On Mon, Jun 11, 2012 at 12:43 AM, Bruce Perens <[email protected]> wrote: > What legal theory would make a user of an API a derivative work if the API > is not itself copyrightable? > > If there was a case like MySQL v. Nusphere without the contract, this is what I'd argue. Note I'd avoid saying "derivative" like the plague. I'd point out (assuming the following is true for sake of argument):
There's ample documentation that the licensor intended this license not to reach compiled or collected works linking this software to proprietary components, and that extra licenses were required at that point. There's ample evidence the licensee was aware of all this. I'd then argue that whether or not it's a derivative work is not at issue. What is at issue is whether the licensor intended the license to allow the behavior in question and whether the licensee knew or should have known this. By distributing the code, they need copyright permission. That's not in dispute. Therefore, the behavior is outside the GPL and at least contract violation if not copyright violation. This may not be a derivative work but it's not really permitted by the GPL. Compiled/collected works require permission to and some of these are "based on" in the view of the GPL even if they are not "based on" in the way that term is used in copyright statutes. I don't know if that's a winning argument (assuming the ample documentation is there). But if I wanted to argue it, that's the case I'd make. Best Wishes, Chris Travers _______________________________________________ License-discuss mailing list [email protected] http://projects.opensource.org/cgi-bin/mailman/listinfo/license-discuss
