Re: Please explain the FSF copyright assignment thing

[Karl Fogel]

Filipe Silva <address@hidden> writes:
>Ok. I get that. But why, why it is legally required? What harm will
>descent upon us all if an official component of emacs is distributed
>which hasn't had it's copyright assigned to FSF?

Well, there's a complex legal history here that I don't know the full details 
of.

However: You can imagine that if the FSF starts distributing some code (say, 
part of Magit) and then one of the authors of that code suddenly says "Wait, I 
never actually said that my code was distributable under GPL-3.0d; you don't 
have my permission to do this!" then the FSF finds itself in the position of 
infringing someone's copyright.

Or suppose the FSF has to enforce the copyright of Emacs in court for some 
reason.  It can only do that if it is the copyright holder.  Therefore, in 
preparation for possible future outbound enforcement, the FSF would like to 
first make sure it *is* the copyright holder.

I am not a lawyer, and I can't say whether these fears/goals are realistic.  
Many free software projects -- most, I think -- have moved away from copyright 
assignment, and toward other mechanisms, such as "developer certificate of 
origin" (DCO).  There is also an intermediate thing called a Contributor 
Licensing Agremeent (CLA) that doesn't actually transfer copyright, but still 
makes the necessary promises to reassure the receiving party (who is also the 
redistributing party).

This topic is complex.  You can read

  
https://julien.ponge.org/blog/developer-certificate-of-origin-versus-contributor-license-agreements/
  https://sfconservancy.org/blog/2014/jun/09/do-not-need-cla/

about CLA versus DCO.  The FSF is one of the few project stewards to still ask 
for an actual assignment of copyright.  I'm not sure why they do, but they've 
been asked before and their continued insistence on doing it is probably based 
on some kind of experience.  The topic only comes up several times a year on 
this list alone :-).

Best regards,
-Karl

>On Jul 13, 2017 15:37, "Karl Fogel" <address@hidden> wrote:
>
>    John Yates <address@hidden> writes:
>    >Perhaps you should ask why Richard decided that Magit, among all
>    >of the many non-FSF copyright assigned packages that emacs users
>    >recommend to one another, was so intolerable as to justify
>    suggesting
>    >mounting a - to my mind doomed - competing project.  It was not
>    as if
>    >all of those recommendations are for some proprietary or
>    non-GPL-V3+
>    >package.
>   
>    Richard didn't do anything special in this case.  He's fine with
>    Magit being distributed as free software, and fine with Emacs
>    users downloading and using Magit.  People are still free to
>    "recommend [Magit] to one another", just as with any free
>    software.
>   
>    The copyright assignment thing is solely about Magit being
>    distributed *as an official component of the GNU Emacs
>    distribution*.  All the packages that are part of Emacs itself
>    have always needed these copyright assignments.  I'm not
>    commenting on whether this is legally necessary or not.  I'm just
>    saying that the request here is exactly the same as the FSF has
>    always made for every other package that would be distributed as
>    part of Emacs.  Magit did not get singled out for special
>    treatment.
>   
>    -K
>