[Foundation-l] Copyright issues of wikimedia projects

[Toby Bartels]

Daniel Mayer (maveric149) wrote:

>Toby Bartels wrote:

>>It's all well and good to speak metaphorically of the "life"
>>of a work of art, or of an idea, or of a meme, and so forth.
>>But when I spoke of freedom above, I was speaking ''literally''.
>>Freedom is a property of living creatures like human beings,
>>not of inanimate things like encyclopaedia articles
>>(except, of course, in a metaphorical sense).

>Knowledge should not be owned, nor should the expression of that knowledge,
>IMO. But we live in a world where the norm *is* ownership of that expression,
>so we play the game but reverse the rules in order to free the content. 

Yes, I understand the reasons for copyleft.
We limit minor freedoms of the document's readers
(like combining with another document that has a different copyleft licence)
in order to keep major freedoms for everybody
(like distributing a derived work based on the original document).
The copylefter reserves copyright rights only for the purpose
of preventing other people from reserving copyright rights,
analogously to using violence only in self defence.
(Personally, I don't use violence even in self defence,
and I don't reserve any copyrights to any of my work.
But I refuse to condemn those who choose differently from me.)

>>Mav wrote:

>>>If derivative works are not ensured freedom from
>>>proprietary control, then there is much less of a positive feedback loop
>>>improving that content.

>>And this statement is FALSE -- or at least not necessarily true.

>The statement is true since all improvements that create derivative works from
>copyleft content can be re-incorporated back into the original. That is
>positive feedback. However if the original were under an attribution license,
>then many of those derivative works will be proprietary and thus any
>improvements made to them could not be re-incorporated back into the original.
>That would not be positive feedback.

I understand this argument, but it is not valid as it stands.
To make this argument work, you have to explain why
a copyleft document will always attract as much derivative work
as the same document distributed under a free non-copyleft licence.

This is certainly not true in general!  Taking the Snow White example,
the Walt Disney Company, given its profit motives, would not make
as many derivative works from public domain (hence free but not copyleft)
fairy tales if its derivative movies had to be free as well.
Yet the free fairy tale community gets some feedback from Disney,
since the Disney characters are now available for some uses in parodies.
This is an extreme example, and all things considered,
I would be much happier if Disney didn't make any movies,
if that meant that the movies that other people ''did'' make
would all be free.  But I will not pretend that copyright
serves no purpose at all -- it ''does'' encourage creation
by people motivated by profit (as we all are to some extent).

>Let's not just assume that copyleft content would fail to attract attention.

I am not making this assumption.  I am only considering the ''possibility''
(and that only relatively -- copyleft might attract ''less'' attention).
You're assuming that a copyleft document will attract as much attention
as a free-but-not-copyleft document; that is the only way that your argument
beginning "The statement is true [...]" above will go through.
I am refusing to make any assumption one way or another.
Thus I am not advocating that WikiNews use a non-copyleft licence.
I am only advocating that the Wikimedia Foundation adopt a policy
that (implicitly is enough) allows for a non-copyleft (but free) licence.

I like this text suggested by Erik (Eloquence):
<http://mail.wikipedia.org/pipermail/foundation-l/2004-May/000222.html>

"All text on Wikimedia projects will be under a license which allows free
distribution and modification by anyone. The image copyright policies are
set by individual projects."

(I guess that I'd like the image language to be somewhat stronger,
but I'm not really sure how to improve it -- that's famously tricky.)

>*If* a copyleft Wikinews project did not do well
>and we suspected it was due to the license, then we could talk about the
>possibility of changing. Due to the nature of news articles, a future change
>of license would not be such a big deal.

This is true.  I'm not very worried about how WikiNews is licensed;
but it's fresh ground for coming up with hypothetical scenarios.

>But I completely see a point about the GNU FDL - a printed newspaper could
>never abide the FDL's requirement to include a copy of the license.

I'm not sure that it's even that bad.  Standard policy with the GNU GPL
is to place the license in a separate file, and the GNU FDL allows aggregates,
so a newspaper should be able to include several FDL articles in its pages,
each with the brief statement that its licence may be found on page A2.

>The CC
>by-sa with an upgrade clause (which is oddly not part of the 1.0 version)
>or my proposal for a GNU FCL would be much better suited for that.

The CC 1.0 licences had several annoying flaws, like no upgrade clause.
(Similarly, the GNU 1.0 licences had several annoying flaws,
but they are much older so we don't remember that. ^_^)
The CC-sa 2.0 licences now do have an upgrade clause, like the GNU FDL
(which of course is a very good thing for your GNU FCL proposal).

>Since Wikinews content would primarily flow to, instead of from Wikipedia, an
>interim solution could be to dual license Wikinews articles under both the GNU
>FDL (for one way compatibility to Wikipedia) and the CC by-sa (for a
>lightweight copyleft license that could be used by print media). Downstream
>users would have to choose just one of the two licenses. This would fork
>derivative work development between the two copyleft licenses and none of that
>work could be re-incorporated back into the original dual-licensed article, but
>we plan to fix the incompatibility issue anyway (that is why this would be an
>interim solution).

I wouldn't want to rely on being able to fix the incompatibility,
so I'd want any interim solution to be viable indefinitely.
That said, this solution probably ''is'' viable indefinitely.
If a dual-licensed project ever wants to change to a single licence,
then it can do so any time that the users wish.

>>>The GNU FDL is copyleft, so IMO anything 'like it' must also be copyleft
>>>(such as the CC by-sa).

>>That's obviously not an absolute statement;
>>the English term "like" is very vague.

>Exactly, thus I stated 'IMO.' It is my interpretation

OK, fair enough.
The relevance of the vaguness is that the statement as it stands
is not explicit, so that your interpretation isn't the only possibility.
But this thread began with Anthère asking for a clarification,
so we're discussing how the statement ought to be clarified.
If it's clarified, then there will no longer be any vagueness.

>I will not support any license whereby a downstream user could make a
>derivative work of our content and be able to not grant us the same freedoms
>over that work.

Of course, I might support one (and indeed have, on various occasions
unconnected with Wikimedia).  Someday I may even try to convince you to.
But right now, I only want to convince you that the Wikimedia Foundation
shall not adopt a policy that it will ''never'' support one.


-- Toby