Re: Interpreting the GFDL GR
Scripsit Glenn Maynard <glenn@zewt.org>
> On Wed, Mar 15, 2006 at 06:28:37PM +0100, Henning Makholm wrote:
>> I assert that this interpretation is most faithful to the arguments
>> presented by proponents of Amendment A during the discussion. In
>> particular, when confronted by arguments that a literal reading of the
>> GFDL leads to nonfreedom, proponents of the winning text did not
>> generally challenge the nonfreedom of the literal reading, but instead
>> argued that we should not use the literal reading because that was
>> "obviously" not what the FSF meant when drafting the license.
> This isn't a case of taking an *overly literal* reading of a license and
> ending up with something incorrect. Rather, the literal reading and the
> natural reading are in agreement, and there's no twisting involved.
I must have failed at expressing myself clearly - I do not attempt to
distinguish between "literal" and "natural" readings here.
> Read in English, naturally by a native speaker, the license clearly
> applies restrictions against "chmod", etc, and the above
> interpretation does not come from the license.
I agree on both counts. Yet rather than taking the GR to mean that
restrictions against chmod are OK in general, I think the GR says that
the GFDL should not be taken to imply restrictions against chmod. If
that leads to using an interpretaion that does not come from the
license, then so be it - it's a lesser evil than deciding that free
software does not need to be chmodable.
> We happen to have a clarification from one copyright holder (the FSF),
> but I don't understand how that can be extended to everyone else. The
> FSF does not have the power to offer binding statements of intent on
> behalf of all of the users of its licenses.
It is always dangerous to make an univocal statement about a license
rather than a concrete work.
However, I think that one would be reasonably safe in assuming that
somebody who thinks the GFDL is a reasonable license to use for their
own work must be someone who, for whatever reason, imagines that the
license does not restrict chmodding.
This is not an avenue that I would advocate if the DDBGR had not
forced us to, but I currently think it is the least harmful position
that is compatible with respect the outcome of the GR. Feel free to
show me that I'm wrong; I'd love to be convinced otherwise.
> (I still wonder: the FSF has an upgrade mechanism for its licenses, has
> known about this problem for years, and has acknowledged it as a problem.
> Where the hell is the fixed license? The only reason Debian is expending
> so much time on this is because of the FSF's stonewalling.
I suspect that the reason the FSF does not fix the minor problems is
that they are adverse to admitting that any criticism from Debian's
direction can have merit, given the underlying differences in
philosophy as regards the invariant sections.
--
Henning Makholm "Det må være spændende at bo på
en kugle. Har I nogen sinde besøgt de
egne, hvor folk går rundt med hovedet nedad?"
Reply to: