August 5, 2006

Torvalds' comments on GPLv3 committees refuted

Author: Bruce Byfield

Linus Torvalds had harsh comments about the committees organized by the Free Software Foundation (FSF) to help it draft version 3 of the GNU Public License (GPLv3). However, so far as NewsForge can determine, none of those actually involved in the process agree with Torvalds' assessment that the FSF isn't listening to feedback.

Shortly after the second draft of the GPLv3 was released, Torvalds was quoted in a story by Stephen Shankland as saying "The FSF doesn't even seem interested in any feedback. They set up several 'committees' to get comments from various industry players, and everything I've heard about the process is that they then ignored them all and did what they wanted anyway."

When Newsforge asked for confirmation on the quote, Torvalds responded, "It's pretty accurate. I'm not at all happy with the way the GPLv3 thing has gone, and to me it looks very much like RMS [Richard M. Stallman] knew what he wanted when he started, and the discussion has been mostly about just the details in how to explain it. I'm sure the discussions have clarified (and hopefully thus improved) the license, but the end result seems to simply be inferior to the GPLv2, at least as far as I can tell."

Considering that more than 130 people are on the four committees involved in the drafting process, Torvalds might find a few who agree with him. But, if so, they weren't among those who responded to my requests for comments.

The Software Freedom Law Center, which is overseeing the drafting process declined to respond to Torvalds' comments, perhaps not wishing to feud publicly with such a major stakeholder as Torvalds, but committee members were quick to leap to the defense of the process. Far from Torvalds' view of the committees as a token show, those who responded firmly assert that the FSF has gone out its way to consult them.

Luis Villa, former director of the GNOME Foundation says, "In general, the process has felt very responsive, within the limits of this being the FSF's license and not ours. They have every right to ignore the feedback, but they've given every impression of listening carefully and thoughtfully, and I think the updated draft reflects that."

Ira Heffan, an associate with the law firm of Goodwin Procter and an expert in free licenses agrees, saying, "From my vantage point as a member of committee A, the FSF seems to have spent an enormous amount of time and effort listening to the concerns of the different stakeholders." Heffan describes the general process as a negotiation between the FSF and thousands of people.

Nor did the responding committee members report hearing their colleagues complain about the process. "I've not heard others complain," says Louis Suárez-Potts, chair of's community council and community manager. Others make almost the same comment in almost identical words, including Villa and Don Armstrong of Debian.

Talking more specifically about the process, Zak Greant of ezPublish and Mozilla notes that the FSF not only paid attention to his initial input on section 6, which describes the requirements for the distribution of source code with binaries, but "sought additional clarity from me." Greant adds that he has also been consulted "On other issues that I was not engaged with."

Benjamin Mako Hill, a leading Debian maintainer with a broad experience of free licenses, also believes that the committee process is working well. "I've been approached by probably half a dozen people," Hill says. "Many had issues they wanted resolved and I helped them file comments. A few others wanted to be on the committee, and that was possible as well. Membership would have been open to Linus, but I assume that he was too busy to participate."

Furthermore, talking about changes between the first and second draft, Hill states that, "You can trace almost every change back to a comment."

Hill went on to criticize Torvalds' comments directly. Referring to the language on Digital Rights Management (DRM) that is one of Torvalds' main objections to both the first and second draft, Hill says, "The DRM comment is much less controversial this time around, and there's been a huge amount of movement on the part of the FSF."

Hill admits that the FSF's consultation a few years ago on the GNU Free Documentation License "was basically a sham. There was a call for comments and almost all were ignored completely." But, he suggests, the changes in GPLv3 between drafts one and two prove that, this time, the FSF is listening. "The process for GPLv3 is unprecedentedly open," he concludes. "I know of no other license that even bothered to consult people in anything resembling this process."

Hill acknowledges that more work remains, particularly in the language on patents. However, he goes on to say, "Of course, at the end of the day, there will be one license and not everyone will get exactly what they want. Richard M. Stallman will certainly be one of those people [who doesn't get exactly what they want], and Linus might be too. It's going to be a process of compromise for everyone involved."

Torvalds' comments should probably be read in the context of his statement that his particular concerns, notably the mention of DRM, have not been addressed. Yet, in fact, they have been. In an interview with NewsForge, Eben Moglen, who heads the revision process, specifically mentioned that some changes in the second draft were specifically made in response to Torvalds' concerns. Apparently, the changes were not enough to satisfy Torvalds, who has ideological differences with the FSF, but that is no reason to condemn the process itself as hypocritical.

True, the FSF sometimes shows a streak of authoritarian rigidity. Yet that is all the more reason not to condemn it in this case, when it is genuinely consulting the community. Torvalds' comments are both unfair and unsubstantiated.

Bruce Byfield is a course designer and instructor, and a computer journalist who writes regularly for NewsForge,, and IT Manager's Journal.


  • Legal
Click Here!