Slashdot Mirror
MySQL Changes License To Avoid GPLv3
munchola writes "MySQL has quietly changed the license it uses for its database to avoid being forced to move to the forthcoming GPLv3. CBRonline is reporting that Kaj Arno, MySQL VP of community relations, revealed the license change on his blog, noting it was made 'in order to make it an option, not an obligation for the company to move to GPLv3.'"
Why not just stay with GPLV2?
They did. They set their license to be GPLv2 only.
Given the GPLv3 (however it's written) cannot reduce the liberty provided by the GPLv2, one can only imagine that those who cross out "or later" are fearful that the public may be given more liberty by the GPLv3 than they'd prefer them to have.
What kind of liberties are there to be afraid of?
The GPLv3 can only restore the liberties suspended by copyright, patents, and the DMCA. It cannot grant any additional liberties, e.g. to inspect the publisher's premises, or to sleep with its CEO's daughter, etc.
-- Ed Avis ed@membled.com
From GPLv3 draft 2: "(For instance, if the work is a DVD player and can play certain DVDs, it must be possible for modified versions to play those DVDs. If the work communicates with an online service, it must be possible for modified versions to communicate with the same online service in the same way such that the service cannot distinguish.)"
bad idea. this means when BluRay comes out, someone cannot take existing DVD-based GPL code and mod it to work with BluRay if it won't also work with DVD.
"This License permits you to make and run privately modified versions of the Program, or have others make and run them on your behalf. However, this permission terminates, as to all such versions, if you bring suit against anyone for patent infringement of any of your essential patent claims in any such version, for making, using, selling or otherwise conveying a work based on the Program in compliance with this License."
i never like the wording of this one this one. (a) "if you bring suit" is not the same as "if you win the suit". and (b) why would a patent suit end said permission? if the patent suit is in one country, development could then just move to another country under which the original patent has no jurisdiction. basically, i'm not in agreement with patent suits winning in ANY case where the patent owner has only created a design but no implementation. if the inventor of an idea does not plan on implementation him/herself in a reasonable time, it is HINDERING THE WORLD to let them have a patent, because all they have done is lock up an idea from being developed. if they want to start a business and implement their idea, however, great.
So, are they planning on adding features that will be incompatible with GPLv3?
Sure, every change they've made or will make that's licensed GPLv2 only. You can indeed download the most-recent "GPLv2 or higher" code, but if you then patch it with any "GPLv2 only" code you can only redistribute under GPLv2.
SMQ 90AE4B2BC4F6BEAF7340F0B40BA2DEF7340F6BC2D0392
Kaj Arno, MySQL VP of community relations, revealed the license change on his blog, on December 22, noting that the license for MySQL 5.0 and 5.1 had changed from "GPLv2 or later" to "GPLv2 only". As he explained, this was "in order to make it an option, not an obligation for the company to move to GPLv3".
To be more clear, what they have done was **take away the option** for other people to distribute the code under a GPL version higher than GPL2. Under the old wording, the company has no obligation to distribute their code under GPL3, even after GPL3 is released.
MySQL AB could continue distributing the code under GPL2 and leave others with the option of distributing their derivatives of it under GPL3. So their concern is what **other people** might do with their code, and has nothing to do with MySQL AB being forced to do anything.
What they are actually concerned with is a forking of their project by a group of developers who prefer GPL3 over GPL2. Anyone so inclined could still create such a fork using a version of MySQL distributed under the old wording (including the most recent current versions, if they received the code under the old license). They just wouldn't be able to use any code from MySQL that came after the license change. It would have to be a clean break, with no sharing of code after the license change.
IANAL, but it would seem to me that given a program consisting of two modules, one GPLv2+ and one GPLv2-only, the result would be distributable only under GPLv2.
Distribution under the terms of GPLv2 would satisfy the licensing requirements of both modules, but distribution under the terms of GPLv3 would violate the license of the v2-only module. The GPLv2+ module could be used as part of another program distributed according to the terms of GPLv3.
Note to ACs: I usually delete AC replies without reading them. If you want to talk to me, log in.
I'm laughing right along with you.
It's difficult to overstate how strong the ramifications of that "or any later version" clause are. For all intents and purposes, including that clause when you license your code is the next best thing to assigning copyright of your code to the FSF, since it allows them to re-license your code as they see fit.
Of course, the FSF likes you to assign copyright to them anyway, but failing that, they booby-trapped the license with language which most people have not been aware enough to question. Some true believers like to respond to that with "but I trust the FSF", but that's nonsense: how can you trust an organization that doesn't exist yet, i.e. the future FSF. What happens when Stallman dies, for example?
The fact that the FSF recommended this language is, to me, an example of overzealousness which the FSF should be ashamed of. We see these sorts of tactics used by politicians and legislators all the time. It's disappointing to see an organization like the FSF, which seems to be founded on some real principles rather than "let's see who can grab the most power", stoop to these levels to promote their agenda.
No doubt FSF members and fans would justify this language on a kind of "means justifies the ends" basis. But the problem is that the language in question represents a centralization of power with no real checks or balances. No moral organization should want such a thing.
More than that. It's the language that typically accompanies the copyright statement, and is the formal indication of the license granted by the copyright holder. The recipient can choose which set of rules they redistribute under, but I'm not at all sure they can change the actual license chosen by the copyright holder.
To be clear, If write a program and license it under v2+, I think you can:
What I think you cannot do at all, though, is change the license selection in my copyright statement to say "v3", or "v2", or anything at all other than "v2+".
Note to ACs: I usually delete AC replies without reading them. If you want to talk to me, log in.
I'm sorry, but I agree with Linus here, this is not abuse.
In fact, you could argue it is a security measure, a liability limiting measure, heck even a safety measure.
If some company implemented some hardware that runs in a critical area of some system (say, an airplane's fly by wire system) and just allowed any software whatsoever to run on it, and someone put new software on it, many people could be injured or killed very easily. To me, it is perfectly fine for the company to ensure that the software running on that system is qualified by them, both to help ensure safety and also to limit their own liability.
The FSF has contractual obligation to me, and I guess thousands of other contributors, that limit what they can do to with the GPL. When you donate software to the FSF, you get a signed contract back stipulating what they can do with it. It obviously leaves some freedom to the FSF (as seen with GPL3), but it is not a blank check.
Why would an airline allow anyone to install arbitrary, uncertified software on mission-critical hardware? And why would the vendor of that hardware have any interest in preventing the airline from installing whatever software they want?
But,really, that's all beside the point because GPLv3 would not prevent the hardware vendor from implementing cryptographic signature checking that ensures that only authorized code is placed on that device -- GPLv3 would only require that the vendor give the signing key to the airline that bought the hardware or, more realistically, allow the airline to use their own keys. That way, random attackers can't install random software on the hardware, and the legitimate buyer of the hardware can do whatever they want with it.
In short, your scenario has absolutely nothing to do with GPLv3. GPLv3 wouldn't disallow the benefits of code signing for safety-critical applications, and there's no realistic scenario under which the legitimate owner of safety-critical hardware should be denied permission to modify it by the license (other contracts, or even government regulations, may well deny them permission to do so, but that's a separate issue).
Note to ACs: I usually delete AC replies without reading them. If you want to talk to me, log in.
In any event, you can read my longer post on this article (a related topic).
If I write a kernel driver and it gets incorporated into the official Linux source tree, and I don't explicitly release copyright of it to the Linux kernel project, then they need to talk to me before changing the license on that code. But if I license it under the GPLv2 and allow the clause that the code may be distributed under "any future version of the GPL", then they could switch to GPLv3 without having to get my permission again.
GPLv3 is considered problematic at this point because of a few things the FSF is trying to do with the new license. They're seeking to prevent people from exploiting GPL code in ways that still manage to lock out user control (Tivo is the typical example - they provide the source for the GPL code they're using, but you can't change the firmware on a Tivo yourself because they have hardware that requires crypto signatures on the installed software) and they're seeking to limit the ways software patents can affect free software. (For instance, you can't write a piece of code, release under the GPL, and restrict people's access to that with software patents. I think they're also looking to prevent people from using a piece of GPL software and simultaneously attacking others' use of it with software patents.) Various people have various problems with these clauses - and while I appreciate what they're trying to accomplish I'm not sure it's entirely a good thing, either.
---GEC
I'm but the humble pupil, seeking to snatch the scratchbuilt pebble from the master's fully articulated hand