A colleague asked me the following question:
For software produced here (UMichigan) that is going to be released under an open source license, who should be listed as the copyright holder in the source code? Would it be “The University of Michigan”?
Here is my answer:
I am not a lawyer but I do a lot of Open Source stuff in many projects. This is what I do and what people in Sakai generally do.
(1) If the project is primarily written by me from scratch, I use Apache 2.0 as the copyright in my own name. Anyone can reuse or include the code I build (including me) because of the license with a simple acknowledgement – with the right license the “owner” does not matter. If someone wants to switch the copyright to a GPL license – the copyright owner can give that permission and one person is easier to ask than to ask a university. The answer may be yes or no with a person – with a university you are most likely never to get any answer. Using your own name is not seen as arrogance unless someone asks if they can change the copyright and you say ‘no’. An example of this is if I build a new experimental part of Sakai I will initially make it copyright me Apache 2.0 – if the code matures and ends up in the core of Sakai I voluntarily change the license to be Copyright the Sakai Foundation using the Educational Community License 2.0 as that is the preferred copyright for Sakai core source and it is considered tacky and arrogant to maintain your name in the copyright of Sakai core code. I emphasize that it is not tacky to use your own name initially. It *is* considered tacky to refuse to allow the name to be changed when asked by a reasonable open source organization / project.
(2) If I am writing in the context of a project like Sakai, Moodle, or ATutor I simply use the copyright of the rest of the code with a clear indication of the extent of my authorship where appropriate. If you say anything about keeping your own name in the copyright in their code base – they will punt you as far as they can. They will kick you out of the project and talk about you behind your back for years to come.
(3) If I am writing a tool for an organization that I am doing consulting for, I insist on the code being copyright them, but Apache 2.0 – so I and others can reuse the code with no further permission except for an acknowledgement. I have not faced a situation where a client insists on me building significant code that is not open source – but frankly I would likely not even be willing to write code for such an organization. I do advise closed-source organizations on how they should go about writing code – I just won’t write their code.
(4) If I am building a tool where multiple UMich authors are involved and in particular paid staff are involved and neither (1), (2) or (3) apply, I copyright the code University of Michigan Apache 2.0.
Interestingly, w.r.t. Sakai, UM has a master agreement that lets them take Michigan contributions and change the copyright to Sakai Foundation. Indiana, and Stanford have similar master agreements. To my knowledge, MIT and Berkeley never executed such agreements and so bits of code still have those universities named in copyright statements in the code. You will not see a Stanford, Indiana or Michigan copyright anywhere in Sakai code – that is considered a badge of honor by the cool tribes. Interestingly, deep in one tool, we use some Stanford code for a concurrent hash map that was built well before the Sakai Project and used in Sakai. We did *not* request change the copyright in that utility code because it was clearly not produced as a contribution to the Sakai project.
All this is evidenced by the Sakai acknowledgements page:
The most awesome and giving schools are in the first list because they are the ones that contributed and transferred copyright or simply contributed with the foundation’s name as copyright in their contributions. They of course are in alphabetical order so as not to characterize any contribution as more important than any of the others.
Probably more than you wanted to know :)