Re: Plugins, libraries, licenses and Debian

[Jeremy Hankins <nowan@nowan.org>]

mru@kth.se (Måns Rullgård) writes:
> Jeremy Hankins <nowan@nowan.org> writes:

>> If you want a simply answer, the answer is: "No (insert disclaimers
>> here)" as others have pointed out.
>
> As someone said, writing is always allowed, it's distribution that's
> restricted.

True as far as the GPL is concerned, but not true generally.

> There have been some indications that a source
> distribution is allowed, even if a binary distribution is not.  Could
> someone clarify?

I must have missed the message that talked about this.  My understanding
is that the only case this might apply is when the source isn't actually
intended for compilation (e.g., it's in book form).  The idea is that if
I distribute a work in source form that requires an incompatible lib I'm
simply trying to get around the law, and generally courts don't like
that.

>> The rest of the discussion is only appropriate if you want to
>> understand why that is.  But it has to do with intent, sneaky ways
>> one might try to get around the GPL, how provable your position is in
>> court, and (perhaps most importantly) how deep your pockets are.
>
> I use plugins for purely technical reasons.  If, as a side effect,
> otherwise incompatible libraries can be used, it's all the better for
> the users of the program.
>
> I don't generally trust courts, so I'd rather not end up there.

Then you (like most of the rest of the world) are best off taking the
conservative view and assuming that they're still incompatible even as
plugins, even if you think you have a valid argument to the contrary.

> Once again, we end up at the words "derived work".  Where should I
> look for precise definitions this term?  For the record, I am doing
> this work in Sweden and Norway, in case it makes a difference.

You could look at the law, but I'd be surprised if it were specific
enough to answer your question.  It certainly isn't in the US; here
you'd need to look at case law, and even there you're not going to get a
clear answer because there isn't one.  Frankly, the most comprehensive
attempt to answer the question that I know of is on the FSF site, though
they clearly do have an angle.

-- 
Jeremy Hankins <nowan@nowan.org>
PGP fingerprint: 748F 4D16 538E 75D6 8333  9E10 D212 B5ED 37D0 0A03