On Tue, Jan 06, 2004 at 09:30:41PM -0600, John Goerzen wrote:
> While we cannot (and should not) prevent others from doing things we
> find distasteful, we are certainly capable of refraining from doing
> those things ourselves.

We certainly shouldn't do things we find distasteful, but there don't
seem to be so many people the feel that way that we can't continue
maintaining non-free ourselves.

I'm sure there are people who find Gnome distasteful, but they manage
to put up with other people working on it and using it. Why can't you
do the same? It's quite easy to find perl distasteful too, and not too
hard to come up with reasons why it's detrimental to its users; but we
can make those arguments while still continuing to support it.

> > In addition, I don't see how you can reasonably claim that "we" consider
> > non-free software to have been licensed unethically; I certainly don't,
> > and the fact that Debian distributes it at all tends to indicate that
> > Debian as a whole has not had any ethical problems. Personally, and I'm
> Point taken, though it is sad to hear that.

Well, Richard Stallman's also come out in the past few years to claim it's
ethical to license documentation non-freely. That's sad too I suppose,
but I don't think it's reasonable to call him unethical because of it.

> > > It should be apparent that the cause of Free Software is not advanced by
> > > promoting and supporting non-free.  
> > It's no such thing. There are at least three ways in which the converse
> > is the case:
> >   * non-free allows us to establish an upstream/downstream relationship
> >     with authors of non-free software, which we can use to encourage
> >     authors to license their software more freely
> This could still be possible for the maintainers at nonfree.org, or on
> project mailing lists, or any number of other ways.  Moreover, new
> incentives (or disincentives) could become available.

Perhaps. Personally, I don't see why people in nonfree.org would bother
arguing -- it'd just be more work for them if they managed to succeed,
since they'd have to repackage it for Debian.

But personally, I don't think a nonfree.org is actually likely; I think a
range of random apt sources for each individual program, some maintained
by upstream, others by random folks, perhaps others on people.d.o,
would be the result.

> >   * making non-free software easily available as an add-on to a fully free
> [ snip ]
> It seems that both of these would hold either way.

Well, they might hold if there was a well publicised, high quality archive
for non-free software that was well integrated into Debian. But then
that would act as just as much of a crutch as non-free currently does,
and we'd have no control over it, which, I thought, was the main thing
you were trying to avoid.

> > You're welcome to argue that, on balance, the Cause is better served by
> > dropping non-free, but arguing that it's in no way advanced by non-free
> > is patently untrue.
> "Better served by dropping" seems the same as "not advanced by keeping"
> to me; in any case, you may ascribe a macro view to the comment :-)

Well, no. Walking advances you, but running serves you better if you
want to get their more quickly. It's simply wrong to say that "you're
not advanced by walking". Likewise here.

> > > non-free part of our archive.  Imagine, then, how much greator those
> > > effects would be by completely banning that software from our project
> > > until it gets a Free license!
> > How about you provide some evidence for thinking it'd be greater,
> > rather than just asking us to imagine it? Personally, I think it'd be
> > marginally worse, since we have less opportunity to build a relationship
> > with upstream.
> I do not have any evidence for that, of course.  We do not know for
> certain, nor can we.  

We certainly could. You could have a couple of non-free software authors
who've said "Thanks for distributing this, without your support I wouldn't
have continued maintaining it", or you could compare how easy it is to
convince people over their better judgement to ease up their license
enough for non-free, compared to getting stuff from non-free into main
when they're not specifically trying to write DFSG-free software.

> People arguing against this claim to know for
> certain that having non-free is a larger stick than not having it;

Uh, no. I think having the DFSG, and the ability to back up all its
points on purely pragmatic grounds, is enough to convince most people
to license things freely, and that the rest have sufficient reasons not
to that they're not going to care about our "sticks".

Personally, I don't think comparing the size of the sticks with which we
beat people is a mark of the "friendship" that free software is meant to
promote, and proprietary software is meant to diminish.

> an
> assumption which I seriously doubt; but again, neither of us has done a
> study on the topic :-)

Well, I've been involved in a couple of discussions which have resulted
in software being relicensed in a DFSG-free manner. I don't think any
of them would've been improved by Debian not having non-free.

I'll happily take harder evidence of my own anecdotes, but anecdotes
trump unempirical, Platonic idealisation any day as far as I'm concerned.

> > > More importantly, we will finally show that it is possible to build a
> > > world-class operating system from only Free components.  
> > Why do you think that Debian main doesn't already demonstrate that?
> Many people seem to find Debian main unusable without Debian non-free,

This wouldn't change if Debian simply drops non-free.

> and to assume that non-free is also part of the distribution, or 

Personally, I think this is a pointless word game. I don't care if
people think Debian distributes non-free, or they think it's part
of "the distribution", or part of "the project", or whatever else,
especially if they understand the difference between what's in "main",
"contrib" and "non-free". I don't think I know anyone who uses Debian who
isn't aware of that difference.

> to fail to fix the deficiencies in main because non-free is "good enough".

Why do you think this wouldn't remain true, with the replacement that
"nonfree.org is "good enough""? 

While I can understand people being motivated by "no free software that
does this", I can't really imagine anyone being motivated by "there's
software that does this, that I can use now, but it's not in Debian"
who wouldn't already have been motivated by it not being free.

> > > These restrictions are not good for users.  
> > And personally, I'd rather let our users and developers decide what's
> > good for them on an individual basis. If they don't think non-free's
> > good on balance, it's simple to choose not to work on it.
> There is no removal of choice here; they are still free to put whatever
> they like in sources.list.

Sure there is. Right now they can upload non-free packages to Debian,
or create a site of their own. If the resolution passes, they can only
create a site of their own. Right now, they can download non-free packages
for crafty from Debian, or you can compile it yourself, or you can find
someone else who's pre-compiled it for you. If the resolution passes,
only two of those options are available.

> Of course, you would have us also decide that providing non-free is
> better for our users.  

Providing a choice is (almost) always better; that's the whole reason
why we like freedom in the first place. But that's not the converse:
you're claiming that not using non-free software is always better; the
converse would be trying to force non-free on them. Giving people the
*choice* of using non-free or not is much better.

And the best way of giving people that choice, is to do it by continuing
to reuse Debian's existing infrastructure: it's less risky, it's more
efficient on the free software community's resources, and it's highly
likely to result in better outcomes for our users.

> > > I submit that the best possible long-term outcome for our users is one
> > > in which non-free software ceases to play any part in their lives.
> > I can't imagine why you'd think anyone in Debian would disagree with that
> > statement, or would need it "submitted" to them.
> I have learned that I no longer can take that for granted.

Or you've fallen under a delusion that makes you think erroneous thoughts
of your fellow developers. It might be a good idea to put your reasoning
up, in case it turns out that it's wrong and can be refuted. Or perhaps
you can find someone who'll say "I think that non-free software should
play a part in our users' lives forevermore" ?

(I don't think you'll find anyone that'll say that, because given the
same software, a free license is always better as far as its users are
concerned. The only argument for keeping non-free software around forever
is that it's easier to develop non-free software than free software;
and while there're probably people in Debian who'll concede that's the
case now, I doubt you'll find many who don't think it's a problem that
can be fixed in the future.)

> > > We have to bite the bullet sometime, and now is a great time to do that.
> > We don't actually have to "bite the bullet" on this ever.
> That's fine, but it seems that there have not been serious proposals
> that would achieve it any other way.  If that can happen, that's good
> too.

Huh? We don't have to bite the bullet on this issue, and remove useful
non-free software from Debian. There's nothing that requires us to
do that; though that isn't to say we mightn't choose to anyway.

That there aren't any serious (non tyrannical) proposals to ensure that
there is no useful non-free software on the planet doesn't mean we should
"bite the bullet" on random things, just because we feel the need to
bite *something*.

> > > I have also said that I believe all operating systems suck.  I still do.
> > > Debian sucks also, though I like to say it sucks less.  I want Debian to
> > > be the first operating system that I can truly say "does not suck."
> > Then drop non-free from your sources.list, if you think that's all it takes.
> No, that is not all it takes.  I am saying that we can make it happen
> with some work.

If you don't think "main" alone is non-suckful, what, exactly, makes
you think dropping non-free will cause it to become non-suckful?

You seem to be assuming that it'll cause a bunch of people to setup
nonfree.org and maintain the packages there. I can't see that being as
efficient, let alone more efficient than what we do, so there's a loss
in resources for Debian there, that I don't think can come at no cost
to the free software community.

We get an extra "stick" to say "non-free? no Debian for you! goodbye!",
but I can't see any indication that that would cause any significant
amount more free software to be available.

We get to have a closer alliance with the FSF, but I can't see how that'd
actually cause us to get any more free software either.

We get to not have these discussions every few years, but I can't really
imagine anyone in Debian seriously thinking that the existance of large
dicussions actively discussions are causing that much harm to free software.

So I'm at a loss to see any reason why you think you'll be able to say Debian
doesn't suck after this resolution passes.

Cheers,
aj

-- 
Anthony Towns <[EMAIL PROTECTED]> <http://azure.humbug.org.au/~aj/>
I don't speak for anyone save myself. GPG signed mail preferred.

               Linux.conf.au 2004 -- Because we can.
           http://conf.linux.org.au/ -- Jan 12-17, 2004

Attachment: signature.asc
Description: Digital signature

Reply via email to