Re: GPL on rendered images

[Brian Thomas Sniffen <bts@alum.mit.edu>]

Andrew Suffield <asuffield@debian.org> writes:

> On Mon, Dec 13, 2004 at 10:17:14PM -0500, Glenn Maynard wrote:
>> The issue isn't whether the conversion itself creates a derivative work,
>> though.  The issue is whether the "preferred form for modification" is
>> that C code, now that I've converted it, stuck the Pascal code in cold
>> storage never to be touched again, and made substantial modifications
>> to to C code.
>
> Those two issues are the same thing. That was the point. While they
> aren't normally defined in terms of each other (neither causes the
> other), they're testing the same things, so they are equivalent.
>
> When the result is a derivative of the original, the preferred form
> for modification will also include the original.

I don't believe this is true.  Consider Linux 2.6, a work derivative
of Linux 2.4.  But only part of the source of Linux 2.4 is included in
Linux 2.6.  Much has been excised.  Only the source for the ultimate
work is required, not the source for all things of which only part is
included.  Very few processes in programming are all-or-nothing
permutations.  Quite a few processes in image and sound development
are all-or-nothing, though, which is part of the problem.

-Brian

-- 
Brian Sniffen                                       bts@alum.mit.edu