On Wed, Feb 16, 2000 at 11:07:28AM +1100, Don Sanders wrote: > After thinking about the the work based on the Program issue some more I've > decided everything I wrote originally is correct. Any response would be > appreciated. ... > If you define the work as I have then reading through the terms of the GPL > sequentially starting at Section 0, continuing on to Section 1 and then > proceeding to Section 2 I think we all agree that the terms of the GPL can be > satisfied. > > It is not until Section 3 is reached where under your interpretation (but not > mine) the Program is redefined to be the complete source code that there is a > problem.
I disagree with this point. When a modified work is created (including a binary), you now have two different works and you must consider their licenses independently. Where you say "the Program is redefined," I'd say "a new work is created which is also licensed under the GPL." Where you say "the Program remains constant," I'd say "the old work is also licensed under the GPL". Unless you claim that it's possible to distribute the new work (created when the original work was modified) without a license, I don't see how your argument holds together. -- Raul
