[OSGeo-Discuss] Software Copyright ownership

[Christopher Schmidt]

On Sun, Dec 13, 2009 at 03:58:01PM -0800, Arnulf Christl wrote:
> What do folks think about software Copyright ownership? OSGeo could  
> suggest that project steering committees move the Copyright of their  
> software under the hood of OSGeo as GeoTools and others already did. In  
> some cases the respective project steering committees might not be able  
> to do such a thing because they do not own it in the first place. Is  
> that a good situation?
>
>
> To my knowledge (this needs verification) Apache has all its projects  
> under their own license and owns the copyright for the code. 

To be honest, after reviewing the way the Apache Foundation works, it's
not atall clear to me that we have a strong desire to follow their model.

A brief search turned up:
 * No list of contributors to the project
 * No obvious maintainership of a list of sources of code
 * No obvious documents indicating copyright assignment of any kind 
   to the Apache Foundation.
 * No obvious requirements for Copyright Assignment (Only License 
   Assignment, via CLA.) 

> Very  
> straightforward. Outside of Open Source in the standards arena Google  
> has given the Copyright of KML to the OGC - which is good. 

It's not clear to me to what extent 'KML' has a copyright; Google states that
"KML is an open standard officially named the OpenGIS® KML Encoding Standard
(OGC KML). It is maintained by the Open Geospatial Consortium, Inc. (OGC). "

It is not clear to me that maintainership of a specification document
is at all equivilant to copyright of code.

> In  OpenStreetMap not doing this has resulted in arguably unresolvable
> licensing problems.

In OpenStreetMap, having all contributors agree to place data under
a copyright license has caused a problem. ALthough if the data were owned
by some entity, that entity would be able to change the license, it is
not clear to me that the problem is that this wasn't the case; instead
the problem was that the license chosen was not the right one (though it
seemed the best option at the time) because it didn't do what people
thought it did.

> A recent example that shows what problems can be caused by not clearly  
> separating Copyright, business ownership and trademarks is MySQL. The  
> company "MySQL AB" operated the project "MySQL" and sold the product  
> "MySQL" under a dual licensing schema. Now MySQL AB (the company) is  
> owned by Oracle. Who owns the copyright of the project MySQL? And what  
> happens to the trademarks?
>
> I generally do not agree with Monty[1] (twittered by James Fee) and  
> believe that he has other motivations and should have taken preventive  
> measures up front to avoid what is happening right now. But whatever the  
> outcome of this fight it for sure has been damaging to the project.
>
> In a response/comment from Groklaw (twittered by Paul Ramsey):
> http://www.groklaw.net/article.php?story=20091021164738392
>
> "And on what legal basis would anyone have authority to change the  
> license, other than the copyright holder? Are you seriously suggesting  
> that a regulatory body decide the license instead of the copyright  
> owner? What a reckless idea."
>
>
> My private opinion on this issue is pretty clear: Move your Copyright to  
> OSGeo - all of it including trademarks, logos and designs. That is what  
> OSGeo is there for. Get it out of corporate reach, it is none of their  
> business (great analogy, hehe). 

> Is their any advantage of keeping the  Copyright under a private property?

Depends. There may be more trust in some private properties than others. 
So far as I'm aware, OGC is a private property, you you argue that putting
KML under OGC was a good thing.

OSGeo is also a private property. It is a foundation, managed by an elected
board -- but so are most companies. (OSgeo isn't even, so far as I know,
a registered nonprofit organization at this time.) What makes OSGeo a better
steward for code than organizations which have managed code for years --
or in the cases of some projects, decades?

> What I get back from corporate users of Open Source software these days  
> is the same, they would rather have the Copyright sit with a (real)  
> non-profit like OSGeo than anything else.

How is OSGeo a real non-profit?  

I don't see a strong reason to change the methods of projects that have
successfully managed many years of code contributions. The best people
to make those decisions are people who have successfully managed those
projects.

It's great that OSGeo now feels comfortable managing the copyright of 
projects, but it's not clear to me what that actually means. Who is the
person who controls the copyright? Who makes decisions about how it is
managed -- and what happens if someone disagrees with those decisions?

I think that it would be lovely to create an environment where projects
feel that giving copyright over to OSGeo makes their lives -- as project
managers -- easier. I'm not convinced that is currently the case; the lack
of obvious documentation on how projects should give copyright to OSGeo,
and what it means when it happens, seems to me like it creates a void in
which projects might feel uncomfortable about giving copyright to OSGeo,
for fear of what that might mean. Improving that, through solid documentation,
seems a great first step in making projects feel more comfortable with
that process; this is certainly true for me as a contributor to OpenLayers.

Best Regards,
-- 
Christopher Schmidt
Web Developer