[Foundation-l] Re: [Wikitech-l] Re: licensing issues, BSD doc license vs GNU

Thu Jan 6 20:40:35 UTC 2005

On Wed, Jan 05, 2005 at 05:49:46AM -0800, Jimmy (Jimbo) Wales wrote:
> Edward Peschko wrote:
> > Right.. I understand the distinction between the software's license
> > and the content's license - but what I was wondering is does 
> > the wikimedia *foundation* have a policy against accepting a 
> > sister project's content with a different license than the Gnu FDL? 
> 
> I'm not sure what you mean here by "sister project".  That term is
> something that we reserve for all of our internal projects.  There is
> no process whereby an outside project can become formally recognized
> by us in anyway, and I'm not particularly keen on doing so, just
> because of the time and energy that would be required for a process of
> scrutiny.
> 
> So far, all of our projects have been GNU FDL.  Wikinews is currently
> using public domain, because there is an ongoing discussion and
> learning process about what sorts of things we might need to do in
> that case.
> 
> There are projects which are friends, but it is informal.  Wikitravel
> for example, is not a wikimedia project, but we are friends with Evan.
> 
> --Jimbo

I understand the reasons for both the preference for GNU FDL and 
the 'friends vs sister projects' distinction, but I believe the project 
in question warrants some exceptions. Its called 'wikiresearch' 
(bring the power of wikis to research projects) - and I think that there 
three basic reasons why the foundation should make an exception on both fronts 
in this case:

    1) 'wikiresearch' is about as generic a project as they come. There
       have been rumblings about it on wikibooks (where it is listed 
       as a 'candidate for deletion' because it doesn't fit the wikibook
       mold. Hence, ultimately it makes sense as a foundation project.

    2) mediawiki as it stands is not a 'good fit' for research projects, both
       software wise and license wise.

       For example, the GFDL license and complete openness would scare away 
       most researchers - they want to control how their content is modified
       and viewed. For them, the more important thing is to get *open comment*
       on their works, not open edits - as well as control *when* it is made 
       public.  This requires IMO something more like the BSD license 
       (which can be sublicensed), as well as mods to the mediawiki software
       to allow people to have their own subprojects which they can admin.

       Likewise, they would want to ultimately take the content that they've 
       created online and publish it, as well as be able to get people to cite
       it.  Again, this requires mediawiki mods.

    3) If this is truly going to work, it is going to need three things: as much 
       freedom as possible to 'get it right' in the form of experiments,
       close working w/mediawiki developers to make sure that any mods made to 
       mediawiki could ultimately be reincorporated back into the main 
       mediawiki branch, and as much exposure of wikiresearch as possible 
       to the outside world by its ultimate inclusion as a sister project. 

In short, I'd like to be able to experiment with mediawiki and modify it so 
that it fits the research problem domain, try it versus a test research 
project, and treat all this as a 'dry run' - and then take the work that 
was done and encorporate it back into the main branch. 

However, if not using the GFDL would interefere with this, or people have 
qualms with using the BSD license && sublicensing wiki content, I'd be 
willing to discuss it... And of course comments on any of the above 
are most welcome...

Ed

(
    ps - 
    in any case, I've applied for two sourceforge projects: wikiresearch 
    and energywiki... I'll let people know when they come through and there
    are mailing lists available..
)