Re: [tlug] JFS file system license

["Stephen J. Turnbull" <stephen@example.com>]

Ian Wells writes:

 > RMS seems clear on what he wants, but has made a complete mess of
 > describing it.

That's because what he wants is not obviously what the law will give
him.  He doesn't dare express that precisely in public.  Thus his use
of legal racketeering to get what he wants anyway.

 > > It's a virtual address space, and it means program code and program-
 > > provided data.  I/O is not part of the work covered by the program's
 > > copyright, although of course it may have its own copyright and
 > > attendant licensing.
 > 
 > A case in point: find me either legal precedent or law that supports
 > this interpretation.  He may have taken legal advice on the subject,
 > but it certainly seems to be 'I say this and it shall be so' from the
 > peanut gallery.

*sigh* You take things much too literally.

The real criterion is that any expressive work fixed in a medium is a
copyrightable work.  "Expressive work" is a term of art in law, and
judges know what it means, and the rest of us don't until they tell us.
OK?

The "exec boundary" rule is not law, but a very convenient rule of
thumb.  It works *because programmers create their expressive works,
by and large, within that boundary*, and because linking, the most
prevalent way of aggregating various software components, works within
that boundary, specifically *by copying various works into a single
medium, thus creating derivative works under copyright law.*

For a counterexample, you could imagine a networked client-server
protocol.  This is (obviously) spread across several virtual address
spaces, yet could be considered a single work.  Or not.  It depends on
the facts of the case.  If the protocol were private and subject to
change at any time on the whim of the authors, and the client does
nothing useful without the server, then you could argue (and RMS has
done so half-heartedly at least once that I've seen) that strong
copyleft could infect clients from the server.  If the protocol is
public, on the other hand, then it makes sense to say that a program
written to implement a client of the protocol is a separate work from
any server implementation.  But without a concrete, very detailed
description, I can't tell you which I think it might be, and of course
a judge might say something quite different.

 > [*] Some days I find it incredible that we have this whole set of
 > ethics with intellectual property which is entirely fabricated.

AFAICS "intellectual property" has no ethical standing, except that
derived from the legitimacy of the legal system that contains it.
It's a purely economic construct.  Bono's widow and her economically-
induced tears ("oh, no, I'm going to be *poor*") notwithstanding....

 > I can create something, and at no loss to me

This is a red herring; it has nothing to do with the economics.  Zero
marginal cost is just a special case of constant marginal cost; all of
the economics works exactly the same whether marginal cost is zero or
positive.

 > I can create an identical copy, and give it to someone else.