> On Sun, May 02, 2004 at 11:39:23PM -0400, Raul Miller wrote:
> > I thought someone had said that we might reject programs which violate
> > the spirit of the DFSG even if they seem to comply with the letter?
On Mon, May 03, 2004 at 12:31:38AM -0400, Glenn Maynard wrote:
> You seem to be contesting the spirit of the DFSG, if only as an exercise.
> I think the spirit of DFSG#3 is clearly "must allow all modifications".
If discussing this ambiguity is contesting the spirit of the DFSG then
that is indeed what I am doing.
> > Actually, I don't understand how #4 doesn't work with #3 where #3 is
> > taken as meaning that it must be possible to distribute some derived
> > versions of the copyrighted work.
>
> In most of the cases I can think of, the logic starts becoming complicated
> enough that I think "that's stupid; that's why these are guidelines". ("If
> DFSG#3 only requires some modifications to be distributable under the same
> license, and DFSG#4 requires all modifications except patches to be
> distributable but not necessarily under the same license--wait, that's
> stupid.")
>
> So, backing up a bit: can you suggest a license clause which would be
> allowed under the "some modifications" interpretation of DFSG#3 and
> disallowed under "all modifications", while still passing DFSG#4, and
> not falling under "rules lawyering" like the above? That is, an example
> showing that the different interpretation is meaningful. (I think the
> lack of one is why this is somewhat difficult to discuss.)
Sure -- imagine that there exists a clause in the terms of some copyright
which prohibited the distribution of some class of modification to some
part of the work which very clearly wasn't the source code to a program.
Here, there can be no security problems because of this lack of permission
(because the restrictions don't apply to the program), and permission
has been granted to distribute the work (clause 1 of the DFSG). And,
in general, all the other DFSG freedoms must still hold.
--
Raul
