Re: Gnus Manual License
Hubert Chan <hubert@uhoreg.ca> writes:
> On Thu, 12 Oct 2006 11:17:01 +0200, David Kastrup <dak@gnu.org> said:
>
>> Feel free to ask yourself at copyright-clerk at gnu dot org in
>> order to get an official statement.
>
> I will do so if I ever feel like modifying a GNU manual for my own
> purposes.
And in the mean time you will be happy to spread FUD. Either you are
interested in getting the question answered, then you should ask, or
you are not, then you should not spread misinformation.
>> Wrong. The copyright holder gives his permissions only to copies
>> he distributed himself. That the copyright holder at some point of
>> time or other gave somebody else a license with different
>> conditions does not mean that you can give yourself that license on
>> a whim. Not even when the copyright holder permitted some other
>> party to redistribute. _IF_ and only if you got your copy
>> _accompanied_ by a license and _understood_ to be under such a
>> license, _then_ you have the right to redistribute under the
>> conditions granted by that license.
>
> Wrong. If I get my copy by other means, and am able to obtain a
> different license from the copyright holder,
A license for that particular copy you have, and only if there are no
third party rights involved. It will be pretty much impossible to
warrant, anyway, that you arrive at the original licensable version
again after replacing all copyright notices.
> then I am also allowed to redistribute under the conditions granted
> by that license. The license does not always have to accompany the
> copy. It can be obtained through a separate channel (as long as
> that channel is legal).
But you can't just get any old GPL. It has to be one intended for
your copy, and that means that you would have, in case of doubt, ask
the copyright holder explicitly for such a license. Which puts you
back at square one: asking is always an option.
> I'm not talking about recreating a license out of the blue. I'm
> talking about obtaining permission from the copyright holder for a
> copy obtained from a third party.
>
> If I download some software from Debian that's licensed under the
> GPL, and I can get the copyright holder to let me use it under the
> terms of the BSD license (say, in exchange for $[unspecified sum]),
> then he cannot turn around and sue me if I distribute the software
> under the terms of the BSD license.
When asking explicitly. But you can't just say "I know this has been
distributed elsewhere under the BSD, and even though I can't persuade
the original person getting it to give me a copy, let us just presume
I could have done so".
>> You can't just hack into my server and steal the software, claiming
>> that because I licensed it to somebody under the GPL, anybody
>> should be able to attach the GPL to a copy he has acquired without
>> a license.
>
> Straw man. Did I ever say anything about hacking into a server? The
> documentation that I obtain from Debian is exactly the same as what I
> would get from the FSF's own _publicly_ available sources.
It won't be the same. For one thing, it will have had the licenses
removed. For another thing, the FSF might have ceased distribution in
order not to have software distributed under older licenses. Limiting
distribution channels of free software is perfectly feasible.
> They give it to _everyone_ under the GFDL.
They _once_ gave it to everyone who _bothered_ to download it at _one_
time.
> To sue somebody under these situations would be completely stupid.
You are the specialist for suing under completely stupid conditions.
>> Debian developers having their own work licensed under the GFDL?
>
> Most Debian-related changes to documentation are probably very
> minor, which are probably too small to fall under copyright, or
> Debian-specific things. If they had to pass a larger chunk upstream
> to, say the FSF, I assume they would have to sign the standard
> copyright assignment. Most other upstreams do not care much about
> copyright assignment, but if a Debian developer pushes his changes
> upstream, and upstream is known to license the documentation under
> the GFDL, I don't see why the Debian changes would not be licensed
> under the GFDL.
But you are conveniently glossing over the fact that Debian is a major
redistribution channel. People will tend to have Debian or
Debian-derived disks at home, and they won't be able to use those
copies under the GFDL. It is the public who is refrained from using
the license intended for the work from upstream. The FSF (as opposed
to Linux kernel development) requires copyright assignments for
strategically important contributions.
> Does it really surprise you that there might be some Debian
> developers who are not so fanatical against the GFDL that they would
> refuse to have any of their work GFDLed? Would it surprise you to
> hear that even I would not be opposed to contributing to a GFDLed
> document, and having my own contributions licensed under the GFDL?
It does not matter. If a minority of 30% opposes the GFDL, then 30%
will have the option of boycotting it.
The GNU public licenses were created for the exact purpose of not
being dependent on goodwill alone.
> [... snip anecdote ...]
>> So just judging from factual experience I have as project
>> maintainer with Debian developers, I can tell you that your
>> assurances about Debian developers in general don't hold. Even a
>> Debian vote does not assure that the people ending up in a minority
>> position of a vote will be interested in implementing or supporting
>> what others have decided.
>
> That's one developer.
Of the two active in AUCTeX development who I know to be Debian
developers. The other Debian developer is the Debian package
maintainer himself.
> My experience is different. Of course there will be people who
> refuse to package GFDLed documentation, under any circumstances,
> just like there are people who refused to remove GFDLed
> documentation until after they were forced to by the vote. Anyways,
> I don't really care too much to debate this point, if we're just
> going to swap anecdotes.
You were making claims about what Debian developers would do, and your
claims do not even hold up to the very limited exposure to Debian
developers I have. I have a strong counterexample even in that.
Whether anecdotal or not, it definitely makes the point that your
claims don't hold.
>>> And again, how does the GPL present insurmountable barriers to
>>> mass printing?
>
>> Maybe you should ask that question to the publishers who refused
>> printing GNU manuals under simpler licenses. There is a _reason_
>> that the GFDL was created.
>
> Yes, I agree that there is a reason the GFDL was created. I don't
> think that "the GPL makes it too hard for printers" tells the whole
> story. (And no, I'm not trying to imply some sort if insidious
> motivation for the GFDL, just in case you got that impression.)
Then you should spell out what you think would be the whole story.
--
David Kastrup, Kriemhildstr. 15, 44793 Bochum
Reply to: