Re: [Lynx-dev] Re: GPL

[David Woolley]

> hmm.  It seems that all google is able to show me in this context is not 
> "software industry", but comments on the GPL.  Perhaps you can point to 
> some substantiative material from industry (perhaps not).

It's generally implied by the wording of commercial licences.  If it
weren't the case that a statically linked program was a derivative work
of its components, they wouldn't need to give a royalty free (or with
royalty licence) to distribute software linked against a static version
of a library, and, as the program has a symmetrical relationship to the
library routine, linking it with the library would remove the copyright
on the program!

As I said, the contentious bit is when dynamic linking is used, and
that generally isn't an issue for commercial licences, because they
are trying to restrict the distribution of the library, not its
use by recipients of code linked against it.  In any case,
the issue is about the interpretation of "derived work", for dynamic
linking.  I am pretty sure that the GPL could have been written to
require that effective function should be possible with a library
with a compatible licence.  I haven't studied GPLv3, but it may well
be more specific on this point.

Of course, most commercial licences purport to be contracts, so are
free to be more restrictive than pure licences.