Salon: Public money, private code

If you're a student at one of these benighted Universities, and you have an idea for a project you'd like to be sure they can't abscond with, you might try this:

Get somebody else who is not encumbered to initiate the project. Get them to create a project on SourceForge or somewhere, with a prominently explicit GPL license. Contribute to the project, designing and implementing the publication-worthy "hard parts" yourself. Make sure your adviser is documentably aware of the terms of participation in the project (e.g. include the project docs as an appendix in your progress reports).

Ideally the public project would be bigger than just your own contribution, so that the recognizably useful product includes others' work, while the academically interesting part is yours alone. Instead of initiating a project, you may be able to get the same effect simply by contributing to an existing project.

This doesn't help projects that are already mired, but it might help keep new projects from getting mired.

. . .

Another, unrelated remark: it was pointed out that it's hard to point to projects that would have flowered if only they had been unencumbered. However, it is much easier to point to projects that didn't flower until they finally got free. We need to tabulate those.

At UWA, the policy is that students own the work they do (of course, you wouldn't own contributions from your supervisor). So in most cases, there would be no problem releasing something as free software. I don't know if this is standard practice for Australian universities though.

Stanford has a policy that all copyrighted material is owned by the creator (i.e., me) unless it's specifically work for hire (which work done under an RAship normally isn't.) They also have a good patent policy: any invention may be placed in the public domain. (Stanford will take a chunk if I license it out for profit, though.)

Although this flexibility would allow me to take publicly-funded research private, it also doesn't stand in the way of making my work free, either. I'd rather have say over what happens with my creations rather than have the university decide one way or another.

One of the reasons I am self employed is that I've long intended to develop my own software products, and I don't want an employer (academic or private industry) to try to take ownership of my work.

I do release some stuff freely (not as much as I would like) and this avoids the same problem.

Usually when I am consulting, the work I do for the client is a work for hire, so I don't get to keep the rights to it, but it is also clear they don't have any right to work I do on my own time.

I distantly remember that CalTech had an annoying patent agreement that we all had to sign before we started studies there. But get this - when I was a graduate student TA at UC Santa Cruz I had to swear a loyalty oath in front of a UC employee before I could start my employment! I thought that went out in the fifties!