Actually, where the gray area becomes is how much capital the OP has to defend his copyright versus how much the university wishes to take it from him. I have personally witnessed where a company tried to steal the copyright from an employee because they published their work after starting, even though the employee started working on the project prior to joining the company. It became a very hairy mess because the burden of proof was then on the employee to prove that they started the project prior to joining the company. I never did hear how that turned out, as I left the company prior to the issue being resolved (and definitely because of that issue).

I was also on the receiving end of similar treatment with a different organization; however that company's lawyers backed away very quickly. I published the project, including source code and binaries, on SourceForge several years before starting with the company. Thus the chain of ownership was very clear about that project.

But, again, I am not a lawyer. This is just my personal experience with corporate corruption, and I presume that it extends to universities as well, since they are also trying to make a profit.

Where this becomes a "gray area" is if the OP actually *published* or made available the original GPL'd work prior to starting at the Uni. TFA does not directly state if the OP published the original work.

If the OP had not published the original work under the GPL, combined with any applicable employment contracts (including the "Employee Handbook" equivalent), then the university can make a case stating that the copyright belongs to them and they have full rights to determine the license. That is where lawyers and documentation get involved, and the case usually goes to he who has the highest paid legal team.

If the OP published the original work under the GPL on an accessible site, then project creation dates, contracts, et cetera can be subpoenaed by the OP's attorney if the case goes to court, showing a full chain of ownership where the OP licensed the software to the Uni under the terms of the employment contract. However, the employment contact may include an "exclusivity clause" stating that any software licensing of preexisting works is made exclusively to the university, or, if unable to be made exclusive, cannot be incorporated into custom software created for and on behalf of the university. If there is an "exclusivity clause", then the OP may be in a whole heap of legal fecal matter that is on a collision course with the mass airflow device.

Disclaimer : IANAL - Always follow up any information on the internet with a licensed attorney in your jurisdiction.

What about SourceSafe or Autodesk Vault? If I'm not mistaken, both of those applications implement this "new and innovative" technology, and have for years.

Prior art, patent busted. Now the question is whether the USPO will look at that (not likely). But Autodesk and Microsoft should fight the patent (or Microsoft will ignore it just like they're ignoring the GLPv3.