[OSGeo-Discuss] role of foundation with regard to licensing

[Christopher Schmidt]

On Wed, Nov 18, 2009 at 11:12:43AM -0800, Tyler Mitchell wrote:
> On Wed, 18 Nov 2009 11:16:13 -0500
> Frank Warmerdam <warmerdam at pobox.com> wrote:
> 
> > >> I know that in practice, this is probably the way things already
> > >> are. Why rock the boat?  Why assign copyright to OSGeo in the
> > >> first place [2]?  
> > 
> > The primary reason to assign copyright to OSGeo is to make it easier
> > to relicense in the future.  It is very hard to relicense a project
> > with copyright held by many contributors.
> > 
> > There are also reasons not to assign license, foremost being the
> > paperwork overhead involved in contributions agreements for all
> > contributors.  Some contributors are also hesitant to surrender their
> > control over their contribution.
> 
> A few questions about copyright have come across my desk or
> face-to-face at events this year.  Frank, for the sake of
> others on the list, could you give us an overview of what does it mean
> to be an "OSGeo project" if OSGeo itself does not hold the copyright?  
> 
> I think the question was geared toward whether or not OSGeo could
> guarantee future appropriate licensing of a product that
> it has arms-length influence over - or would a non-complying project
> then be rejected somehow?

First, once code is released under a given license, that license can't
later be 'removed' in any meaningful way: OSGeo will always be able to
maintain and distribute code which was openly licensed at any point, 
which is a requirement of becoming an incubated project. As far as 
that goes, OSGeo could always continue to provide a home for open source
code that has ever gone through the OSGeo incubation process.

If a project were to attempt to 'rescind' its openly licensed status,
I believe that it would be the job of OSGeo to: 
 
 1. Work to prevent such a change from happening. Given the communities
    we're working with, I think there would be very strong social
    pressure against any incubated project going from open source to
    closed source; changes like these typically simply end up leading to
    a fork, and OSGeo could continue to provide a home to a community
    built around the open source project.

 2. If all else failed, it would be possible for OSGeo to maintain
    resources for the open source code, but would probably do best to
    retire the project, similar to how MapBuilder was retired (but
    obviously for different reasons).    

Being an OSGeo project means that the project is a participant in the
OSGeo community. Sharing information, collaboration, and resources with
other projects in the foundation is the primary motivation for OSGeo
projects to continue to participate.

OSGeo is not a controlling foundation; in this way, it is somewhat
unlike the Apache Software Foundation and the Free Software Foundation,
which take a direct ownership over the projects. This means that OSGeo's
role in exerting control over the direction the project takes is
limited. However, OSGeo's role as a parent organization means that OSGeo
can act as a shepard to code/projects, regardless of the directions that
may be taken otherwise.

Becoming an OSGeo project means putting your project out there, and
participating in a shared community. As a result, you get to exert some
control on OSGeo, and OSGeo works with the project to help it succeed.
If a project were to take a path away from open source, OSGeo would act
as a shepard for the project unless it was no longer in a position where
it made sense to do so, at which point the project would no longer be a
participating project in OSGeo.

This is just what seems to me to be the most reasonable and logical
approach to the situation as it stands today.

Best Regards,
-- 
Christopher Schmidt
Web Developer