Re: A transition plan to fsf-linux.org
On Wed, Jan 28, 2004 at 01:20:20PM -0500, Branden Robinson wrote:
> On Mon, Jan 26, 2004 at 10:10:43PM +0100, Sven Luther wrote:
> > On Mon, Jan 26, 2004 at 12:08:43PM -0500, Branden Robinson wrote:
> > > On Fri, Jan 23, 2004 at 06:11:38PM +0100, Sven Luther wrote:
> > > > On Thu, Jan 22, 2004 at 12:39:56PM -0500, Branden Robinson wrote:
> > > > > and all GNU documentation shall use the GNU FDL henceforth." Equally,
> > > > > it doesn't serve us to say "You'll take our non-free section away when
> > > > > you pry our cold, dead hands from it."
> > > >
> > > > Nope, only when a free alternative for all of its content has been
> > > > written.
> > >
> > > All of its content right now? Is there some freeze on adding new
> > > packages to non-free I hadn't heard about? If so, that's odd, because
> > > an ITP of a non-free package has been made on debian-devel within the
> > > past few days.
> > Sure, but hopefully there will be no more non-free software in some
> > point of the future :))
> Is that more or less likely to happen in the absence of a disincentive
> to produce it?
I think that keeping or removing non-free will have null effect to the
quantity of free vs non-free software that is available, nor on the
ressource that are going to pe poured into developing free software.
The mail you cite, which seems totally unreadable to me right now due to
the use of some strange character by the mail archive, is probably a
corrobation of this state of things. Not sure though, since i got only
garbage in galeon.
And anyway, my statement was that there would, in some time in the
future, be no incentive to work on non-free stuff, because everyone had
access to free alternatives of higher quality.
> *Would* dropping non-free serve as a disincentive to produce non-free
> software? Before making up one's mind, it might be worth considering
> the opinions of Ken Lunde of Adobe Systems Inc., when a (very polite)
> relicensing request was made of that company.
> It would appear that at least in some cases, the retention of the
> non-free section serves as an incentive to preserve the non-free
> licensing of a work.
Mmm, maybe i misunderstood then. But this is a problem of discussion
with the upstream author, and i quite understand the tentation of
wielding the 'remove non-free' stick to menace those people in the hope
of them relicencing the stuff. I have my doubts of this happening, they
will only find other excuses (oh, this is on non-free.org, should be
enough, or we host it on our web servers, just write an install
package). But anyway, by considering these cases only, and i think you
can count them on the fingers of one hand, you don't take into account
all the other packages where this is not nearly the case, those were it
is not in upstream's hand, those whose licence got lost (i recently
removed the old ocaml bignums, because the licence got lost in the
DEC->Compaq->HP mess, and there is nobody at HP with time enough for
searching for the paper. Upstream did a clean and free reimplementation,
so it is not a problem, but still). And seriously do you think nvidia
will revisite their licencing schemes because you remove non-free ? And
imagine the support nightmare you will have when the nvidia packages are
no more ?
> It is scenarios like this that make me question the prediction that all
> non-free software will just wither away if only we leave it alone.
No, who said leaving them alone ? My proposal, which was clumsily done,
and somwhat subsumed by Raul, was to keep non-free, but explicitly as a
staging area only until a free alternative does exist, or the licence is
freed. Putting the onus on the upstream author for the licence issues,
by publicly archiving the discussion about freeing the software, and
encouraging people to look for free alternatives. As soon as such a free
alternative exist, or the licence changes, we scrap the package from
non-free, and to bad for the upstream author who didn't want to free
This takes energy and people willing to follow it, but i believe it is a
much more strong incentive than removing non-free. This makes it clear
that they are second zone citizen in the FOSS (or whatever) community,
and that we encourage active work to make them obsolet and favorize
their competence (i would say concurence in french, but you don't say
that in english, i think).
By removing non-free, and relegating the distribution to third party
archives, you are ignoring them, and much more likely to have the effect
of 'leaving them alone' that you cite there.
> Where "leaving it alone" means continuing to endorse it by providing it
> on our mirror network, that is.
Yeah, sure, will you don't consider distributing the stuff on a third
party archive as leaving them alone, since you will give the users of
actual non-free software no choice than turning to these third party
Let's not be hypocrit, and continue distributing non-free, but put a
much bigger pressure to either free the code or replace it by free
alternatives, and you will hurt the upstream much more than by removing
non-free, after all, you will encourage their competence, and make them
loose market share.