News.com looks
into GPL v3 concerns. "Eben Moglen, general counsel of the Free
Software Foundation, said Thursday that there shouldn't be a problem in
persuading Linux developers to migrate to GPL 3, as the license will be
developed with their input. "I don't think it will be a difficulty,"
Moglen said. "When the FSF finishes its work to produce the first
discussion draft of GPL 3, there will be an extended comment period, which
will be a chance for everybody to have their say. We will take as long in
listening as people need to take.""
(Log in to post comments)
Posted Mar 26, 2005 13:10 UTC (Sat) by copsewood (subscriber, #199)
[Link]
This is an historic opportunity to enhance and defend software freedoms, which will make a great difference to the world our children inherit from us. There are very few in the modern networked world who can afford not to use or license free software.
The coming battle against software patents and copyright term extensions which go beyond economic logic is not yet won. This is a time to rally behind the leaders of the free world to support their efforts, in ensuring that those who legally use free software do so under terms which prevent these beneficiaries from discriminating against its development.
As far as the threat of forks and divisions are concerned, it is not possible to make ommelettes without cracking eggs. There will be projects whose leaders are persuaded to adopt the new license, and this will become increasingly popular as the threat against our entire community grows. The risk of having to expend some effort replacing what a few unpersuadable and obstructive copyright owners will not allow use of within these projects will be small in comparison. Project leaders who decide to prepare for a change to the new license might accept only dual-licensed code contributions during a changeover period.
There might also need to be some management of risk in situations where the original copyright holders are uncontactable or deceased. In this situation the risk of hostile parties being able to prove ownership of such copyrights, and having willingness to assert these in a damaging way against a project which upgrades to the new GPL3 is likely to be small. The risk of allowing relative weaknesses inherent in continuing with the GPL2 license to be exploited by our enemies might be greater.
Our commitment must be to replace seriously contested code incompatible with the license chosen for free software projects. This will remove economic incentive to those who would contest our code, other than to create legal nuisance to those they perceive as competition solely for the sake of doing so.
We will of course need to apply careful scrutiny to the new license when it emerges. But this review must be applied to the task of maximising the benefits to the free software world of applying this license in as many projects as possible. Let us go into this review with an absolute determination to keep the free software community working together on this, and to give no ground to the those gathering against us who would see private gain though prevention and curtailment of the free software world.
GPL v3 and the Linux kernel?
Posted Mar 26, 2005 14:06 UTC (Sat) by Duncan (guest, #6647)
[Link]
Perhaps the most significant piece of software this will affect is the
Linux kernel itself, because it has traditionally been licensed under GPL
v2 only, not the common "2 or later" wording, unless a contributor states
otherwise, AND Linus has always allowed contributors to maintain ownership
of their copyrights.
While that has had the practical effect of making it impossible for
anyone, like that guy who tried to buy a BSD licensed copy of the kernel,
to actually do so, due to the practical impossibility of tracking everyone
with copyright ownership in the kernel down and getting their permission,
even if Linus /were/ to lose his head and fall for such an offer, and I
expect most LWN readers will consider that a /good/ thing, it will /also/
mean it will be years before the kernel switches to the GPL v3, even if
they decide it's a /good/ thing and embark on a concerted campaign to do
so. Eliminating or rewriting any portion found to violate someone's
patent or copyright rights isn't that big a deal, but rewriting the entire
kernel to a new license will take some serious time, even if it /is/
decided that is the way to go, and it's rather likely that one or more
major contributors will continue to insist on the GPL v2, for whatever
reason.
I expect, therefore, that even if it's generally agreed that the kernel
should move to GPL v3, most submissions will have some form of "2 or 3"
language in them for perhaps half a decade, if it doesn't become more or
less permanent.
However, Moglen /has/ previously stated that v3 is being designed to be v2
compatible, tho it's difficult to see how, if v3 is going to add the oft
mentioned web code openness guarantees or any sort of patent defenses,
because that is additional restrictions normally not considered allowed by
the GPL v2. If he succeeds in somehow pulling it off, tho, dual licensing
GPL v2 and v3 shouldn't be necessary, unless the code contributor wants to
explicitly allow the v2 terms, because v3 will be compatible. Even a
simple explicit compatibility statement, the solution taken in some other
cases, won't work, if the additional patent defenses and web app source
requirements are indeed to be in v3 and hold any weight at all, because
that would be equivilent to v2 anyway, meaning why go to the trouble of v3
in the first place, unless it relaxes restrictions, which I find difficult
to envision.
It'll certainly be interesting to see what it looks like, that's for sure,
and equally interesting to see how the various projects currently using
the GPL v2 specifically react, as well whether those using v2 or later,
now change the wording to v3 or later for new code and where the code
owner allows.
Duncan
GPL v3 and the Linux kernel?
Posted Mar 27, 2005 6:41 UTC (Sun) by piman (subscriber, #8957)
[Link]
Despite the common adage that the Linux kernel is under "GPL v2 only", the kernel is heterogeneously licensed, sometimes even incompatibly or infringingly so. Much code is already licensed as "GPL2 and BSD" or "GPL2 or later", or sometimes "THIS CODE IS PROPRIETARY INTELLECTUAL PROPERTY OF LARGE CONGLOMERATE AND IS NOT FOR PUBLICATION". I predict it won't be as difficult to get a kernel compatible with v3 as most people think it will be. I also predict that kernel developers, largely, won't care.
The copyright licensing of derivative works...
Posted Mar 27, 2005 11:55 UTC (Sun) by hummassa (subscriber, #307)
[Link]
Must abide to the terms of the licensing that gave permission to the
derivative works to be created *first*. No, this is not as easy as it
seems. Everything that could be deemed (thru abstraction, filtration,
comparison) as derivative on the GPLv2'd code can only be distributed
under the GPLv2 terms...
Posted Mar 26, 2005 17:06 UTC (Sat) by piman (subscriber, #8957)
[Link]
> When the FSF finishes its work to produce the first discussion draft of GPL 3, there will be an extended comment period, which will be a chance for everybody to have their say. We will take as long in listening as people need to take.
The GNU FDL had a similar comment period, and all comments about (and before, and after) were ignored; indeed, the FSF even reneged on their promise to summarize the comments in a timely fashion (as in, before the next version of the FDL was published), let alone act on them. When people tried talking to RMS about the FDL after the comment period, he told us GNU licenses were none of our business.
Posted Mar 26, 2005 18:01 UTC (Sat) by madscientist (subscriber, #16861)
[Link]
I agree; the handling of the GFDL does NOT inspire confidence in the process, regardless of how heartily Mr. Moglen praises it. Actions > words and all that. I guess we'll see!
Posted Mar 27, 2005 0:59 UTC (Sun) by dlang (✭ supporter ✭, #313)
[Link]
the recent issue with the EMACS documentation (another project asked if the EMACS docs could be put under a better license then the GFDL and were told to get lost) would indicate that they haven't learned from that experiance.
Posted Mar 27, 2005 17:11 UTC (Sun) by coriordan (guest, #7544)
[Link]
Now now Micheal Moore, lets tell the whole story.
An XEmacs developer (remember that project that hired away FSF's Emacs release manager near to a release, the project that doesn't keep proper records of contributions, thus making GPL enforcement hard or impracticle, the project that... and there's two sides to that story), asked for the entire Emacs manual to be GPL'd.
Stallman pointed out that this would make the application of the GFDL pointless, and he offered to instead relicense parts of the manual, and he asked the XEmacs developer what he would like relicensed.
The XEmacs developer started complaining on the Emacs development list, so Stallman said forget it.
Posted Mar 31, 2005 17:10 UTC (Thu) by jschrod (subscriber, #1646)
[Link]
I read that email exchange on the emacs-developer list. That flameware was started and fanned by many people, but not by the XEmacs developer (Ben Wing) who posted the original question about reuse of the Emacs manual.
In particular, he did _not_ ask for the entire Emacs manual to be GPL'd, that was David Kastrup and others -- who develops elisp code for FSF Emacs mostly, not for XEmacs (AUC-TeX, et.al.). Ben asked if the XEmacs team could use the Emacs manual under the old license that it had formerly.
Concerning the Emacs schism, if our fellow readers want more information: the XEmacs Web site has information from both sides. jwz's Web site has the biased anti-RMS view. I don't know of any further RMS publication where he presents his side, besides his writing that's available from http://www.xemacs.org/About/XEmacsVsGNUemacs.html.
Summary: Watch your own biased representation, please. It doesn't help calling others Micheal[sic!] Moore and then not getting your own facts right.
Posted Mar 27, 2005 5:53 UTC (Sun) by piman (subscriber, #8957)
[Link]
If the FSF does not have the resources to publish, after a period of three years, a summary of comments sent to them (under the good faith assumption the comments would be published!) about their license, then perhaps they should reconsider how they allocate their resources. It doesn't take a lawyer, or even an expert legal amateur, to summarize comments.
Certainly the status quo would suggest the FDL is -- and has been, for a long time -- a much more pressing issue than GPLv3. It's like spending a year spell-checking your comments when your program is corrupting hard drives.
Posted Mar 27, 2005 17:33 UTC (Sun) by coriordan (guest, #7544)
[Link]
I don't remember the details, but you say that comments sent were "under the good faith assumption the comments would be published".
How would FSF know that everyone had such an expectation? Maybe they had to assume that some comments were sent on a good faith assumption that comments would remain private?
...so why not publish with names withheld? Maybe even this would be too little privacy for some.
If you just want that your comments were published, you can publish them yourself.
Decisions like this don't go on for 3 years. Clearly FSF made a decision at the time and moved on to something else.
(I don't remember the whole process, I'm just replying based on the information you've given.)
I disagree that the FDL is more pressing than the GPLv3. GPLv3 is needed to better protect free software developers from patent litigation, protect free softare users from Trusted Computing control, and prevent free software owners from writing their own license (because they wish to have the benefits of the coming GPLv3 now - or because they need a license which is in their language).
Posted Mar 27, 2005 19:44 UTC (Sun) by vmole (guest, #111)
[Link]
How would FSF know that everyone had such an expectation?
Perhaps because the request for comment said that they would summarize and publish the comments? Perhaps because representatives of the FSF have repeatedly said they would do so? AFAICT, none of the issues raised by various people on debian-legal, which were submitted to the FSF, have received any response other than "we don't think that's a problem". Which is, of course, their perogative, but it makes a sham of "we'll take input after the initial draft".
Posted Mar 27, 2005 21:31 UTC (Sun) by piman (subscriber, #8957)
[Link]
My complaint is not that they didn't publish the comments, it's that they all-out ignored them. Obviously this includes not publishing them, but had they not published but still *acted* on them I wouldn't be complaining nearly so loudly.
I have no faith that the GPLv3 comment period (and subsequent release of a crappy license) will not be a repeat of the FDL debacle, in which RMS declared that the FSF's licenses are nobody's business but the FSF. That I have no faith in it doesn't mean it won't happen; but it means the FSF is going to have to actually do it, before they get any kind of plaudit from me.
> I disagree that the FDL is more pressing than the GPLv3.
A hundred non-free manuals *right now*, versus a few uneasy companies being uneasy for a few more months? Give me a break. Besides, the FSF could solve the whole issue by making the next version of the FDL equivalent to the GPL. That takes no real effort. They've made it clear they're going to ignore other opinions on the issue.
Free software users have been hurt and will be hurt more by non-free documentation, than waiting a little longer for IBM to dump their next pile of dead projects on us.
> and prevent free software owners from writing their own license...
Of course. All we need to do is write one more all-encompasing stand... I mean, license, and then no one will ever need to deal with all the other standar... I mean, licenses, again! RMS can wave his hands and remove all irrational hatred of the GPL, NIH syndrome, and other stupid things.
It is disappointing to watch good programmers make the same mistakes with licensing that they would never make with code.
Posted Mar 28, 2005 7:44 UTC (Mon) by pphaneuf (guest, #23480)
[Link]
We will take as long in listening as people need to take.
Considering that Debian developers are a significant portion of the people involved, this means the GPL v3 should come out around 2009 or so. So no worries, people.