September 22, 2006

Kernel developers declare GPLv3 dangerous

Author: Joe 'Zonker' Brockmeier

The GNU General Public License version 3 (GPLv3) draft process took a hit today when a number of prominent kernel developers released a position statement deriding the "dangers and problems" with the GPLv3.

Ten developers have signed their names to the position paper: James Bottomley, Mauro Carvalho Chehab, Thomas Gleixner, Christoph Hellwig, Dave Jones, Greg Kroah-Hartman, Tony Luck, Andrew Morton, Trond Myklebust, and David Woodhouse.

The developers acknowledge the GPLv2 as a catalyst for helping foster the development community that works on the Linux kernel and has made Linux a success over other free operating systems. Due to the success of the GPLv2, the developers state that they are "reluctant to contemplate tampering with that licence except as bug fixes to correct exposed problems or updates counter imminent dangers."

The position statement takes issue with the GPLv3 draft, and says that "there's no substantial and identified problem with GPLv2 that it is trying to solve," while the new draft introduces several problems.

GPLv3 gripes

In particular, the letter specifies three provisions in the GPLv3 draft that are considered objectionable; the DRM clause, the additional restrictions clause, and the GPLv3 draft's patent provisions.

It's not shocking that the DRM, or "Tivoisation" clauses, are called out by kernel developers as unwanted. Linux creator Linus Torvalds has expressed his distaste for the DRM clauses on several occasions. While the developers agree that the use of DRM is "deeply disturbing," the developers say that "the essential freedoms of section 3 [of the paper] forbids us from ever accepting any licence which contains end use restrictions."

Another concern is the GPLv3's provision to allow additional restrictions above and beyond those in the GPLv3 itself. The developers argue that the ability to choose additional restrictions for the license "makes GPLv3 a pick and choose soup of possible restrictions which is going to be a nightmare for our distributions to sort out legally and get right."

Perhaps most importantly, the paper points out that the GPLv2 promises that future versions of the GPL will be "similar in spirit to the present version." However, the developers say that adding DRM restrictions "is tantamount to co-opting the work of all prior contributions into the service of the FSF's political ends" which is "a fundamental violation of the trust" placed in the FSF to produce future versions of the GPL similar in spirit to the current version.

I did attempt to contact the Free Software Foundation (FSF) and Eben Moglen, but they declined to comment on the letter at this time.

Kernel developers polled

In addition to the position paper, Bottomley posted the results of an informal poll of 29 kernel developers rating the desire to use the GPLv3 in the kernel. The options range from +3, "I wouldn't want to use v2" of the GPL to -3 "I wouldn't want to use v3."

How were the people chosen? Torvalds explained in a post to LKML that he basically chose from the "first screenful of people" who sign off on kernel contributions, plus Alan Cox.

Only one developer surveyed, David S. Miller, was neutral on the topic. The remainder of developers polled, including Torvalds, Alan Cox, Ingo Molnar, Al Viro, and Arjan van de Ven rated the idea of using GPLv3 negatively. None of the kernel developers polled rated the GPLv3 positively.

Though nearly all of the developers polled rated the GPLv3 negatively, only 10 signed on to the position paper. Bottomley says that this is because they felt it necessary to include all of the issues that came up during the discussion of the GPLv3 on the LKML, but some of the developers were uncomfortable endorsing all of the concerns that were raised.

Why now?

I spoke to Bottomley, the kernel developer who posted the paper to the Linux Kernel Mailing List (LKML) about the position paper and why they chose to release it now. Bottomley says that there was no reason in particular for releasing the position paper now except that "it's become clear that there hasn't been a huge amount of debate" over the GPLv3 and it looks like the draft process is proceeding "without any real debate or dissent."

Torvalds echos this in a mail to the kernel mailing list:

My personal opinion is that a lot of the public discussion has been driven by people who are motivated by the politics of the discussion. So you have a lot of very vocal GPLv3 supporters. But I think that the people who actually end up doing a lot of the development are usually not as vocal, and have actually not been heard very much at all.

In some sense, the poll is a way for the people who actually do a lot of the work to show that the FSF doesn't speak for necessarily even a very big portion of actual developers.

Bottomley also pointed out that this was not meant to be a public announcement, per se, but that discussion is done openly because the kernel is not a closed community. This contrasts with the GPLv3 process, where discussion is being done in public and behind closed doors in the GPLv3 committees.

Torvalds was invited to participate in the GPLv3 but either declined or did not respond to the invitation. Bottomley says that he's unaware of any other kernel developers having been invited to participate in the GPLv3 process, though he did say he has not made it a point to ask around.

Bottomley also acknowledges that it's an open question how developers could move to a new version of the GPL even if the kernel developers were emphatically in favor of the license. The Linux kernel is licensed under the GPLv2, without the "any future version" clause that most projects (particularly those managed by the FSF) use, and it would be difficult to gain consent from all kernel copyright holders to switch the license.

Scrap the process?

It's unclear what effect, if any, this will have on the GPLv3 draft process. Bottomley says that he hopes that the FSF would hold off on the new license if there is enough opposition to it.

The position paper states the fear that the GPLv3 could cause a "Balkanisation of the entire Open Source Universe upon which we rely," by forcing projects and vendors to fork packages "in order to get consistent licences."

To avoid this, the authors ask that the FSF "abandon the current GPLv3 process before it becomes too late."

Click Here!