Re: Evolution copyright assignment: Storm in a teacup

[Daniel Veillard <veillard redhat com>]

On Sun, Aug 08, 2004 at 01:39:13AM -0400, Richard Stallman wrote:
>     People then can choose to use Evolution of the forked Evolution.
> 
>     The same exact scenario would happen with say, Libxml: if someone
>     insisted that their contribution to Libxml had to be under the GPL and
>     Daniel (the maintainer) refuses to accept it, they can always fork the
>     tree and have their own improved edition.
> 
> In principle, Daniel could do this.  But is that really likely to
> happen?  I don't think so.  Are contributors to libxml currently being
> asked to assign copyright?  If not, then most of them are doing

  no, but let's forget the specific details of that project. In general
the whole GNOME project has been build without any Copyright assignment.

> precisely this--and Daniel is not rejecting their contributions
> on account of this.  He's not even talking about doing so.

  Right, but I think miguel was just looking at an example.

<specific>
For libxml2 we changed the Licence 2+ years ago, there was still at
the time a limited set of relatively well identified contributor. In
practice trying to go though the exercice of contacting every one of
them now would be quite difficult (even though all contributors and
contributions are logged in the ChangeLog).
</specific>

  I don't want to discuss the legal aspects of Copyright assignment myself,
but I think it is interesting to try to analyze their effect on project
development, especially dynamics. They are relatively heavy and basically
put a barrier of entry on anybody wishing to contribute to a project. This
pushes more of the work on the core developper: one can make a bug report
without signing a Copyright assignment but making a patch and getting it
accepted requires to go though that paper/printing/reading/sending loop,
most people won't bother, as a net effect I think it reduce the prospect
of growing the community of developpers. On the other hand some things
are easier on the long term if you have collected all Copyright assignments.

  Seems to me it is worth only for very large projects where the duration
of use is expected to be decades and building a replacement in case of legal
problem would be near impossible. The GNU libc and compilers as well as
Mozilla gecko engine are good examples, with dozens+ man-year of work put into
those and reused a lot. Seems to me GNOME has been developped in a far less
monolithic fashion and except for glib/gtk+ none of the libraries is really
critical to the point of not being replaced, we actually have a trend of
discading/rewriting some of our components at every major release.

  The board will have to look at this issue, I hope they won't only limit
their analysis to the legal aspects but also the developments and momentum
aspects of including Evolution as part of the GNOME release things like:
   - integration within the desktop
   - existing momentum (usage and developpers)
   - scripting APIs
   - (potential) requirement of C#/Mono
   - impact on other lightweight or more modular mail/calendaring projects
are to me as important for the project as the Copyright assignment point.

To me as long as we have garantees that the code will remain available as
free software and the mechanisms set up to integrate with the whole desktop
are generic enough that other applications can use them too, I think 
integrating Evolution as part of the GNOME release will be an excellent idea,
but this is just a one person opinion.

Daniel


-- 
Daniel Veillard      | Red Hat Desktop team http://redhat.com/
veillard redhat com  | libxml GNOME XML XSLT toolkit  http://xmlsoft.org/
http://veillard.com/ | Rpmfind RPM search engine http://rpmfind.net/