gnu-misc-discuss
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Gentoo Linux copyright / CDDL question


From: David Kastrup
Subject: Re: Gentoo Linux copyright / CDDL question
Date: Wed, 18 Oct 2006 09:58:43 +0200
User-agent: Gnus/5.11 (Gnus v5.11) Emacs/22.0.50 (gnu/linux)

"Mike Cox" <mikecox_linux@yahoo.com> writes:

> Question to the nice folks in g.m.d.......
>
> I have a couple of nice patches I would like to contribute to
> the Gentoo project, but I have some doubts. As some people
> probably already know, Gentoo requires that their contributors
> transfer copyright of all contributed material to the
> Gentoo Foundation, as described here:
>
>   http://www.gentoo.org/proj/en/devrel/copyright/index.xml
>   http://www.gentoo.org/proj/en/devrel/copyright/assignment.pdf
>   http://www.gentoo.org/proj/en/devrel/copyright/assignment.txt
>
> However, some passages in this document seem very suspicious to
> my untrained, non-lawyer, eyes. Could any of the helpful folks in
> g.m.d please take a look and pick it apart. My questions in
> particular.....
>
> (1) Why would they need copyright for patches for programs they're
> just packaging, but didn't write themselves? (FAQ item #3)
> The problem I see here once a patch for any 3rd party application
> is submitted to bugs.gentoo.org, Gentoo owns copyright and the
> original contributor loses all rights over their code, including
> the possibility to use it in their own, potentially differently-
> licensed projects. Is it advisable to submit any non-trivial
> patches to Gentoo's bug tracker?

Personal choice.  It does seem curious that there is no reverse grant
to use as desired (which is what the FSF usually does) for one's own
use.

> (2) It says there legal action is only possible if all copyright is
> owned by a single entity. Isn't it rather true that any of the
> copyright holders can take legal action should the need arise?

Correct.  But in particular in the case of patches, a lot of people
might not be interested in pursuing a violation, and the passages for
which they would be suing might be too small to be really
significant.  Also, in case that the defendant has managed to get a
piece with his own copyright in, Gentoo will be unable to proceed
because of "unclean hands".

> (3) They are using the term "Intellectual Property." Isn't that term
> generally frowned upon in the Free Software world?

Pretty much, yes.

> (4) In other places they claim they're doing this because
> GNU, Debian, and the BSDs are requiring copyright transfer as
> well [1]. Well I know that some GNU projects recommend it (emacs,
> libstdc++), but by far not all; and I have heard first hand that
> Debian not only does not require it, but they don't even have an
> entity one could assign copyright to even if one wanted to. (Does
> anyone know about the BSDs?)
>
> Also, it has come to my attention that some of the work
> contributed to Gentoo is grossly missattributed; the most recent
> incident being dealt with in
>
>   http://bugs.gentoo.org/show_bug.cgi?id=150231
>
> Here, the writer of said guide (C.M.) complained back in May
> on a Gentoo mailing list that a different guy (who didn't write
> a single word in said document) is listed as the main writer while
> the real writer is only listed as contributor. After nothing
> happened in more than four months, the real writer got rediculed
> as being destructive to the Gentoo community as a whole--just for
> insisting on his copyright. Later on they did give credit--but
> somewhere in a footnote in an appendix, and hoped they would get
> away with it.

Mistakes happen, and there may be more sloppiness than malice involved
at least on the administrative level.  Assignments may keep the buck
rolling even in such situations, even though not rarely with a large
loss of trust and enthusiasm.

> Read more about it here
>
>   http://bugs.gentoo.org/show_bug.cgi?id=42130#c5
>
> I checked both and it seems the guy's right. He says he'll reveal
> more "when the time is right."

I don't like strategic games like that.  There is no reason why he
should not report what he feels is wrong right away.

> So here's my question: under these circumstances, how advisable is
> it to submit any patches to bugs.gentoo.org? We do have some fixes
> for two really longstanding bugs, but feel very uncomfortable about
> submitting them to Gentoo, even if it's against the spirit of Free
> Software and all.  What do the experts here think?

If there is significant value in the patches and you want to leverage
them for political pressure, there might be a reason to withhold them.
Personally, I don't really like this sort of power play.  If those
patches would be important to upstream outside of Gentoo, it would not
seem like a good idea to assign them to Gentoo when you might want to
contribute them where they would benefit more people.

I don't see that Gentoo goes against the "spirit of Free Software":
after all, they warrant certain licensing conditions.  But it would
certainly be more reassuring if they provided a grant-back of power,
and if they had clauses in place for the case where Gentoo was
dissolved.

> On a related subject, we have developed a kernel mode driver for our
> Microsoft Wireless Optical Mice with Tilt Wheel Technology, complete
> with device file in /dev and all needed ioctls necessary for
> querying signal strength, battery life status, and other misc.
> statistics, for Linux, that we have now ported to the OpenSolaris
> kernel (hence the xpost to c.u.s). My question is, is the same type
> of copyright transfer needed to get this driver into OpenSolaris
> proper?

No idea.

> How would we handle code shared between the Linux and OpenSolaris
> driver?

Ask for a reverse assignment of rights to maintain your code under
GPL2 (optionally "or later" for the case that this will become an
option at some point of time) in Linux.  It will be (C) Sun then, but
the Linux kernel policies don't mind mixed copyrights.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum


reply via email to

[Prev in Thread] Current Thread [Next in Thread]