Re: BSD license vs. Evolution assigment (was Re: Copyright assignment)

[David Sugar <dyfet ostel com>]

I believe copyright assignment comes out of limitations in copyright law 
as well as some case precedents.  Let's take a simple example.  Say 6 
different people contribute to a copyrighted work, and one of them then 
creates a proprietary version of the work.  In copyright law, generally 
there is no quanta for anything below the "work as a whole".  That is to 
say, it would be hard for a judge to rule that lines x-y belong to 
person 1, lines a-b belong to 2, or that person 1 has a x% copyright 
"interest", person 2 has a "y" %, etc.  Rather, copyright cases will 
generally be judged on the "work as a whole".  So what possible outcomes 
could happen?

Well, in the worst case, there are a few legal precedents where each 
contributor was given full rights to the "work as a whole".   This 
outcome would suggest that a court might rule that what the one 
contributor did was legal.

A few precedents from some other cases suggest that none of the other 
individual contributors may even have legal standing to sue individually 
unless all of the (remaining) contributors choose to do so.  A related 
example of this was shown with the trouble Mozilla had when trying to 
change license, and the need to get the approval of all past contributors.

So we are stuck with a legal system where copyright has a rather low 
level of granularity while software can have a large number of 
contributors.  Copyright assignment as such solves this problem by 
assuring there is a limited number of copyright holders and that there 
is one (or more) entities that have clear legal standing rather, than 
having to locate each and every contributor.  Of course, it is also true 
copyright assignment depends on trust.

Personally while I generally believe in the value of copyright 
assignment, I also believe it is best to have copyright held by more 
than just one trustworthy entity.  Two or three is perhaps the right 
number, as its small enough where its easy to establish legal standing, 
and it provides a natural counterbalance against the possibility of any 
one entity later choosing to take code private.

Danilo Šegan wrote:

> Yesterday at 22:20, Elijah P. Newren wrote (on d-d-l):
>  
>
> > > With all due respect - but this is UTTER TOTAL CRACK. The situation between 
> > > a project that is based on BSD / MIT licence (like say libxml unless i'm
> > > misremembering what Daniel did with the licencing situation) and a copyright
> > > assignment giving a way for proprietary versions of code are entirely different.
> > >      
> > Entirely different?  No, it isn't.  I admit that there is one difference
> > (anyone can make the BSD code proprietary, whereas only one entity can
> > make the code proprietary with copyright assignment).  
> >    
>
> In a sense, this is exactly the point.  With BSD-licensed code, you
> get a sense of trust — there's no single "privileged" entity, so the
> company who contributed most tells you that you're as good as they
> are: you can make *their* code proprietary, in the same way they can
> make yours.  So, there's a symmetry.
>
> With pure GPL (without assignment), there's also a symmetry: each
> contributor is able to later relicense code they wrote under different
> license, but they cannot relicense stuff they didn't write.
>
> With GPL + assignment, we get many risks.  So far, I have developed
> enough trust only in FSF, perhaps Gnome Foundation (but it doesn't yet
> accept copyright assignment).
>
> Of course, none of this should block Evolution from getting included
> in the desktop: as soon as we get enough significant contributors not
> willing to sign copyright form, we'd probably get a fork we'd used in
> desktop.
>
> I'm sorry this discussion never moved off d-d-l, where it is clearly
> out-of-place.  I won't respond to d-d-l threads any more.
>
> Cheers,
> Danilo
> _______________________________________________
> foundation-list mailing list
> foundation-list gnome org
> http://mail.gnome.org/mailman/listinfo/foundation-list
>
>