[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: General Resolution: Removing non-free



On Wed, Jun 07, 2000 at 10:58:53PM +1000, Anthony Towns wrote:
> > I support my analogy by asserting that Debian is not a general store.  We
> > don't sell everything.  In fact, we can't.
> 
> We're not? We don't?

Uh, I'm sure nobody will have any trouble thinking of some fairly prominent
pieces of software in the world that Debian does not -- and cannot legally
-- distribute.

> Sure, we don't stock some speciality items that we just can't get,

I "can get" quite a few pieces of software that I could technically package
for Debian (albeit crudely), and yet which would land me in court if
attempted to distribute normally as a package.

> Obviously, I'm not so convinced. Debian specifically says that some
> people have to eat unhealthy foods, and they should go right ahead and
> do it. They should try making their diet in general healthy, but if you
> really really have to have pork and grits every now and then, we won't
> chastise you for it. I don't think a health food store would do any of
> that, at least not up-front.

Nothing in the GR says that Debian will chastise its users for using
non-free software.  I don't know where you're coming up with this aspect of
your analogy.  In fact, I suggest you review the new title of section 5 in
the amended GR.

Most of the rest of your digression on the food analogy seems solely to be
sarcasm, so I'll skip it.

> On to whether this breaks part of our social contract.
> 
> First: if it does, well, that's not necessarily Wrong: especially if the
> free software community and the Debian user base think it's a good change.
> All it means it that you (as a proposer or a second) really ought to
> ask them first, ideally, IMO, in a more quantifiable way than "yeah,
> the people I talk to tend to dis/agree with it".

Okay.  Who wants to set up a mechanism for collecting opinions on the
subject?

> > Interestingly, [for point 5], all word of future commitments vanishes, and
> > we are left with descriptions of the status quo.  Whether these constitute
> > promises is left as an excercise for the reader [...]
> 
> Arguing that they're not written as promises when they're specifically
> part of our `social contract' seems pretty weasely to me.

A brave assertion.  But then you have to think about this carefully-crafted
document and wonder why it doesn't read like this:

  5. We Will Support Programs That Don't Meet Our Free-Software Standards

  We will maintain "contrib" and "non-free" areas in our FTP archive for
  this software.  We encourage CD manufacturers to read the licenses of
  software packages in these directories and determine if they can
  distribute that software on their CDs.  We will support the use of
  non-free software in Debian, and we will provide infrastructure (such as
  our bug-tracking system and mailing lists) for non-free software
  packages.

So?  Why doesn't it read like that?

> > 9) We will place [the interests of our users and the free-software
> > community] first in our priorities.
> 
> 13.5) We acknowledge that some of our users require the use of programs
>       that don't conform to the Debian Free Software Guidelines.
> 
> One of John's rationale points was, I think, intended to imply this
> isn't correct any longer: they don't need netscape, or acrobat, or
> anything anymore. Perhaps it isn't. I don't agree, and I suspect most
> people here don't either, by the mere fact that most of us probably
> still use non-free software.

The rationales are not submitted for seconds, or a vote, or for any kind of
ratification.  How about sticking to the proposed GR?

> > 16) we support [the use of non-free software in Debian]
> > 17) we provide infrastructure (such as our bug-tracking system and mailing
> > lists) for non-free software packages.
> 
> Exactly quoted, these are: ``Thus, although non-free software isn't a
> part of Debian, we support its use, and we provide infrastructure (such as
> our bug-tracking system and mailing lists) for non-free software packages.''
> 
> Let's rephrase them slightly, putting them in future tense:
> 
> ``Thus, although non-free software isn't a part of Debian, we will
>   support its use, and we will provide infrastructure (such as our
>   bug-tracking system and mailing lists) for non-free software packages.''
> 
> This isn't, of course, how it's written.

Has it ever occurred to you to ask why?  The Social Contract was argued and
argued and argued about, and yet somehow this final clause managed to fail
to get firmed up into unambiguous, forward-looking, solid committments to
future action, unlike every single one of the preceding sections.

Coincidences are amazing things.

> It is, however, in the language of a promise, as Branden said above.

As a matter of fact, I didn't say that.  I said:

> > Interestingly, after this point, all word of future commitments
> > vanishes, and we are left with descriptions of the status quo.  Whether
> > these constitute promises is left as an excercise for the reader -- but
> > I'd remind the reader that promises, vows, or contracts bind people to
> > action in the future (affidavits are documents where you swear to the
> > truth of events in the past or current status)...13 promises and 4
> > things that can be interpreted as promises only with an expansive
> > definition.  But, not one to shy away from a challenge, for the sake of
> > argument I will be liberal and accept just such an expansive
> > definition.

Very clever of you to take something I said "for the sake of argument" and
now present it as an unqualified assertion.

> As it's a part of our social contract
> --- our promise to the free software community --- this is how I read it.
> It certainly seems a pretty short and reasonable step from "We will
> be guided by the needs of our users..." and "some of our users require
> [non-free software]" to "we'll distribute some of this non-free software",

Is it any less of a reasonable step from "We will be guided by the needs of
our users..." and "We acknowledge that some of our users require the use of
programs that don't conform to the Debian Free Software Guidelines" to
"While we will not distribute such software itself,We will support users of
our system who develop or run non-free software."?

> and indeed, that's the step the social contract takes. If one of those
> assumptions changes (in particular, if our users decide they don't care
> about non-free software at all), then sure, it's appropriate to change
> the conclusion too.

So the social contract should not be immutable... well, that distinguishes
your position from some of the opponents of the proposed GR.

You cite the wishes of our users as one criterion for change.  Are there
any others?

> But even though we didn't ignore that line, did we reduce the utility of
> Debian? No, we drew the line, and *kept* Debian's usefulness.

How foolish of me to not notice that Debian's usefulness will be utterly
destroyed by passage of this GR.

> So what, then, is this precedent?

The very fact that we didn't write section 5 as I first hypothesized above.

Most people seem to recognize the existing section 5 as a balance between
competing interests.  Will the time ever come to review that balance?  Has
nothing changed in the past 4-5 years?

> > > and thus would marginalize us into a non entity.
> > This does not follow.  Giving up some of a substance, like "market share"
> > is not the same thing as giving up all of it.  Furthermore, I do not think
> > you can even prove this assertion except retrospectively.  We do not
> > collect statistics on our users, so it is difficult to make quantitative
> > stataments about the demography of Debian users in general, let alone
> > perform longitudnal analysis.
> 
> But we do collect statistics on our users: 
> 	http://www.debian.org/~apenwarr/popcon/

Are these statistics strong enough to be used for making policy decisions
of any sort?  Remember what they say about lies, damn lies, and statistics.
If your methods are poor, your numbers are meaningless.  The popularity
contents results are fun to read but they have to be kept in perspective.
The population base is self-selected.  Furthermore, the figures you present
are at best a fraction of Debian *developers* let alone Debian users.  How
much faith are you willing to put in these numbers?  Deciding how to
organize a CD is one thing (cf. the package description); this can
reasonably be considered quite another.

> I am fundamentally opposed to dropping software --- even non-free
> software --- to satisfy a political whim, when that software is useful,
> has been packaged to work within Debian and is legally redistributable.

To characterize it as a "whim" is a belittling attack.  Replace "whim" with
"principle".  Would you then say the same thing?

> I don't like having some semi-official repository for KDE .debs, but I
> understand it because of the license situation. I don't like having a
> special repository for Gnome debs, that people get recommended because
> they're better than the Debian Gnome debs. I don't like have a couple
> of repositories for IPv6 debs.

You're welcome to your preferences.

> By being a single all-encompassing distribution our user's have lots of
> benefits: one place to go to for support and filing bugs, easy to find
> what software's available, an assurance that it'll all work together,
> and if it doesn't there won't be too much fingerpointing about whose
> fault it is. Separating all these little things into different little
> organisations that don't care about each other isn't good because it
> removes all the benefits that coordination brings.

Funny, I thought we were a bazaar, not a cathedral.  Furthermore, I think
you overestimate the coordination of our existing arrangement.

> And for what? To change our stance from "we recognise non-free software
> exists, that it's sometimes necessary, and we'll help" to "we don't want
> to hear about it, go somewhere else"?

Of course, how silly of me to overlook the words "we don't want to hear
about it, go somewhere else" in John's proposal.

> To save a a measely gigabyte or two of disk space when the distribution
> increases by twice that each year?

Ask Jason, he's the one who had to shut down the mirror on samosa because
the disk was out of room.  Moreover, I think space requirements on disk is
but one of many factors that needs to be taken into account when
considering this proposal.  Perhaps it's fashionable to take a conventional
political stance, and promote fear among people ("THEY'RE TRYING TO TAKE
YOUR SOFTWARE AWAY!"[1]), but I think we should take a more measured
approach.  That can't even really begin until there's an understanding
about what the GR is and is not, and what it will or will not necessarily,
or likely, do.  And we seem to be a ways even from that.

[1] Ah, I can manufacture quotes, too.  This is fun!

-- 
G. Branden Robinson            |    Murphy's Guide to Science:
Debian GNU/Linux               |    If it's green or squirms, it's biology.
branden@ecn.purdue.edu         |    If it stinks, it's chemistry.
roger.ecn.purdue.edu/~branden/ |    If it doesn't work, it's physics.

Attachment: pgpQEWiGtREPk.pgp
Description: PGP signature


Reply to: