>> The Free Software Foundation [...] asserts that an executable that uses a dynamically linked library is indeed a derived work. Which if you think about it i

>> The Free Software Foundation [...] asserts that an executable that uses a dynamically linked library is indeed a derived work. Which if you think about it is really a very problematic assertion, for a couple of reasons. First, it doesn't seem to be related at all the the definition of "derived work" in the Copyright Act (in the US) or the examples given for other kinds of derived works, which all involve incorporating creative elements from the original work into the derived work. These elements might be translated or rearranged or otherwise transformed, but they're still necessary for the work to count as derivative. Linking against a dynamic library has more in common with referencing another work in your bibliography—which so far as I am aware has never been considered enough on its own to consider the referencing document a derivative work. But more importantly: The FSF also maintains that a clean-room, from-scratch open-source reimplementation of a proprietary library with a compatible ABI is not a derived work of the proprietary library. (For example Lesstif vs. Motif, or Wine vs. Windows.) A program which is dynamically linked against some library is only dependent on the library's ABI, not the specific implementation of the library, and can be used at runtime with either version. So is it a "derivative work" of the original proprietary library, or the open-source reimplementation? And of course the same logic could be applied the other way around for a program which can be (but need not be) dynamically linked against a GPL library—it could just as easily be linked against a proprietary reimplementation of the same library with a compatible ABI. And why should it make any difference for the purpose of declaring something a "derivative work" whether the reimplementation under a compatible license already exists, or merely could exist? The part which is asserted to be "derivative" hasn't changed at all. I am not (a / your) lawyer, yada yada, but the "dynamic linking implies derivative work" argument has always struck me as being at odds with both the plain meaning of "derivative work" and actual practice when the situation is reversed.