what does constitute "linking"?

[Bradley M. Kuhn]

Allan Hardy wrote at 15:53 (EDT) on Thursday:

> That said, your question isn't really about linking, its about what
> constitutes a derived work, 

Quite correct.

> This is a very old question and it's been discussed I'm sure thousands
> of times.

Indeed.  It comes up, almost on cue, on some email list related to the
GPL every few months.  This list appears to be "the place" this quarter
of the year. ;)

> You are in an area where there is no case law, no judges rulings to go
> by, for the most part.

Actually, there is various case law in the United States.  What I think
you mean to say is that no court has applied this case law on copyright
derivative works of software to a situation involving the GPL.  BTW,
this old interview with my colleague, Dan Ravicher, might be of
interest on this subject:   http://lwn.net/Articles/62202/


My advice to Janez is to get a lawyer who specializes in
software copyright and FLOSS and ask them this question.  Asking a list
like this will give you a bunch of information, none of which will help
you analyze the legal risk of what you seek to do, since it won't be
legal advice.

Meanwhile, I think other posters have a good point: Why don't you just do the
right thing *ethically* regardless of the specific license requirements, and
share your software under a license that respects your freedom and
everyone else's?

I'm often amazed at the trouble and effort people go to when seeking the
legal boundaries of copyright derivative works.  I suspect that more
energy is expended sometimes on that consideration than is on the
proprietary code itself that might take advantage of the edges, if such
edges can even be determined.

BTW, IANAL and none of this is legal advice.
-- 

   -- bkuhn