Josh Triplett <[EMAIL PROTECTED]>: > > So before Wine was created, anything which uses a Windows library was a > > derivative of Windows? > > Yes.
There are so many theories on this subject that I am perpetually confused, but I don't think that is what is usually claimed in the case of GPL libraries. I think the usual claim is that the program that uses the library plus the library is a derivative of the library (which is obviously true) and also a single work even when the parts are distributed separately (which is at least plausible). In the case of a typical Windows library that's not a problem because: 1. Only Microsoft and its agents are distributing the library. 2. The library is not available from public servers. 3. There is explicit or implicit permission to link the library with arbitrary programs. However, in the case of a a GPL library it is possible to argue that the person distributing the program is encouraging people to fetch the library from a public server and link it with the program, and therefore that person is in effect distributing the GPL library in an unlicensed manner. There are all kinds of hypothetical circumstances that might strengthen or weaken that argument. For example: 1. The argument seems stronger in a case where the program and the library are being distributed by the same person or by people who are in some way working together. 2. The argument seems weaker in a case where the program is being distributed in source form only together with an explanation that the program is for research use only until there is a compatibly licensed substitute for the library. Obviously I have no idea in what circumstances the argument might be accepted by a court in what jurisdictions, but I think that's roughly what the usual argument is, and it doesn't directly imply anything about the situation with a typical Windows library.
