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

Re: Proposed General Resolution : IRC as a Debian communication channel



On Tue, Nov 06, 2001 at 01:24:45PM +0100, Raphael Hertzog wrote:
> Le Tue, Nov 06, 2001 at 06:07:49PM +1000, Anthony Towns écrivait:
> > The burden of proof is always on the person making a positive
> > claim. "Prove that he's never read this book", versus "Prove that he
> > has read this book", eg.
> You're making the positive claim : « #debian-devel is out of Debian's
> scope »

Which indicates nothing more than that you didn't understand what I said.

Please, stop, and try to understand it.

A positive claim is something that can be easily verified by a simple
argument.  "He read that book", can be verified by a photograph or a
video recording of whoever it is sitting in a chair, intently gazing
down at the book, or having him explain the book's plot in his own words
and recite sections of it from memory. "He has not read that book",
on the other hand can only be demonstrated by exhaustively cataloguing
every thing he's ever done in his life, and demonstrating that not one
of them happens to be reading the book.

You'll note, in particular, that the two tasks require very difficult
workloads. In particular, you'll note that if the question is "did he read
that book", and you're trying to decide whether the guy who says "yes,
he did" is right, or the guy that says "no, he didn't" is, then you'd
probably want to listen to the guy making the positive claim first since
that'll take less time, and if his evidence stands up under scrutiny,
there's no point worrying about the other guy at all.

Oh, and for reference, the only claim I've made in this thread is that
you should actually argue your point, which is a positive one, and,
you'll note, which I'm justifying as I make it.

> If you could concentrate on the real issue, that would be better.

It'd be lovely, wouldn't it. Unfortunately, without a postive argument,
there's just nothing to be said. I could start right now making an
authoritative examination of everything that should be official and
Debian sectioned, but by the time I was finished (at which point I could
demonstrate that #dd wasn't in that list) it'd be a few years later and
Debian would've died because woody still wouldn't be released.

> Of course I have :
> - unofficial means "not referenced in any debian documentation"

If that's exactly what it means, then in the interests of clarity,
you'd've been better off saying "documented" instead of "official".

> - official means "documented" and therefore #debian-devel would follow the
>   policy that Debian has decided for it

Again, that's a tautology. The "policy" that Debian's "decided" for it is
"whatever the channel opeators say, goes".

>   (and I propose the "open" policy
>   that let operators kick only on signal/noise consideration)

Which is all very well, but you need to *argue* that point, damn it,
not blather on about "officialness", and mention it in brackets as though
it were some sort of self-evident corollary.

> > in person. The channel could just as easily have been called "#foobar"
> Unfortunately it's called #debian-devel and people think that it's for
> person interested in debian development

Excepting the fact that it's not documented publically anywhere, and
is a hidden (+s) channel, so that people don't know about it in the
first place.

> > and still have the same function. Which is to say, the channel's utility
> > and effectiveness have nothing to do with its officiality.
> This can be argued : by officializing it more debian developers will know
> about it and may join it... 

Which is already covered by James emailing new developers as they join.

> by letting future maintainers join (and
> lurk), we'll let them learn faster how Debian works. 

Congratulations! Some actual justification for what you want to happen!
And now, you'll even get some on point rebuttal, instead of meta-rebuttal.
And you'll note it's on the topic of an actual *action* rather than some
random wording change, or way of thinking about things.

This isn't particularly true, though: for a start, we already provide
debian-policy, the developers-reference, several example packages and
helper tools, and the new-maintainer process; they're the "official"
guides to learning how Debian works, and they're available via the
traditional Debian avenues: the package system and email.

But there're a more important purpose to #d-d than teaching newbies:
and that's providing a quick and easy means for developers to talk to
each other. At the moment we come close to guaranteeing that everyone
on the channel's committed to Debian, usually evidenced by them being a
sponsored maintainer/in the n-m queue, or being a registered developer.
For a long while, we kept it that way by ensuring that the channel's
existance wasn't widely known, that way the only people who'd ever join
would be people recommended by someone else cluey. The benefits of having
a channel inhabited solely by people who're already actively involved in
Debian seems pretty clear: don't have to worry about "Debian SuX" trolls,
don't have to worry about answering user questions, don't have to worry
that someone's watching your every word about to jump all over it.

> It's not only about
> the channel utility, it's also about the channel "role" (within Debian).

And as far as that's concerned there's another problem with making irc
at all official, which has been alluded to elsewhere in this thread. IRC
is a *very* poor "official" communications medium. Try summarising an
IRC discussion sometime, or culling out the useful parts of an IRC log
in a way that stands alone and is actually readable, sometime. It's
very difficult to extract the sense from the noise even in a fairly
focussed conversation. That pretty much wrecks it for archiving purposes.
It's also realtime, which means that if two people don't happen to be
available at the same time (because ones asleep when the other's awake, or
because one can only do Debian stuff on the weekends and the other hates
touching computers in his free time, eg) it's also useless. IRC's great
as an unofficial communications medium, but IMO it's too inaccessible to
too many people to warrant being considered "official".

> > So could you. Seriously. Start again, from first principles, and try to
> > explain what you're on about.
> I believe I have been quite consistent with my principles. 

I'm sure you have. Unfortunately you haven't told us what those principles
actually *are*, or why you think other people should hold them, or why
what you want to happen follows from them.

The two sentences I replied to above are at least useful, but they're not
really very effective. Again, please actually *make* your argument. Start
out from a few fundamental principles ("Debian should be open!") and
demonstrate how whatever it is you want to actually *happen* is (or
isn't) in line with those principles. For reference, another you might
want to consider is ("Developers should be able to get things done with
a minimum of fuss").

Seriously: from where I sit you haven't even attempted to make your case,
which means you're not only wasting everyone else's time, but you're
wasting your own too.

Cheers,
aj

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

 "Security here. Yes, maam. Yes. Groucho glasses. Yes, we're on it.
   C'mon, guys. Somebody gave an aardvark a nose-cut: somebody who
    can't deal with deconstructionist humor. Code Blue."
		-- Mike Hoye,
		      see http://azure.humbug.org.au/~aj/armadillos.txt

Attachment: pgptYDulPEK7j.pgp
Description: PGP signature


Reply to: