On Oct 19, 9:47 am, "Alfred M. Szmidt" <[EMAIL PROTECTED]> wrote:
> I am not a lawyer, so I can only offer a "common sense" opinion: > > Which sadly, is not much common sense. I see GNU likes civilized discourse. > When you choose dynamic linking, you are not including the library, > but only its interface in your product. The users of your product > may or may not opt to use it with the library in question. > > You are including code as well, macros for example. If they are part of the interface (e.g. some convenience macros), and the interface is not owned by the library author, how does this even modify anything I said? OTOH if the macro usage is an "interface", but their implementation is not, the situation is similar to C++ templates and static linking. > The > binary is also combined into one big blob when run, But not even by me. The user chooses to combine it. I don't. For all I care, the user may choose to combine it with some other library, which he may even write himself, with a compatible interface. > which means that > it does not only share `interface', but memory and everything else as > would be done during static linking. P.S. Please reply to newsgroup. _______________________________________________ gnu-misc-discuss mailing list gnu-misc-discuss@gnu.org http://lists.gnu.org/mailman/listinfo/gnu-misc-discuss
