Skip to content
Welcome guest. | Register | Login | Add
About | Wiki | Legacy

GPLv3 is to preserve Free Software, not kill it

A group of important Linux kernel developers have recently published a position statement on GPLv3, as reported by Steven J. Vaughan-Nichols in a story with a rather alarmist title: "GPLv3 could kill open source, top Linux dev's warn". In truth, the goal is exactly and essentially just the opposite. I have little doubt that the statement will deepen the controversy already present around the new version of the most popular Free Software license, especially since Linus Torvalds criticized the license on a number of occassions before.

Unfortunately, the statement doesn't strike me as very constructive in regards to the GPLv3 development process. Indeed, it calls for its termination under a rather drastic prediction of what should happen otherwise, " balkanization, which will be manifested by distributions being forced to fork various packages in order to get consistent licenses, has the potential to inflict massive collateral damage upon our entire ecosystem and jeopardize the very utility and survival of Open Source."

So what are their objections? First of all, they believe that GPLv2 has served them well in the past and are hence reluctant to consider any changes. Interestingly though they do acknowledge that if there are any "bugs" in the license, they need to be fixed while at the same time missing to mention two greatest bugs which GPLv2 currently carries within the context of times in which we face the threat of software patents and DRM. In that sense this statement actually contradicts the whole message behind their article:

Since GPLv2 has served us so well for so long, and since it is the foundation of our developer contract which has helped propel Linux to the successes it enjoys today, we are extremely reluctant to contemplate tampering with that license except as bug fixes to correct exposed problems or updates counter imminent dangers.

The thing is that there indeed are certain imminent dangers to counter in context of which GPLv2 indeed does have bugs which need fixing. On the merit of this quote alone we could reject the call for GPLv3 process termination with which their position statement concludes. Something needs to be done to counter Digital Restrictions Management and software patents and FSF is doing it with the revision of the GNU GPL, a license which covers majority of all Free Software.

The motive behind GPLv3 is no different than the motive behind GPLv2. It is to guarantee freedom to all users of the covered software. Changes which are being introduced to GPLv3 have no other purpose but to address what has essentially enabled a legal circumvention of the license terms, as the TiVo case vividly demonstrated. With DRM, it is possible to obey the license while in practice make freedoms which it grants impossible to excercise. By loading GPLv2 covered software on a device which carries a DRM technology which prevents any modified version of the covered software to run properly, you are making a joke of rights that GPLv2 was supposed to protect.

Software patents too are a way to counter the protection of rights which GPLv2 was supposed to guarantee. Software patents today cover so many software designs, concepts and ideas that it is more likely for a programmer to use patented ideas and fall under risk of patent litigation than not. Because of this, software patents are a threat not only to Free Software, but to software innovation as a whole. Creating patent commons and promising not to sue may be commendable, but it doesn't resolve the whole problem and there still remains a significant number of patents which may pose a threat. This too is a danger that needs to be countered and in accordance to its goal of granting and preserving users freedom, GPLv3 is meant to do exactly that.

Certain patent holders do not like that of course, and it is true that some of them have made significant contributions to certain Free Software projects. However, these contributions have not been made out of generosity alone, but were driven by certain interests as well. By contributing they were essentially becoming a part of our community, one which is, or at least should be, governed by the principle of freedom. It is, afterall, exactly that principle which makes our community, our development process and our software so special and interesting to the outside world. It is at the same time exactly the thing that attracted involvement by these corporations.

And if it is in the interest of protecting this freedom, should we really not take certain measures to shield ourselves from software patents threats just because a few corporations don't like it and may retreat their contributions?

Why should we be the ones to give in? Are we really so dependable on them? Why are they contributing in the first place? They contribute because they themselves derive value from cooperating with us. They will have to evaluate what is more important for them, patents which they would have to give up to continue deriving the benefit of cooperating with us, or these benefits themselves. I do not believe the risk of loosing a few corporate contributors trumps the risk posed by software patents in general.

The position statement by Linux kernel devs and maintainers expresses doubts about anti-DRM and anti-software patents provisions being the right way to combat these threats. However, it unfortunately doesn't provide us with alternative solutions either. In fact I have not seen an alternative proposition strong enough to warrant abandoning the one solution that we do have at this point: GPLv3.

They also criticize anti-DRM provisions to be a restriction of possible uses of the software covered by the license and imply that this goes against the goal of freedom which FSF professes to support. But how can that be when the only thing which GPLv3 restricts against is exactly the thing which would be the real restrictor. These are not the freedom restricting "restrictions". These are freedom preserving restrictions.

Overall, the statement seems to show some evidence of misunderstanding of the implications and goals behind GPLv3. Some criticize that because RMS and FSF reject their calls for changes or even in this case termination of GPLv3 that they are not being listened to and that their input is being ignored. What is, however, FSF supposed to do when propositions made go against the stated goal of the license revision? Just because the license development process is open to everyone, as no other license development process has ever been, doesn't mean that the end goal of the license can be affected by those who provide input and criticism. Their input is not supposed to change the goal. Their input is supposed to help find the best way to reach this goal.

While the position statement by Linux kernel devs may bear some marginal value in the sense of emphasizing the existance of opposition to the license and hence possibly influence more effort at resolving certain issues that can be addressed, its drastic conclusion, the termination of GPLv3, is by all means off the mark. There is little reason to warrant ceasing the GPLv3 process compared to the reasons that exist for continuing.

GPLv3 is to build a protective legal wall around the pool of software it will cover which will keep software patents and DRM out and preserve freedom for those who use this software fully excercisable. Those who wish to benefit from the pool (everyone from individuals to corporations) should naturally respect the rules which protect freedom they benefit from for others to benefit from equally as well. That is the only real goal behind GPLv3.

Thank you
Danijel Orsolic
Libervis Network

Note: FSF has sent a clarification in light of recent discussion on GPLv3.


quid pro quo


The funny thing about the quid-pro-quo argument is that it equally applies to the GPL3, if you use different terms of reference. For instance, the patent clause basically says (if I understand it correctly) that if you use GPLed software which may contain patentable code then you agree not to use your patents against the GPLed software. It sounds pretty quid-pro-quo to me. The DRM clause which should really be called the keygen clause since the word's DRM are not in version 2 of the GPLv3 says (if I understand it correctly) that you can't use and modify the code unless you allow for your platform unless you allow others to use and modify that code for your platform. That basically means, if your GPLed code requires a private key to run, then you either need to provide that private key or provide a key generator to generate a new key. Again, fairly quid-pro-quo.

Personally, I think the crux of the issue isn't the quid-pro-quo argument -- it's something simpler. The GPLv2 works as far as his needs go and it's time proven in courts. It would also be a pain to convert the kernel to GPLv3 (though it is possible on a file by file basis to change GPL2 only to GPL2 or greater, and then rewrite any GPL2 only code where the author can't be contacted or refuses to change). The GPLv3 has not been time proven and Linus is uncertain about the extent that the patent and keygen clause will help or hurt the Linux kernel if it were adopted. The only sure thing is that it would cause some confusion in the short term, and that's something he'd likely like to avoid unless there's a pressing need to change. (remember how much pain the Bitkeeper to GIT clause caused?)

wish I agreed


but I don't. DRM isn't going to go away, and GPL3 is only going to marginalize FSF's influence over the open source world because instead of engaging the corporate community in a dialogue, it preaches to them. It doesn't matter if it's a 'better' license for end users if it's so full of zealotry that no one in their right mind is going to use it. 'Better' doesn't mean instant adoption, if we've learned anything from VHS/Beta, Windows/Linux/BSD, mp3/ogg debates.

I'm not saying it's a bad goal, but if the license goes too far then it runs the risk exacerbating the already existing license fragmentation problem.

exactly what they said about GPL 2


GPL 2 is a cancer, will destroy capitalism as we know it, and the entire foundation of all intellectual property laws. Maybe you remember reading those items in the past years? I do.

It didn't and doesn't do those things, it simply creates a different model under which to develop software, one in which the rights of the end user cannot be sold or eliminated.

All that the gpl 3 does is make sure that people's work does not get used in ways that subvert this basic concept, like the tivo box for example, which is simply the first in what will be a steadily expanding wave of gpl 2 products that cannot be used or changed in any way.

There's nothing zealous about this position, it's simply a completely logical next step to counteract the next easily visible threat to free software on the horizon.

Mistaking the ability to see slightly further ahead than most people with zealotry seems to be a consistent mistake people who are opposed to the gpl 3 process are making. The real radical concept is already in place and now considered more or less normal is this: you must share if you use gpl code. You cannot hog it, you must give back. The little tweaks gpl 3 make to this far more profound concept are really fairly trivial if you look at it objectively, and that's quite obvious when you see that only corporations who want to keep the status quo, and the kernel developers who work for those corporations, really seem to have a serious opposition to this.

And of course, no law makes anyone use the gpl 3 license, but I think it will spread quite fast, and will not be fatal, and will simply be a good update. Unless you pay attention to the fud and fearmongers, who are simply replaying the gpl 2 songs they worked on a few years ago when they tried to resist that evolution.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.