight is a wonderful thing.
Another thing the EFF should consider is going on the offense and filing a class action lawsuit against everyone who has presued prosecution under the DMCA or possibly the US Government. In order to do this real harm must be shown, and this is were the real problem is. I personally consider loss of freedom of speech, freedom of the press and fair use, to be real loss. However in post 9/11 America, I seem to be alone, personal freedom is at an all time low and no one cares. In order to carry out a class action lawsuit, the EFF would need to prove real financial loss on the part of the victims and prove direct profit on the part of the defendent. Not an easy task when the RIAA and the MPAA have already characterized themselves as the victims of Hacker, Pirates, Thieves and dope smoking hippies.
Re:Technology and Judges (Score:1)
How many small time companies can afford to pay lawyers while a judge spends day after day trying to get his or her head around the basics of the technology?
How open to FUD is the system while judges don't know what they are really dealing with?
The Supreme Court is a speciality court dealing with constitutional law. The Family Court specialises in family law. This is because they are very complicated areas. In other areas of society we employ specialist professionals (an aeronautical engineer handles different work to a naval engineer or civil engineer), why should the law necessarily be different?
I'd rather have a judge who knows that Word Perfect and MS Word are different products making decsisions whihc have big ramifications on the Net and technological sector!
Re:Technology and Judges (Score:1)
Re:Technology and Judges (Score:2)
Simply because a case involves technology does not mean that the details of the technology are relevent. Copyright infringment is not fundermentally different in the case of this kind of GPL infringement from a publisher failing to pay an author or someone hiring a band to play a venue then trying to change the rider after the performance...
A speciality court could assume a great deal of knowledge on behalf of judges resulting in quicker, cheaper and more informed decsions.
Or it could be a small clique easily bought off by monied interests...
Re:Technology and Judges (Score:1)
The MS EULA isn't a valid licensing agreement. It doesn't even meet the basic requirements of contract law. No court has upheld a EULA as a valid form of contract.
Max
Re:Technology and Judges (Score:2)
Nor should we need special laws for specific technologies either. A well written law should cover doing something old with new technology perfectly well. Only if a new technology makes something new possible do you need new laws. N.B. simply making something easier does not make it "new".
First, the law is The Law. It's interpretation should not change based on the plantiff or its paritcular singular application - a violation of licensing agreement should be enforced the same whether it is MS's EULA or the GPL.
IMHO It makes more sense to group the GPL with publishing contracts or even contracts for live music performances
Qualifications (Score:2)
Re:Technology and Judges (Score:3, Insightful)
Rather the judge is incompetant if they attempt to make a ruling when they don't understand the case. All he or she has to say is something to the effect of "If you can't explain your arguments to my understanding I'm dismissing the case."
Irreperable Injury (Score:5, Interesting)
The GPL is not a gift. It is a barter offer: I allow certain carefully constrained uses of my IP and in return you offer any intellectual property you create some as a result of my good will back to "the pot". In software development, time is a critical element: if one competitor can benefit from the IP of the other and mix in its own IP without disclosure, then the cheater has gained an unfair advantage. If its cheating is part of a business venture, then its entire profit stream is usurped from the good faith party as a direct result of its bad faith dealings.
For this reason the GPL essentially says "if you ever cheat, you get the death penalty". Without this clause, freeloaders could perpetually use gamesmanship and delaying tactics to obtain competitive advantage in the bartering of "hot" IP. To assure that bad faith cannot usurp the original author's valuable IP contributions, the GPL revokes all rights for the smallest infraction in order to assure that any "negotiation" will occur with the position of strength emphatically given to the original author.
The irreparable harm inflicted by not issuing an injunction is that NuSphere can continue to profit from a business model that is based on outright theft of intellectual property. The moment they were notified of GPL noncompliance, section 4 of the GPL revoked their licence. They are distributing and indeed modifying a copyrighted work in direct defiance of the authority of the copyright holder with the intent to profit. That is not only irreperable harm, it is stealing, and it is CRIMINAL if it is done willfully (17 USC 504).
Re:Irreperable Injury (Score:1)
Sorry if you disagree. I don't mean to mimimize the impact, but I do disagree with the term.
By the way, very good comments. I do like the "Death Penalty" analogy. I also the barter analogy. It seems as though many argue that the GPL doesn't put any restrictions on you, but it does. (Perhaps they don't like the word "restrictions" huh? [Grin]) Anyhow, thanks for the comments.
Cheers!
past or current violation? (Score:2)
Re:past or current violation? (Score:2)
Re:Irreperable Injury (Score:2)
Right now legally speaking the company is innocent until proven guilty. (Even if we know otherwise or think otherwise) This is a justice...
The comparison to piracy is incorrect since that has already been determined to be illegal. Remember how long it took to make hacking and cracking illegal.
Re:Irreperable Injury (Score:2)
But copyright isn't new, contracts licencing a third party to make copies of a copyright work (in return for consideration) arn't new either...
Re:Irreperable Injury (Score:2)
Also the GPL has a lot more in common with regular contracts than any EULA. Indeed it isn't an EULA in the first place it's a copyright licence...
To assure that bad faith cannot usurp the original author's valuable IP contributions, the GPL revokes all rights for the smallest infraction in order to assure that any "negotiation" will occur with the position of strength emphatically given to the original author.
Or if you prefer it ensures that the original author maintains the position of strength which copyright law gave them in the first place.
Breaking the GPL is conceptually little different from an author licencing a book to a publisher and the publisher failing to pay in accordance with the agreement.
Re:Irreperable Injury (Score:2)
While you are essentially correct, copyright infringement is NOT theft. It's copyright infringement. Courts have traditionally interpreted copyright in a different light from fee-simple property rights, so this shouldn't be any big news.
Re:Irreperable Injury (Score:3, Insightful)
By the way, by 17 USC 504(a)(1), if NuSphere's actions were "willful" then they are in fact criminal.
Re:Irreperable Injury (Score:2)
So the way I read that (seems pretty clear), if found guilty, NuSphere is liable for MySQL's actual damages and any money NuSphere made off of MySQL. I don't see how this court-enforceable liability makes NuSphere guilty of criminal acts, whether willfull or not.
Can you explain?
Assign everything to the FSF ?!? (Score:1)
(1) NuSphere probably figured they could probably violate the GPL and MySQL would not have the resources to sue them. This is probably true of most free software developers who have not assigned their copyrights to the FSF. After watching this hearing, it seems to me imperative that developers assign their copyrights to some party that will have the resources to enforce it (whether the FSF or some other organization). These violations can only become more common as GPL software becomes more powerful and widespread, and it is quite expensive to litigate against them.
Uh, say what? Does he mean that the FSF has no legal right to assist another party monetarily in courst unless the FSF has the assigned copyright? I somehow don't think legal rules are that restrictive. I read that as, "if you don't give your code to us (the "free" world)", we don't care spit about you if someone hijacks it, GPL or no."
Please tell me I missed something here...
Yes, you did miss something here (Score:3, Insightful)
Re:Yes, you did miss something here (Score:1)
Re:Yes, you did miss something here (Score:1)
The FSF could provide legal counsel at minimal cost to ANYONE they desire. (Remember Paula Jones - she was represented free of charge by a highly partisan group, if I recall correctly)
You can't sue _George_ for injuring _me_. Only the injured party can *ACTUALLY* sue George. *BUT* you can help me pay for or provide legal council for me to sue George.
For the FSF to say they can't help is a crock. Perhaps they choose not to, or perhaps they don't have the funds, but they CAN help.
Cheers!
Additional point/question? (Score:1)
But this brings up another point. If you assign all your rights to the FSF, you wouldn't have the right to then sell your code to some commercial entity (sans any added GPL additions) because you wouldn't have any rights other than those granted under the GPL.
Would it be possible to grant just the GPL portion of your rights to the FSF, and retain the original rights, so you could take your original code private/commercial/closed. (I don't know much about assigning rights to the FSF, so it might be that this is how it works now, but I suspect not...)
This might make it unadviseable to assign rights to the FSF if you might want to exercise your right to close or commercialize your software. Again, I do understand that you couldn't take the resulting GPL code with other's changes and close it. But you could with the original.
Just some thoughts...
Cheers!
Re:Yes, you did miss something here (Score:2)
The injured party can however appoint someone to act as their behalf. Depending on the context this is known as either "appointing a proxy" or "granting power of attorney".
Re:Assign everythi Getting two dozen authors on fo (Score:2)
If all of these contributors assign to the FSF the copyrights on their respective portions of the project, then it is much easier for the case to proceed, since there is just one plaintiff, and no nitpicking over who owns what part.
So I don't think that the FSF will refuse to help you if you don't assign them your copyright; it just makes the proceedings much simpler if you do.
Clarification on GPL vs LGPL for dynamic vs static (Score:3, Informative)
So, I've always gathered that the LGPL exists to permit a program to link to libraries - but my impression was that the LGPL was phrased such that code could statically link to it -- such that the result would be a single executable, and that even under the GPL, it was legally permissible to write code that would rely upon a dynamically loaded library. Can someone confirm this is the case? IE, you could write a program which required that a person have libmysqlclient.so* installed for dynamic loading and that program could be closed source, but you could NOT produce an executable which statically linked that into the executable unless libmysqlclient.so was distributed under the LGPL (which, of course, it is not).
Re:Clarification on GPL vs LGPL for dynamic vs sta (Score:5, Interesting)
Even if the source code is not "integrated", when the result of compiling is an indivisible executable object code, the result is a derivitive work. If I write a library and you use it, the resulting object code is a derivitive of both source files. The GPL specifically mentions and includes this case. The LGPL is offered as an alternative when the harsh consequences of the GPL are not what the author desires. End of story.
dynamic vs static linking (Score:3, Interesting)
All your source -> your executable
All GPL source -> GPL shared library
When your executable runs, it then loads the GPL shared library in order to have access to the required object code. If it doesn't find it, it has unresolved symbols and craps out. It is, therefore, clearly dependant upon that shared library, but it is, by the same token, clearly NOT a derivative work, since 'your executable' is derived without any GPL source code.
The static executable under GPL is a clear cut violation. But that leaves static binaries under the LGPL (I'm unsure), the dynamic under the GPL (unsure, think legal), and dynamic under the LGPL (must be legal -- hell, what would the LGPL be good for if it wasn't?
Re:dynamic vs static linking (Score:2)
Correct, but that's not an issue here, since my understanding is that NuSphere is statically linked.
If I'm wrong about that, then we need to delve deeper. If executable A can be extended by dynamically linking binary B, then the two executables are A and A+B, both of which are obviously derivitives of A when they are in memory. I also think that the law is pretty well settled that in-memory copies of code are "copies", so that even the A+B case is an integrated copy once its in memory, which would still meet the definition of a derivitive work. Moreover, if B was designed from the outset to only be used with A, then A+B is a derivitive.
The judge appears to have asked the right question when she asked "can NuSphere's binary run without MySQL installed on the same machine?" Answer No. Conclusion, this is not two separate programs interoperating, but a derivitive form of one program.
Re:dynamic vs static linking (Score:3, Informative)
These requirements apply to the modified work as a whole. If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works. But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it.
The only way they can get around not GPLing their software is if it is independant and not distributed as part of a whole.
In a nutshell this means that the software must not completely depend on the library to function, it must be able to survive having that library removed and still able to operate or else is it not independant but dependant on the library. Also it cannot be distributed with the GPL'd library as part of a whole program which does NOT necessarily mean statically linked but in the same package! Be that tarball or CD-ROM.
If these two conditions are not met then the software must be bound by the GPL regardless of object or source code, static or dynamic library.
-- iCEBaLM
Re:dynamic vs static linking (Score:1)
Re:Clarification on GPL vs LGPL for dynamic vs sta (Score:1)
Re:Clarification on GPL vs LGPL for dynamic vs sta (Score:2)
Correct.
But the GPL does add additional restrictions if you distribute commercially: you must either provide source with binaries (if you provide binaries on "media", or a transferrable written offer to provide source. If you do not distribute commercially, it is sufficient to provide a reference to where you obtained the source.
One issue is whether pointing to an FTP server counts as providing source "on media" (for the case where, for example, you ship binaries on a CD, with a link to an FTP server for source). My understanding is that the FSF and RMS think "No." This is because network access may be nonexistant or prohibitively expensive for some.
Of course, if you provide both binaries and source on a server somewhere, you're O.K.
Re:Clarification on GPL vs LGPL for dynamic vs sta (Score:2)
Clearly absurd, and unfortunately typical of NuSphere so far as honesty goes. Which is why I am NOT in any way angry at MySQL AB on this case - if you've followed it at all you know that NuSphere have been arses to them repeatedly, while relying on their code to sustain their business, so it's hard to criticise MySQL for just saying enough's enough, cease and desist you dumb f*ckers.
The Lesser GPL is indeed so phrased, and that is why Stallman recommends its use only when a pre-existing proprietary library beat us to the punch so far as functionality goes. If there is a reasonably priced proprietary library that gives you function X, then a GPL library that gives you function X isn't particularly attractive. So in that case it should be LGPL instead.
As to the rest... I think it's pretty clear to anyone with a small fraction of a clue that a statically linked program is a derivative work. The question remaining is whether a dynamically linked work is a derivative work as well. That's really the only area where FSF interpretations are likely to face any significant court challenge, and the FSF is and has been aware that their view may not prevail from the beginning - their publications make this clear.
The "official" FSF interpretation, if memory serves, is that dynamically linking a GPL library creates a derivative work. Yeah, that's a period.
Before the peanut gallery starts raving about the FSF being hypocritical here, let's just remember that this is a standard tactic in any legal or political arena, and the FSF would be selling us all short if they didn't use it.
Based on legal history, if it came down to the wire a court would be unlikely to uphold that claim in the event of a program which was dynamically linked against a GPL library but would also function if dynamically linked against another library instead. However, if there was no non-gpl library it could be dynamically linked against instead, the courts would likely side with us. Courts may not understand code, but they do understand pretty easily the difference between a work which has no value without X and a work which can work identically with X or Y.
Even if there is a non-GPL library now, if the work in question was distributed before the non-GPL library was available, the courts would likely call the work a GPL-derivative. Just one more reason to license your library under GPL NOT LGPL unless you are certain your library simply duplicates functions already available.
Not About Linking (Score:2, Informative)
Nusphere lost the right to redistibute MySQL because they made a modified version without the source code for that version being available: "Any attempt otherwise to copy, modify, sublicense or distribute the Program is void, and will automatically terminate your rights under this License."
The MySQL authors therefore have the right to stop Nusphere redistributing their software.
MySQL AB undoubtedly have their own reasons for being so stiff necked.
Amend the GPL (Score:4, Insightful)
Specifically, I would propose that use of GPL software be conditioned on legal stipulation of the proposition that continued violation of the terms of the GPL after notification is "irreparable injury" because it is theft of intellectual property.
Similarly, you should have to stipulate that the "balance of harms" of a violation is completely in favor of the author, and that this applies regardless of any business interest that exists as a result of the exploitation of the licence. The author offered a zero-cost good faith licence to use and extend their IP subject only to the condition that IP extentions be returned to the commons on demand. In many cases the GPL'ing author makes this deal believing that it will produce more value than retaining tight proprietary interest and selling the software for its market value. If someone else builds a business by skirting the extremely minimal requirements of the IP barter arrangement, then any revenue stream that results should be viewed by stipulation of the licence as being misappropriated and unjustly earned. Indeed pirated or stolen.
Thirdly, I would propose that the section 4 "death penalty" provision be made substantially more blunt. In particular, a "take down" process should be invoked (somewhat like the DMCA provision) that says if you recieve a letter of noncompliance from an author that your rights cannot be restored other than by a written order of the author, an arbitrator (see below), or a judge. Some people have noted that if you receive another copy that the licence can be read to re-grant a new and separate licence. All such weasle room should be removed by stating that this doesn't apply if you're licence was revoked for noncompliance.
Fourthly, I would propose that the licence should designate that any disputes be settled by arbitration by the FSF or by the EFF if the FSF is a party with the loser to bear all reasonable costs incurred by both sides.
Not so hasty.... (Score:1)
As far as I can tell, this entire case is MySQL trying to protect its market share. Nusphere's strategy seems to have been to drag its feet releasing source code to make it a more attractive partner than MySQL. It could offer to support MySQL code, but not vice versa. If you're a firm, your choice of partner is a no brainer.
The problem is that the GPL was not designed as a commercial license. Firms are not EXPECTED to make money selling code.
Re:Amend the GPL (Score:2)
Why do you need to ammend the CPL to cover this. Copyright infringement is already against the law. If someone distributes the software in violation of the GPL that is exactly what they are doing.
Similarly, you should have to stipulate that the "balance of harms" of a violation is completely in favor of the author, and that this applies regardless of any business interest that exists as a result of the exploitation of the licence.
Again this would appear to be redundant since there is plenty of statute and case law related to copyright infringment.
The author offered a zero-cost good faith licence to use and extend their IP subject only to the condition that IP extentions be returned to the commons on demand.
It's not "zero-cost", simply that any cost is not in terms of money. To be valid contracts require transfer of "consideration", this includes money, but also includes barter of goods or services, etc.
n particular, a "take down" process should be invoked (somewhat like the DMCA provision)
Why not simply use the provisions of the DMCA?
All such weasle room should be removed by stating that this doesn't apply if you're licence was revoked for noncompliance.
THe only thing which might be a useful ammendment.
Re:Amend the GPL (Score:2)
This would make it unnecessarily difficult to enforce the GPL outside the US (ie, with both parties outside the US). Also, some countries do not have arbitration at all, and it's hard to predict what a judge in such a country would say to such a designation.
Re:Amend the GPL (Score:2)
Also, there is no reason why such disputes have to be arbitrated with both (or even either) parties physically present.
Re:Amend the GPL (Score:2)
I see one risk with such a clause. People in other countries could be tempted to release their software under a license in which they replace "US law" by "law of country X" (with X being their own country). This would make it even harder to enforce the GPL.
On a more personal note, since I come from a country which doesn't (yet) have software patents and a DMCA-like law, I'm quite glad that the GPL doesn't refer to US law specifically. :-)
Re:Amend the GPL (Score:2)
Re:Amend the GPL (Score:2)
Re:Amend the GPL (Score:2)
Because infringement can be inadvertent and accidental. I'd rather have a "grace" period spelled out, than have to rely on the forgiveness of the license holder. 30 days strikes me as reasonable. Of course, any time you put a specific limit on something, it will be too long for some, and too short for others.
The recent affidavit before the court is a testament to the fact that the FSF prefers complience to retribution (as, it appears, do most slashdotters). If that is the case, formally giving up the right to vengeful persecution for the slightest inadvertent infraction, is certainly in the correct spirit.
Of course, this defense should not be available for repeated, intentional, violations opening a 30 day window of opportunity each time (and I suspect that is what you fear). I think that acceptable wording could be drafted to balance heavyhandedness for blatent intentional infringement against grace for temporary lapses, even though reasonable dilligence was exercized.
The bottom line is that while I like the idea of a GPL "death penalty", I also think that some safeguards be in place so it is not abused.
Let me give you an example of an inadvertent infringement: I once sent a RedHat-derived CD to someone (with our own GPL code and source) and forgot to include the Red Hat sources. Technically, I was in violation of the GPL even though I already knew they had the relevant sources. They didn't even want me to send them something they already had. However, the GPL did not cover the relationship betweem me and them, but betweem me and the copyright holders so what they wanted was irrelevant: I either had to provide the sources, or a transferrable offer to provide them. I was not prepared to do the latter, because we were not prepared to provide source to any and all who asked for it and didn't obtain binaries directly from us. So, I made sure I sent a source CD with the next update.
RMS was aware of this slip (we were discussing subtle nuances of GPL complience at the time and the pragmatic problems some requirements caused (hint: making a supportable source distribution is not always easy)) and, IIRC, considered their statement of already having the sources "good enough" to qualify as our meeting the requirement that they have them. However, the FSF was not the only copyright-holder affected, so I made sure to dot the i's and cross the t's as it were and sent sources anyway.
Should I suffer a "GPL death penalty" because of this gaffe? I don't think so, because (a) I was not grossly negligent, (b) I compensated for the harm caused (sent source), (c) acted in good faith. If the case went to a civil court, I'm sure I'd prevail: civil courts are about restitution, and not revenge.
Omiting reasonable "forgiveness" clauses in a license leaves the matter entirely up to the courts, and they might be more forgiving than the copyright holder might wish. Putting them in strengthens the copyright holder's position, IMHO, and does not weaken it.
Bcast2k saw this coming (Score:1)
"...thus likely that she will not actually address the terms of the GPL itself in her decision, which is a good thing, since there was so little expert testimony about it..."
It's as if open-source software is destined to transcend limits of a hobbyists project and mutate into a corporate-sponsored system. Programmers fill a need with MySQL. Pretty soon, it is MySQL that fills needs within other corporations. Broadcast 2000 video editing software from Heroinewarrior.com was the best bet going. It worked, cost no bones, and featured GPL liscensing. Soon enough, dot bombs pop up and Bcast2k is pulled because of liability concerns. Corporation Soon2be Inc., need video editing and fills that need with Broadcast 2000. Who's liable if a setuid/gid root
it's looking scary for hackers who share their toys and get organized / sponsored. pitch your product to company Soon2be Inc., and when they merge with Around4Ever Corp., watch yer back.
Fat Lady Not Sung (Score:5, Informative)
It's clear that allowing NuSphere to continue shipping code they have been shipping for the last seven months won't, by itself, do much more harm, if any, so the GPL issue was left alone.
Trademarks are another matter -- the more and longer they are abused, the greater the harm to the owner.
When the case itself goes to court, the text of the GPL will leave the judge little choice: NuSphere's product really is a derived work, and there is lots of case law about derived works, even with software.
For some background, see my letter to LWN [lwn.net] last year. (Scroll down to the end. Incidentally, it appears I was the first person to tell Monty about this feature of the GPL.) Evidently it took this long to establish that NuSphere just wouldn't figure out where they stand without help from a judge.
Re:Fat Lady Not Sung (Score:2)
Trademarks are another matter -- the more and longer they are abused, the greater the harm to the owner.
Usually the same reasoning is applied to copyright infringement too, though.
My politech response, my bad (Score:1)
My appologies to Mr. Kuhn. I had met him at the open source convention in San Diego, and spoke with him about the FSF, which he is Vice President of -- the man on the street, if you will. He was extremely helpful and quite expedient in helping me with my GPL'd project and understanding license issues. (Thanks Brad - My bad!)
Anyway - I wish I had read this before I came down so soundly on MysqlAB's side: http://www.nusphere.com/misc_stuff/declarationofb
Why does the name Bruce Webster sound so familiar? He must have argued this GPL issue before?
Abe
Section 4 isn't the end of it (Score:4, Insightful)
That's not the end of the story as most people seem to be professing. The real issue comes when someone from NuSphere downloads ANOTHER copy of MySQL including the source, especially if this is a newer or later version or if they obtained it from a different party. Does this new download constitute a new contract between the parties, and if so then all a GPL violater has to do is download a new version and/or download from a different source and wham - new license, new contract, all sins are forgiven.
Failing that argument, the (rather simple) alternative is to create a new and independant legal entity, assign all your IP rights to the new entity (including the trading name) and the new entity downloads a new copy of the source and is immediately in compliance with the GPL.
The real danger here is that the GPL may and probably does have a "built in automatic forgiveness" clause. If you break the GPL then it seems likely that bringing yourself into compliance and relicensing is a trivial exercise.
Note: IANAL... I did see one on TV once though... In Australia... It was a British show... Ahh whatever.
Re:Section 4 isn't the end of it (Score:1)
Someone download MySQL X.XX, that is not employed by NuSphere (could be someones brother for that matter). They go though the code and rewrite portions, add stuff, and rename is to XxSQL and distribute to NuSphere.
I would say that it is legal as long as it is all GPL the entire way. There is NOTHING in the GPL to keep me from renaming a project (AKA CODE FORK).
Sure it would not be MySQL anymore, but there is no reason why they can't track the protocol and ABI. It is GPL remember?
How does that one sound?
BWP
Re:Section 4 isn't the end of it (Score:2)
"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." (my emphasis added)
Please stay home next time (Score:1, Flamebait)
If you want to support someone at any future GPL related hearing, please stay home (unless you're going to act like a grownup).
All we need to do to convince a judge that we're a bunch of kooks is dress like we're silicon valley engineers (or worse), play Quake on our laptops, and snicker when the judge makes technical mistakes. I learned not to be a snot to the authority figure (think: professor) back when I was in college, but apparently, some people never learn.
Trademark Issues??? (Score:2)
THe tradmark issue would apply only to NuSphere attempting to pass their product off as being MySQL. With the GPL issue the relvent matter is copyright violation.
The first linked article completly muddles the two up. The second article does at least mention that there are at least two separate issues to the lawsuit.
Why not in sweden? (Score:1)
Does the GPL not hold outside the US? Do judgements in foreign courts not hold any weight in the US? Or is it simply because they had a better chance of winning a trademark case in the US? (or is their Trademark registered in the US?)
They should have known that a US judge is never going to make a decision that harms a US company to the advantage of a foreign one.
Re:Why not in sweden? (Score:1)
Thoughts (Score:2, Interesting)
From Adam Kessel's Comments:
[The Judge] seemed to think "linking" code was analogous to hyperlinks on the web
In the case of dynamic linking, isn't it? I can link your graphic from your web server inside my web page. In both cases I'm telling the software (either a browser or the OS) where to go to get the rest of the relevant information. In the abstract sense, how is dynamic code linking unlike web linking?
It seemed like she [...] had trouble seeing how [immediate and irreparable injury, loss or damage] was the case
This is the clincher for pretty much any temporary injunction filed based on the GPL. The GPL requires that no money can be due to authors of the work for its use. No money, no damages. No damages, no injunction without a full blown trial.
The only way I can see to get around this would be to offer the software under a dual license where the second license required significant remuneration. That way you could argue that if they failed to comply with the GPL then the second license applied, which would mean a breach involving a lot of unpaid cash.
<Rant On>
Why are we fighting this fight anyway? If we as software developers choose to give our stuff away for the common good, what does it matter if someone who doesn't share our values manages to make a buck or two off of it? If the Open Source / Free Software doctrines are sound, then such folks are going to lose in the long run anyway. Its not as if there is a shortage of folks out there who are willing to be a part of this without being forced. Not for the better part of a decade.
Think about it: If Microsoft could swipe the Linux kernel, wrap Windows around it, and sell it without the source, what would the fallout be? Millions of users would see a more stable and secure operating system, hardware manufacturers would have an easier time of providing drivers for Linux (since they would be essentially the same), and Redmond would be financially motivated to keep coming back to the core linux source tree rather than a single-shot fork. Socially, Linux in the public domain would be far more powerful than Linux under the GPL.
Besides, saying that I "freely" offer you my stuff but you must subscribe to my social values to get it seems a little like a 21st century version of Marxism. Who do we think we are that we should have such power? Wouldn't we be better off to just release in the public domain?
<Rant Off>
Re:Thoughts (Score:2)
Give me all your money. Socially, I can be more powerful with your money than you so give it all to me.
So many people miss the idea of the GPL. It's not just about sharing code, but also protecting the individuals who share their code. It makes sure that noone screws me if I'm doing something nice for them.
It would be horrible if MS stole the Linux source code and never rereleased it. I would never again be able to hack the kernel when I needed or add an unsupported device driver.
It is not marxism, it's capitalism. The GPL is oriented towards preserving individual rights of a piece of code. Public domain abandons individual rights. I think you are a tad confused here.
Re:Thoughts (Score:1)
But that's totally irrelevant here.
Re:Thoughts (Score:1)
[The Judge] seemed to think "linking" code was analogous to hyperlinks on the web.
In the case of dynamic linking, isn't it?
Yes, except the Judge clearly wasn't thinking of things like image tags; she was analogizing the linking of code to the links you click on when you are viewing a web page that take you to other sites.
By the end of the hearing, I think she understood that linking was more complex than underlined words. She also knew there wasn't time to deal with the issue in this hearing.
Re:Thoughts (Score:2)
I'm not sure whether that point is valid. Just because I choose not to sell something doesn't mean stealing it, is ok. If I lend a book to a friend, I don't loose ownership of it - he still has to give it back. The same applies if it's not a book, but say a bicycle which I made myself, and which I lend to all my friends freely.
Courts are quite used to assign values to goods, even if they don't have a price sticker. I doubt they'd be stumped in this case.
Re:Thoughts (Score:1)
Eben Moglen's affidavit (Score:2, Informative)
Judge Saris rocked the house (Score:3, Informative)
Clearly she didn't know or care that this is considered the big "test case" for this culturally iconic GPL. To her, this was just another license and copyright case. The question of whether the GPL is "valid" never came up; it was assumed to be as valid as any license. She knew, probably from the 200 pages of briefs she complained loudly about having had to read, a bit of the history and significance of the FS/OSS movements (when the MySQL lawyer began a diatribe about the importance of keeping software free, she interjected, "Yeah, yeah, I know, it's like a religious movement, it's Open Source!")
It was just as clear that she knew nothing about software and made no bones about it. She started out pronouncing the product in question "Mice Quill". (And then she has to deal with the plaintiff calling it "My Ess Cue Ell" and the defendant calling it "My Sequel"!) And, as has been remarked, when the MySQL lawyer used the phrase "single executable file," she interrupted and demanded that he clarify, "Is that like, when I click the icon for Microsoft Word Perfect, just one window comes up?"
However, like any good judge or lawyer, she was focused on avoiding time-consuming litigation with expert testimony and the like. She saw past all the technical arguments and FS/OSS crusading and recognized this case for what it was: a business relationship gone bad. She stopped in the middle of NuSphere's testimony and asked if the two businesses wanted to work together again. Both sides were taken way aback, there formed these two little clusters of suits as the lawyers and the company reps conferred; each lawyer in turn blustered for a minute about his client's demands and then said yes, they'd love to. (Significantly, this is where the judge got NuSphere to promise to take the EULA off their product.) Judge Saris then got her bailiff to get on the phone and find them a qualified arbitrator *right away* and offered them her courtroom for the rest of the afternoon. In essence, "Will you kids please just get a room and figure this out!"
It was a bit disturbing that she was so reluctant to see irreparable harm in letting a company abuse the GPL like NuSphere did and get away with it. I think that was mainly her wanting to avoid a long trial unless it proved essential, and making NuSphere quit using MySql would basically shut down NuSphere, and would thus require a long fierce court battle.
It is important that even cases like this, dealing with highly technical matters, be judged in a regular court by regular legal types with no prior knowledge of the field. Just like a case between a bricklayer and her construction company, or a musician and his record label. Because they need to be judged based on regular law, in proceedings comprehensible to an intelligent but uninformed outsider. If the judge had known the difference between static and dynamic linking, she would have been an industry insider and would have already had a bias one way or the other.