Re: A possible GFDL compromise: a proposal
MJ Ray <email@example.com> a tapoté :
> > [...] And the affected ones will be users at first (lacking good
> > documentation because of an invariant section that is maybe not
> > something they consider as non-free),
> This is not a new effect: users who disagree with our definition of
> free software already don't have some things they may consider as free
> in Debian.
Not *in* Debian, but *shipped by* Debian. For you, there's no
distinction between GNU Emacs manual and Macromedia Flash?
> > the debian developers (forced to write from scratch new
> > documentation),
> They are not forced to do that, but may choose to.
> > the GNU developers (losting an user base for their documentation).
> This is entirely down to the GNU developers and the remedy is in their
> own hands.
We cannot make any progress if each part claims it's entirely up to
the other part, while both parts have sensible arguments.
> [...] > I agree with your first phrase. Unfortunately I cannot
> agree with the > last one: Debian already recommends non-free
> software to users by > provide apt-get links to these softwares,
> If linking were recommendation, then FSF recommends SCO, the FIPR and
> kelkoo because it links to them on
> http://www.fsf.org/philosophy/audio/audio.html and
> http://patron.fsf.org/2003-patrons.html respectively. This is clearly
This is not at all absurd. When you tell people they can find non-free
software at -this address-, you advertise for this non-free software.
When you do a Ph.D about NSDAP, when you quote Mein Kampf, you're not
advertising for this book (please, do not try to get godwin points by
yelling "godwin bla bla, I have no argument against your logical
To synthetise, you argue that a link is never a recommendation because
somes links are not recommendations. This is truly absurd.
> Debian has non-free software packaged for it on its mirrors, but it
> is not recommended and moves are underway about that, so please stop
> pushing at the open door.
Maybe the word "recommends" is a bit too strong, but is it the
problem? In my eyes, the problem right now is not whether non-free is
appropriate in Debian or not.
The problem to me is simple: we provide non-free stuff (yes, we do
provide it, despite the fact it's not considered as part of debian)
and we're about to claim that GFLed documentation, which may not at
all having any invariant part, is non-free stuff.
It seems weird to me. Someone said that GFLed documentation without
invariant sections can be made non-free if someone getting a copy of
the documentation add invariant sections.
What does it change? Do we consider BSD software as non-free software
because they do not forbid a software to become (wholy) "invariant"?
> As I understand it, invariants are not the only problem
So what are the others problems (on purpose, I ignore problems that
already got a decent answer)?
Not a native english speaker: