Re: Interpreting the GFDL GR

[Glenn Maynard <glenn@zewt.org>]

On Wed, Mar 15, 2006 at 06:28:37PM +0100, Henning Makholm wrote:
>     we shall, for the purpose of the DFSG, assume that this
>     requirement is implicitly qualified by
> 
>          If distribution of Opaque copies is made by offering access
>          to copy from a designated place, then offering equivalent
>          access to copy the Transparent copy from the same place
>          counts as distribution of the Transparent copy, even though
>          third parties are not compelled to copy the Transparent copy
>          with the Opaque one.
> 
>     (I am not sure whether ftpmasters should apply this interpretation
>     when deciding whether they dare put binary packages containing
>     opaque copies of GFDL-licensed works on our ftp servers given the
>     current pool infrastructure. I would suggest not. However, that is
>     not directly a DFSG issue).
> 
> 
> 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.  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.

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.

(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.  So much of
this would be simpler if the FSF would fulfill their responsibility of
fixing their license, which they assumed when they begin proliferating
it.)

-- 
Glenn Maynard