Slashdot Mirror
MySQL AB and Nusphere Go to Court Over GPL
A little fairy whispered in our ear: "MySQL AB is seeking a temporary injunction against NuSphere, even though they've finally released the source code for Gemini and MySQL Advantage. According to the GPL, NuSphere lost the right to redistribute when they violated #3 by not providing the source code originally. The FSF will testify tomorrow in court, according to this Newsforge article." Newsforge and Slashdot are both part of OSDN. We've done a couple of previous stories about the MySQL AB vs. Nusphere conflict: the original story, a follow-up, and a note about a countersuit.
Update: 02/26 21:15 GMT by T :
bkuhn (Bradley Kuhn of the Free Software Foundation) writes: "The FSF has a press release on the matter and affidavit that we filed is also available."
Is this the first legal test of the GPL right now? Will it be setting precedent?
--------
Bleah! Heh heh heh... BLEAH BLEAH!!! Ha ha ha ha...
Once again, let's settle our issues in the courtroom and not in the marketplace. Perhaps this issue will finally come to some closure soon.
Please subscribe to see the more insightful version of th
What will be the ramifications if the GPL doesn't hold up in court? If the GPL hasn't been challenged in court thus far, this could have far reaching implications within the Linux community. If the GPL doesn't hold up, does that mean Microsoft is free to take large chunks of GPL'd software and make it proprietary?
NuSphere may not be 'obscure', but the mysql.org website seems a little deceptive. If I didn't investigate it more, I wouldn't have known the difference between mysql.org or mysql.com just by looking at the front page of the website.
.org websites should ever be for-profit businesses as that is not how that domain was intended to be used.
MySQL.org presents their website in a manner that does not give credit where credit is due.
For one, if you go to mysql.org you will find that it doesn't say whether it developed the software, it just says it's got mysql software available for 'free' download. But the mysql.com guy is correct, in that you cannot download any software without registering with mysql.org first. That indicates to me that it's not 'free' because now mysql.org has customer data to use to market their product to. Plus, it doesn't say: "Hey, we didn't write the software, mysql.com is where you can find that info. We just improved upon it." Besides, I don't think
I thought that by the definetion of FSF that SUN would be required to include the GPL in their SQL app. If they messed up on something like this then they deserve what ever they get.
Personally, I don't have much of a vested interest in who distributes MySQL, since I don't use it, and probably won't any time soon. However, as one of the first court cases to test the enforceability of the GPL, it think that it's critical that MySQL AB win. If they can't take on another small company over a GPL violation, how in the hell can we expect anyone to be able to stand up to a BigCo that decides to rip off their GPL'd code?
Sounds like Bush should add NuSphere to his "Axis of Evil" list.
I, for one, use MySQL all of the time, but had no idea about the history of it, or the fact that NuSphere didn't create it.
Pretty pathetic, NuSphere. I, for one, hope you get nailed in court, and all of your source code gets handed over to MySQL AB. You deserve it!
distribute your derivative work.
The GPL is clear on this point. If, for whatever reason a court of law decides the GPL is invalid in a particular case, then you lose any right to distribute derivative works.
Thus, there is no incentive for a company to try to invalidate the GPL, because without it they have no rights to distribute derived works. The only option a company has is to prove it is in compliance with the GPL.
The one interesting thing here I observe is that once you have been found in violation of the GPL only the holder of the copyright can reinstate your ability to distribute derivative works-- and there is no obligation to do so.
Thus, companies using GPLed code in works they distribute need to pay special attention to compliance as none-compliance carries with it the possibility being permanently blocked from using the GPLed code again.
Admitedly though, it's not complicated. Simply ensure you distribute your source along with the product, and all is well.
Otherwise put, if you want to use GPLed code, you have to share the code of your derivative work with everyone.
Do not spread "09 F9 11 02 9D 74 E3 5B D8 41 56 C5 63 56 88 C0" over the internet, thank you.
So the article said that NuSphere registered MySQL.org, and AFAIK the real MySQL web page is at MySQL.com.. but --
[xaxxon@chopper log]$ nslookup www.mysql.com
Server: localhost
Address: 127.0.0.1
Non-authoritative answer:
Name: www.mysql.com
Address: 64.28.67.70
[xaxxon@chopper log]$ nslookup www.mysql.org
Server: localhost
Address: 127.0.0.1
Non-authoritative answer:
Name: www.mysql.org
Address: 64.28.67.70
Did I mis-read?
Basically, since the GPL is the only document granting you permission to use the software, violating the GPL revokes your rights under it. That means that if it is found that they violated the GPL (which seems a foregone conclusion if the reporting is accurate), they will no longer be able to distributed MySQL code at all. In other words, put completely out of that business.
Now that is something other companies will take seriously in the future. (IANAL and all that of course.)
NuSphere is not at all a small company. It's GIANTLY HUGE ;-))
NuSphere is owned and financied by PROGRESS
http://www.PROGRESS.com/ which is really a
giant company like Oracle...
The Price of Freedom is Eternal Vigilance.
As usual, it is worth pointing out that if the GPL were ever found unenforceable, then you have NO right under copyright law and international treaty to use the code - it doesn't suddenly become public domain.
This is one of the reasons the GPL tends not to get challenged - it's a lose-lose situation for a GPL infringer who challenges it in court:
Scenario a: GPL challenge fails. Infringer has to GPL all derived work of original GPL code that he wants to distribute. Presumably he didn't want to do this, otherwise he wouldn't have taken it to court...
Scenario b: GPL challenge succeeds. Infringer has to stop distributing all derived work of original GPL code, as he now has no rights granted to him by the original copyright holder to use the code.
RMS has sued both parties for not calling the GNU derived application as GNU/MySQL.
"Free Software " declared RMS, "always starts with GNU".
EULAs are a strange beast. They are a non-negotiated contract made through click-through or breaking a seal for something that you purchased already. The theory behind a EULA is that you contract to the EULA. You do not need a license to run software. If you need to clikc Agree to use the software, have you enterred into a contract? That's an INTERESTING legal question.
GPL is MUCH less interesting. By default, you have NO right to distribute software. The GPL is a distribution license.
This Slashdot mental masturbation is childish. The odds of the GPL being overturned and everyone's software under license being made public domain is pretty close to 0%. It is only a concern on Slashdot.
The GPL hasn't been to court because every violator has reached a settlement.
This case sounds like NuSphere is fucked. The portion in question suggests that if you violate the terms of the license the license is voided. This is pretty standard stuff.
Here is the question that the court will answer.
If I break the GPL, I can be sued for damages, etc., and must stop distribution. My license is revoked, etc., etc. Can I then go out, download a fresh copy and distribute under the terms of the GPL? Stallman says no, I'm not certain. That's where this case is questionable.
However, this is a good test case for the GPL. The question of derivative work is interesting. I'm not certain that the linking scenario creates a derivative work. However, since this company distributed a modified MySQL with their additions, they are CLEARLY distributing the work.
They need to establish that they have a separate license or did so under the GPL.
Regardless, the GPL being invalidated would not make things Public Domain. Without license you cannot distribute, so if the license falls, no distribution under GPL v2. FSF releases GPL v2.1 within a week and any provision that includes (or later version) is fine, everyone else needs to update.
Alex
NuSphere is a totally independent subsidiary of the company with about 30 people.
Progress is just their VC.
This does not sound like GPL infringement on NuSphere's part. NuSphere says there are two components to Gemini : a Table Handler and a Storage Engine. The Table Handler source code has been GPL since NuSphere MySQL Advantage started shipping in April. Previously, Storage Engine was only available under a commercial license.
IANAL, etc. etc. ad infinitum, but if NuSphere's description of the product is correct, this does not sound like it violates the GPL in any way. As long as they release the code for the GPLed programs they should be in the clear. The GPL doesn't say you can't release software that works with the GPLed code, just that you can't release software that uses the GPLed code. That includes library calls, but does not include interaction between programs (e.g. through signals, semaphores, etc.).
Feel free to correct me if I am wrong (because if I am I'd like to know how it actually *does* work). As far as the domain name "hijack", unless MySQL actually tradmarked the MySQL name rather than just the logo, I think they have no ground to stand on. The article does not mention whether this is the case or not.
Wow, I guess it sounds like I'm flaming MySQL something heavy. I really have nothing against them, and think they make a great product; I just think they are most likely wrong in this case (assuming that the information I have is correct).
MySQL's time-to-market was better than PostgreSQL. People who have mysql working now are naturally hesitant to switch to something else. Secondarily, mysql (outside of Debian) was easier to install last I checked. This may have changed.
FSF has a press release and the affidavit we filed available on our website.
If the injunction is granted, it sets a strong precedent for certain parts of the GPL. If the injunction is denied, it may be for reasons other than any part of the GPL being invalid.
Because, this case will create precedent for all GPL'ed software, and whether or not corporations can steal the work of other people and call it their own! It's not a matter of which database software is better, but whether or not individual people should be able to create enforcable copyrights for their own software, and expect them to hold up in court.
If the GPL doesn't hold up in this case, expect a whole slew of proprietary packages to start popping up all over the place, each with a surprising resemblance to other, slightly inferior, but still groundbreaking "Libre" alternatives. It will suck the air out of all those alternatives, and once the alternatives are sucked dry, the Open Standards will go with them. (Remember -- Microsoft isn't the only company out there bent on World Domination through Embrase and Extend tactics.)
Your Servant, B. Baggins
It is also acceptable for the license to require that, if you have distributed a modified version and a previous developer asks for a copy of it, you must send one.
and also...
3. You may copy and distribute the Program (or a work based on it, under Section 2) in object code or executable form under the terms of Sections 1 and 2 above provided that you also do one of the following:
a) Accompany it with the complete corresponding machine-readable source code, which must be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange; or,
b) Accompany it with a written offer, valid for at least three years, to give any third party, for a charge no more than your cost of physically performing source distribution, a complete machine-readable copy of the corresponding source code, to be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange; or,
c) Accompany it with the information you received as to the offer to distribute corresponding source code. (This alternative is allowed only for noncommercial distribution and only if you received the program in object code or executable form with such an offer, in accord with Subsection b above.)
I don't see, anywhere in the liscense, a timeline specified as to when any changes to source code must be supplied. Anyone know of a quote from the GPL that specifies that the source must be made available at the time of binary release? AFAIK, they don't have to release the code to anyone who doesn't specifically request it, and who has a copy of the binaries; and I don't think time-frame is brought into it.
put the what in the where?
Alot has happened in the MySQL AB vs. Nusphere fight. First Nusphere took the mysql.org domain and acted as if they were the not-for-profit corp behind the software. MySQL AB was right to be frustrated about the situation. But for what we know, this question got resolved, as mysql.org now points to the original mysql.com site and WHOIS reports MySQL AB as the owner.
So where do we go from here? These guys believe that they can enforce article 4 of the GPL against Nusphere for an infraction to the GPL that occured in the past and which has been corrected since then as Nusphere published its source code. I personnaly believe this is not The Right Way. It totally goes against the spirit of the GPL: software must be free, and although we want to protect the original copyright holder, we must not discriminate against the users.
Many people could try to profit from the GPL in an illegal way, does it mean that we can put them on a blacklist? If they decided to change their attitude and in turn abide to the license, I don't see why we would still punish them. They can be a valuable ressource in improving the software, even though it clearly goes against MySQL AB's agenda.
Hey, if you're going to act that way, why did you release it under the GPL in the first place? By retaliating the way they do MySQL AB makes me wonder who between them and NuSphere has the weakest principles.
Using the GPL as a tool for revenge is definitely not what Stallman & followers originally intended.
Of course, MySQL AB could and should sue NuSphere for misrepresentation, moral copyright infrigment and dubious corporate conduct. They should seek damages, which they are fully entitled to under their country's Copyright Act. I for one find it unacceptable to simply lock Nusphere out.
It is sad to see that the first judicial test of the GPL is witness to such a hijacking. And the worst thing: the FSF has it's arms in the mud upto the shoulders.
This is one of these situations where Stallman actually should open his big mouth like he usually does.
I don't think that the judge will let that injonction go through. Their case is too weak to achieve that. Anyways in the worst case, what will prevent Nusphere from starting another company?
pertools will bring you the thruth.
That's why I think that the whole section is silly. You really can't restrict people from accessing the GPL this way.
I suppose they could file for an injuncton and argue that the company is a shell. Who the hell knows.
Regardless, I think that this Slashdot bullshit about GPL code becoming public domain is beyond silly.
Now, the outrageous claims of some proponents (anything that touches GPL code or is written by people that have looked at it becomes GPL) should get swatted down.
I also don't think that you can try to pull trade secret bullshit (prove you didn't see this) with GPLed code, but we'll see.
Alex
Basically, MySQL AB is arguing that because NuSphere violated the GPL *in the past*, they have forfeited their right to distribute a GPL app, even though the source code in question has since been released in compliance with the terms of the GPL. What are the ramifications if they succeed? Does this mean NuSphere will be unable to distribute any GPL applications? Will the NuSphere contributions be deemed invalid? Will MySQL AB then distribute MySQL with Gemini tables while NuSphere is not allowed to? (Gemini tables are a NuSphere contribution.)
Isn't it sadly ironic? The first time GPL is tested in court, and I hope they loose, for all our sakes. No matter which way the decision goes, the dispute hurts the entire free software community, and will make corporate contributions to code base that much harder to come by. And think of what Craig Mundie from Microsoft will have to say about this! All over what was originally a domain name dispute. (If your new to this conflict, NuSphere paid MySQL a huge chunk of change for the right to distribute MySQL, (that's right, paid for the right to distribute a GPL app.). In the process, they created a mysql.org web page that did not give any credit to MySQL AB, and indeed, only made SQL Source Code available to registered users. MySQL AB charged NuSphere with GPL violation. In response, NuSphere made the source code more available and released the source code for their own proprietary modules. It is still not clear to me, however, what exactly MySQL AB thinks the millions they were paid by NuSphere was supposed to be for. Somehow, the right to distribute a GPL app for a limited time just doesn't make sense.)
MySQL AB, I thank you very much for your hard work and development of MySQL. But please, stop pissing on the GPL to advance your own agenda. You can debate whether the money you were paid gave NuSphere the right to hijack MySQL trademark until the cows come home; but all GPL issues have been put to bed months ago!
It is ironic that a license that is supposed to advocate "freedom" in the development world can be so repressive.
They released the source code so that should be the end of the story. Does it mater when and how? If I play around with the Linux kernel and add some code, are the FSF Gestapo going to come crashing down my door demanding me to release all the code? You call that "freedom" (by ANY definition)? No way, maybe in Cuba! Screw that... that's why BSD rocks because i can do what i want with the code. I don't see how GPL is different from a commercial license because, in the end, you have to answer to an organisation that dictates the "rules".
The question here is whether if party A has distributed party B's source-code without party B's permission, party B can deny party A permission to distribute from then on, no matter what party A does. This is a subtle point. (Item 4 in the GPL applies.) For an example, this is the exact situation that KDE wound up on the wrong side of which RMS pointed out in the "forgiveness" speech. (A speech that ignorant morons all over flamed him about, but that is another story.) Having once not been in compliance with the GPL, you need explicit forgiveness from the copyright holder to regain your rights under it later.
If the GPL wins, it demonstrates that not only does the GPL hold, but a very strict interpretation of the GPL holds which gives copyright holders the power to bargain with the question of whether forgiveness will ever be granted. (This forgiveness is customarily granted by the FSF.)
If the GPL loses, it only casts doubt on whether the copyright holder needs to give forgiveness. There is no question of whether it is legal to distribute in violation of the GPL's terms. (The injunction seeks to stop a distribution that includes source-code.)
IANAL, but I suspect that the FSF will have an easy decision here. But the cynical bastard in me notes that other recipients of the forked code have rights to distribute the whole thing under the GPL. If one of them distributes back to the original creator, under item 6 of the GPL the creator now gets a fresh license. What if the creator then distributes using that brand spanking new license?
Get sued random small ass person who has code under the GPL. Get the GPL invalidated. Shut down all the millions of free software projects that make use of the GPL. Hell, forget Microsoft, how do we know that Theo da Raadt or some other cat stroking evil genius isn't gunna do it (no offense Theo). If the GPL is declared invalid then surely you have to be wrong about this whole "no rights to distribute at all" stuff. Surely 99% of people using the GPL would immediately issue statements saying in essence that they are switching to a BSD license. Ahhh, fun with copyright law.
How we know is more important than what we know.
Reading the FSF Affitdavit addressed some of my concerns. In particular, #25:
FSF's policy with respect to GPL violations is to secure compliance, not damages. When a party has violated GPL, and the violation is called to our attention (which happens on the average some dozens of times each year), we inform the party in violation of its responsibilities, and advise it on the steps necessary to come into compliance. It is our practice that once a party has taken steps to comply, and has entered into confidence-building measures to ensure that future non-compliance will be avoided wherever possible, and rapidly discovered and remedied where inadvertently reproduced, distribution rights under GPL 4 are restored on a cooperative non-judicial basis. In this fashion, I have secured compliance with the license in dozens of cases over the past decade, and have never had to resort to judicial measures of mandatory enforcement. Without the leverage provided by 4, however, parties would resort to repetitive partial compliance, ``capable of repetition but evading review,'' in language the Supreme Court has applied to a different sort of situation, substantially if not overwhelmingly complicating the task of securing reliable compliance with the license.
In this case, since the FSF does not own copyright to MySQL, their usual policy doesn't apply. I still think what MySQL AB is doing is wrong, and is abusing the spirit of GPL, however. Just as importantly, using such a convoluted case to test the GPL in court for the first time seriously jeopardizes the license. (If it is not upheld in this case because of the strange circumstances, where will that leave FSF in enforcing GPL in other, more clear cut instances?)
In a normal contract... there is a point where the contract an agreement is reached, the contract is signed. If a clause in the contract causes the contract to be terminated, it's null & void.
But you see, with the GPL, there is no negotiation. What prevetns someone from re-licencing the software under the GPL again? Nothing. As long as they are currently complying with the terms, they can keep using it.
Why is the GPL different? Because nobody CHOOSES to let someone license it.
... now I don't have to read endless postings of "Nobody ever sued over the GPL" ;)
Pls. note: I'm fully aware of how important any challenge to the GPL could be, particularly the first one; my prediction is not overlooking that fact, but merely stating that I think that it won't come down to a cut and dried decision.
everyone here is saying that this could be the first real legal challenge of the gpl.
which makes me wonder: what ever happened with carmack and the guy (slade?) that was trying to redistribute quake binaries? for all the ranting that went on here about that i never did hear about how that ended up. anybody here know?
If I don't put anything here, will anyone recognize me anymore?
They could acquire any trademarks and other IP held by nusphere, and be back in business in a couple days.
Seems like a minor hassle.
The underlying problem with the GPL remains: you can't keep a competitor from using your stuff.
What does this "written offer valid for any third party" mean? Does that mean everyone in the world can get the source to any GPL'ed program no matter what?
"Valid for any third party" means that anyone who has the offer is entitled to take you up on it.
If you commercially distribute binaries not accompanied with source code, the GPL says you must provide a written offer to distribute the source code later. When users non-commercially redistribute the binaries they received from you, they must pass along a copy of this written offer. This means that people who did not get the binaries directly from you can still receive copies of the source code, along with the written offer.
The reason we require the offer to be valid for any third party is so that people who receive the binaries indirectly in that way can order the source code from you.
If I distribute binaries without sources, can I provide source code by FTP instead of by mail order?
You're supposed to provide the source code on a physical medium, if someone orders it. This means sending it by mail order.
You are welcome to offer people a way to copy the corresponding source code by FTP, in addition to the mail-order option. FTP access may be convenient for the users, or it may not. If the FTP access is convenient enough, perhaps no one will choose to order a copy, and you will never have to ship one. Good for you. But if a user chooses to order a copy of the source, you must fill the order.
The early version of NuSphere was clearly in violation of the GPL and thus _de facto_ renounced their rights to redistribute that version of MySQL. However, it seems that they've cleaned up their act for the next version of NuSphere.
What I'm wondering is if that revocation is permanent, absent forgiveness by the licensor of the MySQL code? It will obviously be so for that particular version of MySQL. But what about later versions of MySQL? New code, new license. Does their initial unrepentent infringment of the GPL mean that they are forever barred from redistributing newer versions of MySQL whose license they haven't violated?
All versions and variants of the BSD license require that credit is given to the copyright holder. The only difference with the "new" version is that credit no longer has to be given in advertising material that mentions use or features of the code.
Credit still has to be given in all documentation provided with the software, as well as within the source code. "Give credit" in this case really means that the whole software license, including a disclaimer, has to be reproduced with every copy.
All works are automatically copyrighted. I may give someone a limited right to distribute my work, tied to a number of conditions. Breaking these conditions means you no longer have that right. The issue then boils down to standard copyright law, under which you are absolutely denied any right to distribute the work beyond Fair Use.
In short: Accept the GPL and enjoy the right to distribute my code, or accept plain copyright law under which you don't have that right. There are no other alternatives.
I don't see, anywhere in the liscense, a timeline specified as to when any changes to source code must be supplied.
The timeline is very clear and obvious, at least to anyone who can read english, namely:
You have to make the source code available the moment you begin distributing the GPLed software.
Even if you choose option (b) or (c), the source code has to be available, and provided on request, the moment you begin distributing your derivative work.
Where is the rocket science in this?
The Future of Human Evolution: Autonomy
All that NuSphere needs to do is remove all the MySQL code from their codebase, stop distributing any other version of MySQL, and then download the code from mysql.com and put it back in their codebase.
viola! instant fix. Why does this work? Because by purging MySQL they are make themselves in compliance with the GPL (they aren't distributing MySQL), then when they download it again, MySQL AB is distributing it to them under the GPL. So, they get a second chance. It is true that MySQL AB can refuse to license MySQL to NuSphere, but they can't distribute it to them (which implies granting a license) at the same time they are denying a license.
~smm
This suggests that free software that uses non-free libraries can be distributed commercially, but must include a written offer valid for any third party . Furthermore, any program which makes a system call to run separate programs which are released under the GPL, with a special exception which allows linking proprietary modules under a controlled interface only must be made available with its complete source code under the GNU GPL. Furthermore, a library covered by the GNU LGPL can be distributed commercially, but must include a written offer valid for any third party . On the other hand, a library covered by the GNU GPL can be modified and used internally without releasing it .The GPL is a free software license, and therefore any program which is merely aggregated to separate programs which are released under the GPL can be modified and redistributed under the GNU LGPL . Suppose, for instance, that binaries you distribute for download is covered by the GNU GPL According to section 2 of the GPL, any program which makes a system call to run separate programs which are released under a proprietary license is covered by the GNU GPL . Notice, incidentally, that any program which uses a library which is released under the GPL can be distributed commercially, but must include a written offer valid for any third party For one thing, a GPL-covered program can be modified and used internally without releasing it . This suggests that a library covered by the GNU GPL must be provided on a physical medium On the other hand, modified versions you distribute can be distributed commercially, but must include a written offer valid for any third party . Thus, any program which makes a system call to run separate programs which are released under a proprietary license can be modified and redistributed under the GNU LGPL According to section 3 of the GPL, a program released under the GPL can be distributed non-commercially, but must include a written offer valid for any third party . On the other hand, any program which makes a system call to run separate programs which are released under the GPL, with a special exception which allows linking proprietary modules under a controlled interface only must be made available with its complete source code under the GNU LGPL. For one thing, any program written for a programming language interpreter which is released under the GPL can be modified and redistributed under the GNU LGPL . Of course, a non-free program can be distributed non-commercially, but must include a written offer valid for any third party The GPL says that any program which makes a system call to run separate programs which are released under the GPL, with a special exception which allows linking proprietary modules under a controlled interface only can be modified and used internally without releasing it . On the other hand, any program which is sold commercially under the GNU GPL can be modified and redistributed under the GNU LGPL Thus, a library covered by a special exception of the GNU GPL is covered by the GNU LGPL . However, this assumption is not correct, since a library covered by the GNU GPL must be provided on a physical medium So far, any program which uses fork() and exec() to invoke separate programs which are released under the GPL, with a special exception which allows linking proprietary modules under a controlled interface only can be modified and used internally without releasing it . According to section 1 of the GPL, any program which makes a system call to run separate programs which are released under the GPL, with a special exception which allows linking proprietary modules under a controlled interface only can be modified and redistributed under the GNU GPL The GPL is a free software license, and therefore any program which uses a library which is released under the GPL must be provided on a physical medium . Analogously, any program which uses fork() and exec() to invoke separate programs which are released under a GPL-compatible license must be made available with its complete source code under the GNU LGPL. Note that any program which uses fork() and exec() to invoke separate programs which are released under a GPL-compatible license can be distributed commercially, but must include a written offer valid for any third party . We have already seen that any program which uses fork() and exec() to invoke separate programs which are released under a proprietary license must be made available with its complete source code under the GNU GPL. According to section 1 of the GPL, binaries you distribute for download must be made available with its complete source code under the GNU LGPL. . Clearly, any program which dinamicly links to a library which is released under the GPL can be distributed non-commercially, but must include a written offer valid for any third party It appears that any program which is merely aggregated to separate programs which are released under a proprietary license must be licensed to all third parties under the GPL . This suggests that any program which uses a library which is released under the LGPL must be made available with its complete source code under the GNU LGPL. Nevertheless, the source code for all derivative works can be modified and redistributed under the GNU LGPL . Suppose, for instance, that any program which is merely aggregated to separate programs which are released under the GPL can be distributed non-commercially, but must include a written offer valid for any third party Let us continue to suppose that any program which makes a system call to run separate programs which are released under a proprietary license can be modified and redistributed under the GNU GPL . Furthermore, any program which makes a system call to run separate programs which are released under the GPL must be provided on a physical medium
The reason Eben Moglen has gotten dozens of companies to give up and to submit is because of section 4. Without it, we'd have a lot less free software than we do now. In the past, the threats of nuclear war have been private, but very serious (if you're in the Linux business and lose your right to distribute, say, glibc, you're dead meat).
It's important for everyone to understand that if you violate the GPL, it's not sufficient to just stop violating, you need to get the copyright holder's explicit permission before you can ever start copying, modifying, or distributing the program whose copyright you violated ever again. People got pissed off when RMS talked about "forgiving" the KDE project, but too many people don't realize that from a legal standpoint this forgiveness was required (though evidently only a couple of less-important KDE applications ever had any FSF-owned GPL code in them). Certainly RMS could have been more diplomatic (though maybe not, it isn't one of his talents).
Don't buy any more software from them. That'll show the bastards! Who needs an ACID-compliant, fine-grained access, modern RDBMS anyway? We can live without its beautiful subqueries, strong transactions, multi-language stored procedures, so let's show them!
You can't change the licence, but you can charge *anything* (no, not equal to cost) for getting it from you. Translation: I can charge you $1000 to download MySQL from me, but you're free to distribute it at no cost. Read the licence, it's all in there.
Kjella
Live today, because you never know what tomorrow brings
See, that's where this whole thing seems absurd. In fact, I would argue that I can agree to a new license for the same version.
When I negotiate contracts, we go through multiple revisions. Then we sign something. We could always amend the contract (sign a new one that says how we are changing it) and most have a termination clause.
If I terminate my contract with entity X, we can sign a new contract later. Situations change.
In this case, you have put in an unsigned license that says this licesne is available universally. It allows me to license the code from you to distribute and license others to redistribute.
I broke a section, therefore your termination clause automatically kicks in. The old license is gone. Oh no, I'm without license so I go to get a fresh license from you, and low and behold, there is one with the same terms. I agree this time and I am licensed.
Now, the GPL could stipulate that you lose your rights to all other applications under the GPL licensed from that entity. As the copyright holder, that is okay. Okay, no problem, I have myNewShellCompany download the software, then send me a copy. My company has now licensed it from myNewShellCompany, and your restriction on my agreeing to a fresh license from you goes away.
I don't know, I could see this restriction being less powerful than MySQL AB and the FSF want it to be. Who knows, maybe the courts will enforce it reasonably, and say that it is reasonable to say no more distribution for you if break the GPL.
It makes sense that you can get an injunction and sue for damages if someone ships your GPL code in violation of the license, but I'm not sure that you can prevent them from shipping it within the context of the license given that you have a universal license grant.
Alex
I don't get it. Why is this going to court? The parties worked out their differences, but then MySQL decided to be petty and "punish" NuSphere. This is wrong. I totally support the GPL, and if NuSphere was brought into compliance, why would MySQL not allow them to continue distribution of compliant software? Other than hellishness, I see nothing to be gained. Can't we all just get along?
The problem with what RMS did wasn't what he did legally, we all recognized that. It was his being a jerk about the entire situation.
The KDE Team felt they were within the bounds of the law, FSF felt otherwise. Either there was no FSF code involved or they felt that their case was week, so they focused on complaining and launching a competing project.
When the FSF and Trolltech worked out their differences regarding Qt licensing, RMS issued a statement applauding the change, forgiving KDE and it's users, and cheering on GNOME. Once Qt went GPL, there is no reason for the FSF to support GNOME (which sits on top of libraries with the "bad don't except under special circumstances" LGPL license ) over KDE except for NIH.
RMS handled it with less tact than he normally uses, that is what pissed everybody off. The "forgiveness" could have been done in nice legalese on their website without trying to get it coverage.
Alex
Seems to me that Nusphere is trying to greedily free-ride off the hard work of Open Source developers. Simply put: Nusphere doesn't get it. They are one of many who still mistakenly believe that proprietary software is the only way to make money. As such, I personally hope they get ripped to shreds in court and all their proprietary modifications get forcefully released to be freely assimilated into the GPL codebase as MySQL developers see fit.
I disagree with your interpretation, specifically "that is: it must be supplied together with the binaries at the same time. If the software is distributed without source, then it is in violation of the license agreement." I disagree because of the written offer, which doesn't specify a timeline. Under close scrutiny I'm not so sure the GPL (in which I've placed much faith[and I emphasize faith, as it still has not had its day of scrutiny in court]) will hold up, almost specifically because of the timeline issue. Also the fact that if no GPL version is specified, the holder of the source/binary can use ANY of the GPLs, at his/her discretion. That could very well make everything GPLd to this point moot.
put the what in the where?
Please note, this is the first time I've ever posted to slashdot, hoping this will be a new thread, apols if I've fecked up...
I used to work for a UK interactive tv company "SmashedAtom", we ceased operation due to lack of funding in Novemeber 2001.
During the 8 months I worked for them, we did a lot of work on the OpenTV platform (http://www.opentv.com), who are the basis of the Sky digital platform in the UK, and no doubt lots of others worldwide.
Now, when you purchase (you certainly can't get anything for free) the OpenTV SDK one of the things they supply is a modified version of gcc to compile to what I believe is their own proprietry "java" like byte code, that is interpreted by their set-top boxes.
We asked many times for the source code, believing we had a right to it under the GPL, but
all requests were stalled or ignored. Either the person on the phone didn't understand what we were on about, or no replies were ever made to emails.
Should their modified gcc be available? Pretty academic to me right now, but since they charge
thousands of pounds sterling for what amounts to
a version of gcc...
Regards,
Neil.
neil@neilpearce.com
Slashdot has never intented to be commercial, but now they have the banner ads, which is a necessary evil. However, it's not like they are selling anything directly.
One thing I can't stand is people who buy out all three domains for a name, just because. Now, with all of the extension choices, it's not as common a practice, but it still happens often.
Zodiac Survey
This suggests that free software that uses non-free libraries can be distributed commercially, but must include a written offer valid for any third party . Furthermore, any program which makes a system call to run separate programs which are released under the GPL, with a special exception which allows linking proprietary modules under a controlled interface only must be made available with its complete source code under the GNU GPL. Furthermore, a library covered by the GNU LGPL can be distributed commercially, but must include a written offer valid for any third party . On the other hand, a library covered by the GNU GPL can be modified and used internally without releasing it .The GPL is a free software license, and therefore any program which is merely aggregated to separate programs which are released under the GPL can be modified and redistributed under the GNU LGPL . Suppose, for instance, that binaries you distribute for download is covered by the GNU GPL According to section 2 of the GPL, any program which makes a system call to run separate programs which are released under a proprietary license is covered by the GNU GPL . Notice, incidentally, that any program which uses a library which is released under the GPL can be distributed commercially, but must include a written offer valid for any third party For one thing, a GPL-covered program can be modified and used internally without releasing it . This suggests that a library covered by the GNU GPL must be provided on a physical medium On the other hand, modified versions you distribute can be distributed commercially, but must include a written offer valid for any third party . Thus, any program which makes a system call to run separate programs which are released under a proprietary license can be modified and redistributed under the GNU LGPL According to section 3 of the GPL, a program released under the GPL can be distributed non-commercially, but must include a written offer valid for any third party . On the other hand, any program which makes a system call to run separate programs which are released under the GPL, with a special exception which allows linking proprietary modules under a controlled interface only must be made available with its complete source code under the GNU LGPL. For one thing, any program written for a programming language interpreter which is released under the GPL can be modified and redistributed under the GNU LGPL . Of course, a non-free program can be distributed non-commercially, but must include a written offer valid for any third party The GPL says that any program which makes a system call to run separate programs which are released under the GPL, with a special exception which allows linking proprietary modules under a controlled interface only can be modified and used internally without releasing it . On the other hand, any program which is sold commercially under the GNU GPL can be modified and redistributed under the GNU LGPL Thus, a library covered by a special exception of the GNU GPL is covered by the GNU LGPL . However, this assumption is not correct, since a library covered by the GNU GPL must be provided on a physical medium So far, any program which uses fork() and exec() to invoke separate programs which are released under the GPL, with a special exception which allows linking proprietary modules under a controlled interface only can be modified and used internally without releasing it . According to section 1 of the GPL, any program which makes a system call to run separate programs which are released under the GPL, with a special exception which allows linking proprietary modules under a controlled interface only can be modified and redistributed under the GNU GPL The GPL is a free software license, and therefore any program which uses a library which is released under the GPL must be provided on a physical medium . Analogously, any program which uses fork() and exec() to invoke separate programs which are released under a GPL-compatible license must be made available with its complete source code under the GNU LGPL. Note that any program which uses fork() and exec() to invoke separate programs which are released under a GPL-compatible license can be distributed commercially, but must include a written offer valid for any third party . We have already seen that any program which uses fork() and exec() to invoke separate programs which are released under a proprietary license must be made available with its complete source code under the GNU GPL. According to section 1 of the GPL, binaries you distribute for download must be made available with its complete source code under the GNU LGPL. . Clearly, any program which dinamicly links to a library which is released under the GPL can be distributed non-commercially, but must include a written offer valid for any third party It appears that any program which is merely aggregated to separate programs which are released under a proprietary license must be licensed to all third parties under the GPL . This suggests that any program which uses a library which is released under the LGPL must be made available with its complete source code under the GNU LGPL. Nevertheless, the source code for all derivative works can be modified and redistributed under the GNU LGPL . Suppose, for instance, that any program which is merely aggregated to separate programs which are released under the GPL can be distributed non-commercially, but must include a written offer valid for any third party Let us continue to suppose that any program which makes a system call to run separate programs which are released under a proprietary license can be modified and redistributed under the GNU GPL . Furthermore, any program which makes a system call to run separate programs which are released under the GPL must be provided on a physical medium
1. You do not question the GPL.
2. You do...NOT QUESTION THE GPL!!
3. Don't fuck with the FSF, especially their lawyers.
Zodiac Survey
That is just one of many interpretations of Article 4. The GPL text doesn't explicitly say anything about permanent revocation, and it doesn't say anything about requiring explicit permission from the copyright holder to do anything.
Article 4 is only clear about terminating your license, it doesn't say anything about preventing you from re-licensing another copy. According to the precedents in software contract law, unless otherwise specified, the term of the contract lasts from when you obtain the copy until you get rid of it or the copyright expires. That means it should be legal to simply download a new copy of the software (in the process re-accepting the license terms under a new contract) and start over.
I think you're right that the GPL will be cast aside as a worthless document.
Hopefully the court finds that the mere act of publishing the source code with the intent of sharing is equivalent to publishing to the public domain and any restrictions upon public domain information are unbinding.
This could be the end of the GPL. Thank god.
You do not need a license to run software
This is a common bit of Open Source mythology that was at one time a reasonable approximation of the truth. But it simply isn't the law. Whatever may be the way we wish it were, the cases do not support this proposition.
In fact, the cases support the contrary proposition: that running software typically entails loading content from some medium onto RAM, and thereby constitutes a "reproduction" under Section 106 of the Copyright Act. Thus, without a license, the owner of the copyright has an exclusive right to preclude a user from running the software.
Not that this result isn't controversial -- academics have assailed these cases in law reviews, and the Congress recently limited their applicability in certain cases that are not relevant here (embedded operating system software in connection with purchase of hardware).
But controversy doesn't change the law. While the previous author clearly spouted the "ole FSF party line," that party line isn't consistent with the Copyright Act and corresponding case law.
If it did, then the GPL isn't an issue here at all, because they can claim that their right to distribute a derived work comes from their contract, not the GPL. That's what dual licensing is all about. Then again, IANAL.
Revenge is a dish best served cold -- grits should be served hot!
Nusphere should continue to be able to release their software in compliance with the GPL, however for each copy of their software they distributed without complying with the GPL, they should be forced to negotiate with MYSQL AB etc. and agree on a fee to be paid for each copy of the software distributed illegally.
If precise figures cannot be had, then an estimate should be created based on the length of time NuSphere was distributing product that did not comply with the GPL.
A price per copy should be constructed based on similarly featured products in the marketplace. It could get expensive for NuSphere, but commercial software is an expensive commodity.
As MySQL is open-source, it may make more sense to have NuSphere donate the money to an organisation like the EFF or the FSF, to support Open Source software as a whole.
I gots ta ding a ding dang my dang a long ling long
Let's assume GPL is deemed invalid:
The work is still copyrighted by the owner. If GPL was invalidated, you would need to negotiate an agreement with the relevant owner(s) if you wanted to use it. With any amount of collaborative input ("we used the libraries from project X and the drivers from project Y which was derived from project Z") this would be a real minefield that could see you sued further down the track (ala GIF). In the meantime, a new GPL would be circulated and pretty much overnight, most people would be distributing under "son of GPL".
Me thinking out loud:
I think it would take a ruling that took copyright away from owners to compromise things. Imagine a "Fair and Reasonable Corporate Access to Copyrighted Materials" Act that watered down copright for individuals. Say under the excuse that copyright shouldn't be used to protect DCMA circumvention and such copyrights are therefore forfeit.
Xix.
"Everything is adjustable, provided you have the right tools"
Sure, if I redistribute GPL'd software in violation of the license, the license terminates. But I don't see where it says that I can't re-license the software.
So, according to the plaintiffs, if I violate the GPL (say, by forgetting to date one of my changes, as required by 2a), I can never distribute the program again?
If that's the legal interpretation, then the GPL needs to be fixed.
"It sure was strange to see something on Usenet about me that didn't involve Klingon gang rape." -- Wil Wheaton
Obtaining the code from a different source wouldn't help, as your right to redistribute has been withdrawn by the copyright holder, and applies over all copies.
Yes, but according to section 6 of the GPL, you automatically recieve a license from the original licensor each time you recieve "the Program".
We have said earlier that we welcomed the opensourcing of Gemini when it finally happened. But NuSphere still denies having violated the GPL in the first place, which, if left at that, may set a damaging precedent and is a potential threat to ANYONE developing GPL'd software. I believe it is in everyone's interest that the GPL is a defendable licensing model.
And, for those who may not know, we at MySQL AB have made numerous attempts to settle the case out of court.
Marten Mickos, MySQL AB
What's clear is that NuSphere violated the GPL and lost their rights as per article 4. What's not clear is whether (and why) article 6 no longer applies. Article 6 would seem to allow them to regain their rights by receiving a new copy of the code. The copyright holder's explicit permission would seem to be automatically granted by article 6!
I would further note that article 4 speaks of an action ("any attempt to" blah, blah) which results in automatic termination of the license. I think it's hard to argue that in the absence of an action (an actual or attempted violation), any termination of a new license obtained via article 6 would occur.
But what the heck, I'm not a lawyer or a judge, and your opinion, my opinion, and even Mr. Moglen's opinion won't matter a hill of beans if the judge disagrees.
GPL has never been tested - lots of people feel like it could be challenged. There is nothing else like it - how can you be so sure - friends on the Supreme Court?
Admit it - you are just voicing an opinion based on no real legal understanding of these issues - just layman's logic.
See comments at the end of FEDERAL APPEALS COURT HOLDS SHRINK WRAP LICENSE UNENFORCEABLE on this topic, written by someone who doesn't have to add "IANAL". :-)
That article was written in 1992, so it's very out of date. In the case of a warrentee disclaimer, a different set of notification laws apply. "Service sold separately" is a completely reasonable point of view, given proper notification. Most EULAs attempt to use "notification" as an excuse to take away the user's rights. Only a contract can do that.
It's important to know that all of the caselaw on this subject makes it very important to look at the exact circumstances. For example, if you are installing from one disk onto more than one computer, then you DO need a licence and in that case it is sort of like the GPL: if you don't accept the EULA then nothing else grants you multiple machine use. The "nothing else" part fails in the case of single machine use because there is something else: 17 USC 117, which is not subject to approval by the vendor.
As the owner of a copyright, I can license it however I want. If the GPL fails, then nobody can distribute it. Alternatively, the courts can rule that the GPL allows more than we think it does. They may rule that derivative works aren't as powerful a concept (any linking?) as the FSF would like. These are all possibilities.
However, the right of a copyright owner to license distribution is NOT questionable. If the GPL fails, then anyone distributing is at the mercy of the copyright owners. The GPL doesn't protect the owner (copyright protects the owners) the GPL protects the distributor that licensed it.
Section 4 is a strange beast, that is the crux of this case. If Section 4 isn't legal, the the FSF loses its stick.
This entire case revolves around Section 4, NOT the concept of licensing copyrighted work.
Read the people that think it could be challenged, they have bizarre views that the lawyers on Slashdot laugh at.
There are questions about the GPL, but the onces that Slashdot's laymans voice aren't the real ones.
IBM is putting $1B behind Linux development... Their legal team has likely gone over the GPL. Sure there are questions that only a court can determine (namely, what constitutes a derivative work and is Section 4 legal), but this Slashdot hand wringing is rediculous.
It's popular here to bash the US and its courts... just because ACs get modded to +5 by calling the US government corrupt doesn't make it so.
Alex
> So, according to the plaintiffs, if I violate the GPL (say, by forgetting to date one of my changes, as required by 2a), I can never distribute the program again?
No one is free to distribute software that doesn't belong to them. So you can't distribute GPL'd software without the copywrite owner's permission. However, you *can* distribute software that does belong to you and is not covered by the GPL.
MySQL AB holds the copyright and the trademark to MySQL. They are a business trying to make money just like NuSphere. NuSphere made a mistake distributing the binaries without the available source, but they fixed that and are still being penalized. It seems to me like NuSphere is being punished by a change in strategy at MySQL AB.
I just returned from the hearing for a preliminary injunction on the
NuSphere/MySQL case. Here are some initial reactions, many of which are not
original to me. This is fairly long, and was also posted to the fsl-discuss list.
First, the headline: there was nothing to suggest that the GPL itself or
free software is in danger, although the Judge is unlikely to rule in
MySQL's favor on this preliminary injunction.
The details: Judge Saris was surprised by the number of people attending the
hearing (around 20-30). I suspect she didn't know this was being seen as
the first test case of the GPL in court. She threw both sides off balance
by announcing, before any arguments, that she had more or less made up her
mind.
She said she was inclined to grant the defendant's (MySQL AB) motion for a
preliminary injunction with respect to the trademark issue (enjoining
Nusphere from using MySQL's registered trademark against MySQL's wishes
after their 'temporary agreement' broke down) but was not inclined to grant
the preliminary injunction against NuSphere from using the GPL'ed code at
all. There followed about two hours of arguments, and when the Judge
adjourned the hearing it seems she held the same views.
The Judge did not want to get into issues of fact beyond their most
minimally necessary. Although her technical knowledge was limited, she
realized that it could take several hours and probably several days to fully
explore the technical side of things, and she wanted to make her decision
today after this hearing, particularly since people had traveled great
distances to attend the hearing (including the two MySQL AB developers from
Sweden).
Like most Judges, Judge Saris was hoping this could be settled between the
parties, particularly since the Court is not particularly well equipped to
deal with the more technical issues of 'statically and dynamically
linked code', etc.. She urged the parties to use the remainder of their
time in the courtroom to talk, and was trying to set up some mediation with
a professional mediator or magistrate judge. One analysis is that she
will grant one motion but not the other so as not to put either party in too
powerful a position in an out-of-court settlement.
Although the Court clearly didn't understand the specifics of open source
(she seemed to think 'linking' code was analogous to hyperlinks on the web,
and that a single package was like a single icon you would click on on the
desktop), she did pretty clearly understand what the GPL is about, and
said nothing to suggest she didn't think it would be enforceable. The issue
at hand was whether the facts required immediate injunctive relief. For
preliminary injunctive relief, the Judge looks at:
(1) likelihood of prevailing on the merits
(2) immediate and irreperable injury, loss or damage
(3) balance of harm to moving party vs. harm to non-moving party lies in
moving party favor
It seemed like she might be convinced on (1), but she had trouble seeing how
(2) was the case, so she was not particularly interested in hearing a lot of
argument about whether or not the case would eventually prevail on the
merits. She also was concerned that the injunction from using MySQL code at
all would essentially destroy NuSphere's entire business (despite the fact
that they are a subsidiary of a $300M company), and thus found it hard to
believe that the balance of harms would lie in MySQL AB's favor.
It is 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 (Eben Moglen was present but was not given a
chance to testify). Instead, I expect she will focus on the lack of proof
of irreperable harm in a short opinion denying the motion.
The trademark issue was more clear cut. Under Copyright Law, non-permitted
use of a trademark is presumptively irreperable harm. MySQL AB and NuSphere
had previouly had a provisionaly contract granting NuSphere the right to use
MySQL's trademark; however, the Court found it convincing that that contract
terminated (1) because it had a horizon date, after which further
arrangements would need to be made, and (2) because NuSphere ceased making
payments to MySQL AB. NuSphere wanted to argue that MySQL AB should be
estopped from the claim because they took 18 months to bring suit about it
(therefore, in NuSphere's view, 'acquiescing' to the continuation of the
Contract) but MySQL AB was in good faith negotiations for most of that time;
once things did break down and MySQL AB learned that NuSphere had
distributed MySQL along with proprietary software, then they did file suit
fairly promptly.
I'd be happy to answer any questions about the proceedings for those who
couldn't make it.
A couple of thoughts:
(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.
(2) The GPL termination clause (section 4) is going to cause problems. Does
the person who violates the GPL permanently lose their license until the
licensor regrants it, even if they bring their use back into compliance with
the GPL? Can they not redownload the GPL'ed software and become a 'fresh'
licensee? What if the original licensor can't be located? Does this mean
there is a 'special class' of people in the world who can't distribute GPL
software for the rest of their lives because of some past action that has
been cured? How are courts likely to view this interpretation? Is it good
policy? These issues were quite present because it appears that NuSphere
has substantially cured their prior violation (although there was a dispute
of fact over whether they actually had fully cured it).
That's it for now. The parties were given until August for discovery,
although the judge urged them again to settle before that. It appears that
a preliminary injunction against NuSphere to stop them from using the
trademark (they will probably change the software name to something like
'NuSphere Enhanced for MySQL') will likely give them something on the order
of 3 months to implement.
---
Adam Kessel (adam@bostoncoop.net)
Who cares? Use a real database...
We have more to fear from the bungling of the incompetent than from the machinations of the wicked.