On Tue, Aug 05, 2003 at 05:13:15PM -0500, Steve Langasek wrote: > On Tue, Aug 05, 2003 at 10:17:54PM +0100, Andrew Suffield wrote: > > > > IMO the GPL is purposefully vague on this point; if someone (not just > > > the copyright holder) can show reasonably that they preferred a certain > > > form for modification, then they've met the terms of the GPL. > > > This means that I can take your GPLed application, modify the binary > > directly (I prefer this form because it is required by my business > > model), and redistribute that - without the source to my changes, > > since I don't prefer that form. > > > I reject all interpretations which lead to this result as > > fundamentally flawed. > > I reject all interpretations which prohibit this result as fundamentally > flawed. The GPL is not a hastily written license; every word is > carefully chosen with the intent of promulgating Free Software, in *all* > its forms. If I make non-trivial modifications to a binary, my > masochism should not make me ineligible to distribute my work under the > GPL. If you can have this, then you can trivially extend it to generating those modifications with another tool - say, a compiler - and then patching them into the binary by hand. This actually quite simple, so would form a gaping hole in the GPL, especially if a) the GPled application is written in C, and b) you are working at the level of entire functions. This remains the form you prefer for modification, because it's the only one which supports your business model - even though at times you work with a form you don't prefer (C source). I do not believe that a court would accept this as a reasonable interpretation of the intent of the copyright holder, and nor do I accept that they would permit the licensee to define the terms in a license - that would lead to absurdity. -- .''`. ** Debian GNU/Linux ** | Andrew Suffield : :' : http://www.debian.org/ | `. `' | `- -><- |
Attachment:
pgpCBAfGSxuBZ.pgp
Description: PGP signature