Why this route?

Posted Sep 23, 2006 2:33 UTC (Sat) by coriordan (guest, #7544)
In reply to: Kernel developers' position on GPLv3 by gregkh
Parent article: Kernel developers' position on GPLv3

Have you submitted these comments through gplv3.fsf.org ? If not, why not?

(Keeping in mind that the issues reported there seem to be being listened to)

One important design feature of the gplv3.fsf.org comment portal is that it requires people to highlight what words and sentences their comments relate to. In public discussions, people can say "I/We don't like the patent bits", but at the portal, you have to say what words you disagree with. The downside is that commenters have to have actually read the draft, and they actually have to have a possibly-valid comment. It sounds like you have read the draft, so this isn't a problem in this case. The upside is that discussion starts from fact and details instead of general ideas.

One example is the length. For 15 years, people said "make it shorter", but when it is opened for revision, how many pointed out words that could be removed? Few or none.

Please, do submit your comments through gplv3.fsf.org, not just slashdot and lwn.net

to post comments

Why this route?

Posted Sep 23, 2006 18:41 UTC (Sat) by sepreece (guest, #19270) [Link] (1 responses)

Minor nit - one of the weak things about the GPLv3 comment portal (unless they've changed it recently) is that you have to comment on a text segment no longer than one sentence, which is often narrower than the arc of a particular thought within the license.

It's not well-designed for raising philosophical or holistic objections.

commenting

Posted Sep 24, 2006 1:53 UTC (Sun) by coriordan (guest, #7544) [Link]

To comment on a whole section, attach your comment to the section title.

Why this route?

Posted Sep 25, 2006 15:21 UTC (Mon) by mingo (subscriber, #31122) [Link] (7 responses)

Have you submitted these comments through gplv3.fsf.org ? If not, why not?

btw., this i see as another problem. The GPLv3 process is fundamentally undemocratic: the President of the FSF retains all rights to set the language of the GPLv3.

furthermore, Linus has stated that he has sent his comments to both Richard Stallman and to Eben Moglen directly, and that they were ignored, see this Groklaw comment of Linus:

Yes. I have emailed both rms and Eben directly. They know my position. They don't care. If they told you otherwise, they lied. Get over it.

Why this route?

Posted Sep 25, 2006 16:07 UTC (Mon) by alexbk (subscriber, #37839) [Link] (4 responses)

Wow, is that anonymous frustrated fella really Linus? Someone should really veryfy this. Posts like http://www.groklaw.net/comment.php?mode=display&sid=2... are borderline insults worthy of someone like ESR, and don't sound like Linus at all.

He's believed to be Linus

Posted Sep 25, 2006 20:24 UTC (Mon) by coriordan (guest, #7544) [Link]

PJ checked during a previous discussion (I guess she emailed him), and he was indeed Linus.

Why this route?

Posted Sep 25, 2006 23:24 UTC (Mon) by h2 (guest, #27965) [Link] (2 responses)

Exactly what I thought when I read that groklaw thread, but it became increasingly clear that the shrill, near hysterical, almost totally irrational voice I was reading was in fact Linus himself. If he was aiming to impress anyone with his arguments or clear reasoning he certainly failed in my case, and I used to respect him much more than I do now, after reading that thread. He showed a bit too much there if you ask me, and the fact that he had bad enough judgement to do that in the first place is also revealing.

I remember well a different Linus, who refused to take credit for the kernel, laughingly saying he just managed to get that credit by absorbing great work from others. That was a Linus I respected, this new Linus 2.0 is one I could happily never read another word from again.

The fact that the people he works with most closely agree with his position should not be surprising, and should not be considered grounds for discussion in the first place. What else would you expect? Of course the core guys more or less see eye to eye with Linus, otherwise they wouldn't be core guys. This has no more significance than polling a position at the democratic or republican or libertarian national political convention then using the results to show broad support for your position.

Why this route?

Posted Sep 26, 2006 1:19 UTC (Tue) by bronson (guest, #4806) [Link] (1 responses)

It shows broad support *among kernel developers* for Linus's position. And that certainly sounds significant to me.

depends on how seriously it was taken

Posted Sep 26, 2006 1:45 UTC (Tue) by coriordan (guest, #7544) [Link]

Well, that depends on how seriously the voters took the poll.

Why this route?

Posted Sep 25, 2006 19:08 UTC (Mon) by alexbk (subscriber, #37839) [Link]

The core of his argument seems to be that the kernel development is based on "fair trade" principle
and not on "freedom for the users" one. But fair trade can be interpreted in many ways, and I don't
see why his interpretation - "I give you source code, you give me your changes" is more valid than "I
give you the right to use my source code with your hardware, you give me the same right".

Why RMS?

Posted Sep 25, 2006 21:38 UTC (Mon) by coriordan (guest, #7544) [Link]

You've raised two new issues there, but I guess the answer to my question is no, you haven't submitted your comments via the comment system. I think it would be useful if you did. Comments submitted there are reviewed by a committee of lawyers, a committee of large businesses, a committee of free software projects, and a committee of general developers who'd shown an interest. Between the four committees, there are about 130 people.

For the two separate issues you raised. Richard has acknowledged that the first needs work. He replied to questions on this in Italy in March ("... We're going to have to replace me somehow, sooner or later.") and in Barcelona in June ("...Most of our community does not appreciate freedom ... So, if we wanted to do a good job of protecting freedom with version 3 of the GNU GPL, we could not let the majority of our users decide what goes into that licence..."). A vote isn't the right thing to do (just like a vote is not the best way to land a plane or do surgery), but yes, we can't always have Richard making the decisions. We need to build a committee which can be trusted to update the GPL in the spirit of copyleft.

For your second issue, it's hard to know what Linus has been saying to Richard. It would be easier to know about Linus' comments if he submitted comments to the public gplv3.fsf.org portal, attached them to the actual licence text, and got them discussed by the committees. And the same goes equally for the other Linux developers and members of other free software projects.

People who are trying to improve GPLv3 can take some hints from the above letter, but it would be far more useful if that letter also had some comments on the text of draft 2.

For example, section 5.2 of the above letter recommends removing section 7b of GPLv3 draft 2 because of a certain list of problems. These problems are mostly known, but section 7b has the benefit of making GPLv3 compatible with more free software licences, such as the Apache licence. Licence incompatibility is a pointless bureaucratic impediment for free software developers, so removing incompatibility should be tried.

So, while highlighting the problems is useful, we should also look at each problem and decide how to minimise it and how much of a problem it is, so hopefully we can fix the problem, or fix it sufficiently, and keep the benefits. This discussion is easiest to have inside the consultation system, instead of via the community and mainstream news outlets.