The GPLv3 Debate: Focusing On the Issues

LXer Feature: 02-Nov-2006 Yesterday we sought to "separate the wheat from the chaff", so to speak. Today, we'll take a look at the heart of the issue - the technical question that has been raised around the GPL's anti-DRM clause. I'm all for fighting DRM. I want the digital content I pay for to be available on my terms, seeing how I already paid for it. The question, though, is whether the GPLv3 will be the most effective tool for combating Big Media's attempt to take away our fair use rights.

Intro

In yesterday's article, I revisited the history of the GPL, if only briefly, to help provide some perspective on how the license came about. I pointed out how much of the debate has focused on peripheral issues, such as certain personalities involved. So what are the real issues? Can we solve this debate?

Keep Focused on the Real Issues

The real issues behind the anti-tivoization clause in the GNU General Public License v.3 are not whether Stallman enjoys smelling flowers or dons a saintly persona to add a light touch to an otherwise serious (and rather dull) issue, but whether anti-tivoization is what we really want, and whether a software license is the best tool to prevent tivoization. With that in mind, let's examine these issues.

When I contacted Tivo's PR guy, he said he would get back to me and referred me to Daniel Lyons' articles on Tivo. Lyons' article, "Toppling Linux" was about as useful as mining for gold in a glass of water. Not only would non-adoption by the kernel team not topple Linux (or the GPLv3), his article actually omits quite a few facts. I clarified those issues in my first article.

His interview with Torvalds on the Tivo issue was a bit more helpful, however. Torvalds says that Tivo built the box, and should therefore decide how to upgrade it. I disagree strongly with that sentiment, especially where the box is owned by the user. Were it leased or rented, it wouldn't be my responsibility. Not only that, but this is exactly the point of Free Software - to give the user control of the software that runs on their hardware. And while Linus doesn't think that activism and technical decisions mesh well, a number of people have discovered technological freedom precisely because someone told them about it. Activism is the only way to spread the word when you don't have a mega-business marketing budget.

A Few Facts & Observations

First and foremost, I would like to raise a question about the developers involved in the GNU project. If these people are so adamantly opposed to the anti-DRM clause, why haven't they broken ranks? Some outside the FSF have rightly expressed concerns about the GPLv3, in light of Torvalds' reservations. But I would be far more concerned about the hopes for the GPLv3 if the GNU Project developers were deserting Stallman and threatening to fork their Projects. I haven't seen any hint of that as yet.

The anti-DRM clause has been frequently mis-labelled as the anti-tivoization clause, which rather narrows the scope of the discussion. I wonder if Torvalds is right about the anti-DRM clause affecting only the kernel? If true, I would question the wisdom of a provision that has such a limited impact, especially if the kernel developers refuse to use it. That would lead to a dead clause in the license - a waste of space. I think Torvalds' statement is true, but only for Tivo. We would be short-sighted to not consider other applications.

Would not the anti-DRM clause also affect other GPL'ed software used in other devices likely to deploy DRM measures? For example, a developer tries to implement Rhythmbox on an portable audio player, with DRM code integrated into the app. Right now, it certainly seems possible, even if unlikely. The anti-DRM clause would prevent Rhythmbox being used that way, presumably because the developer made a conscientious choice to block such efforts. If the developer doesn't choose the GPLv3, then it's a moot point. If the developer chooses the GPLv3 because it's "popular", and later complains that it's "too rigid" or "political", then she chose the license for the wrong reason.

Torvalds raises another interesting point in his interview with Lyons. Even if Linus convinced the hackers to migrate to the GPLv3, Tivo could continue using the GPLv2 kernel. This should cause one to wonder just how effective any software license could be against such a rotten trick. With Free alternatives, there's just not much need for a Tivo. So it stands to reason that, although someone can exploit a seeming loophole in the GPL, people can still negate the exploits by refusing to run those. A growing number of people are refusing to run Windows, and this wouldn't be any different.

Tremendous Challenges

I am not so much arguing against the anti-tivoization clause, as wrestling with this issue in a way that I hope will help others see the various angles. Tivo did what Stallman set out to oppose to begin with - robbing users of the freedom to control the software that runs the hardware they own. I find that utterly dispicable. Still, if the GPLv3 only affects the kernel, and those developers don't migrate to the GPLv3 license, then there may not be much point in having such a clause in the license. Perhaps targeting Tivo in the Defective By Design campaign would be a better strategy.

That is not to say that the GPLv3 will destroy Linux, or that the FSF will lose a large following, or even that use of the GPL will fall off the face of the earth. One would be a fool to think such things, much less say them. There are simply too many people who identify wholeheartedly with the GPL's spirit and terms. At worst, the GPLv3 simply will not have quite the impact that it currently has. Everything else, though, will largely be status quo. One effect the GPLv3 (as it stands now) might have is to more effectively separate those who choose a given license for convenience' sake from those really buy into freedom for all users. After all, freedom for all users necessarily means a little less freedom for developers.

Look, I really wish I had the answer. Honest. I wish I could just spell out a few simple words that would solve the problem and make everyone happy. Still, I'll take a stab at this. I suggest that the FSF consider leaving the patent clause in place (with any necessary clarifications), but either re-write the anti-DRM clause, or remove it for now. They could also delay releasing the GPLv3 until they have studied the issue a little more. If we find an effective method for fighting DRM through software licenses, then let's add it in.

I haven't covered every nook and cranny - just a couple of points that seem worth thinking about. I'm sure readers will see other issues that need to be addressed. Hopefully, readers will also be able to offer some insight on how the anti-DRM clause, as it stands, might prove effective. Perhaps someone can step forward with a revised clause that is (more) effective. Don't attack the various personalities. Just focus - if only for now - on what's important, not who's weird. There will be plenty of time later for a good, old-fashioned flamewar.