Debian Developers - NM queue and groups [Was: NEW queue and ftp-master approval]

Author

NM queue and groups [Was: NEW queue and ftp-master approval]

Joel Aelwyn

2005-01-25, 5:53 pm

On Tue, Jan 25, 2005 at 08:25:22AM +0100, Marc Haber wrote:
> On Mon, 24 Jan 2005 22:28:25 +0100, Goswin von Brederlow
> <brederlo@informatik.uni-tuebingen.de> wrote:
>
> And, as usual, no communication with the outside takes place. Which is
> a real mess.

<hijack mode="thread"/>

On the flip side, whatever was broken with the NM queue in the past (and
people have a wide variety of opinions on what that might have been), it
appears to have started moving relatively steadily, coincidental with the
addition of another DAM.

Perhaps that really is coincidence, or perhaps it is causal; maybe there
are things we can't see going on inside[1], but whatever changed, it
appears to have had a positive, and so far consistant, effect.

So, three cheers for... uhm... whatever fixed it.

[1] Which is a separate rant, and frankly, I think Debian needs to be
clear about what we really mean by "We won't hide probles" in our Social
Contract - it talks specifically about our bug database, but if that's
what we mean by "problems", we should say "We won't hide bugs in software
we maintain". The usage of "problems" implies a far broader expectation
of things like transparent process, especially when a process appears to
be hung or broken, and we don't seem to even be able to decide, amongst
ourselves, which we mean - perhaps because different camps want each to
be true. But I hesitate to even open the can of worms of "another Social
Contract revision" so soon. Still, people should think about what they
believe this means.
--
Joel Aelwyn <fenton@debian.org> ,''`.
: :' :
`. `'
`-

Andrew Suffield

2005-01-25, 8:48 pm

On Tue, Jan 25, 2005 at 10:52:48AM -0700, Joel Aelwyn wrote:
> [1] Which is a separate rant, and frankly, I think Debian needs to be
> clear about what we really mean by "We won't hide probles" in our Social
> Contract

It's a literal statement. We won't hide them. As always with the
social contract, do not attempt to assume the inverse is true. Just
because we won't hide them does not mean we're committed to going out
of our way to make them well-known and easy to understand. It is not a
commitment to some higher notion of transparency, but rather merely to
avoid *obstructing* transparency.

Complaining that you didn't know what the issues with the NM process
were is precisely equivalent to complaining that you didn't know about
some random bug which nobody had filed. Nobody was hiding anything,
it's just that nobody bothered to document the problem; they're very
different things.

--
.''`. ** Debian GNU/Linux ** | Andrew Suffield
: :' : http://www.debian.org/ |
`. `' |
`- -><- |

Joel Aelwyn

2005-01-25, 8:48 pm

On Wed, Jan 26, 2005 at 12:06:06AM +0000, Andrew Suffield wrote:
> On Tue, Jan 25, 2005 at 10:52:48AM -0700, Joel Aelwyn wrote:
>
> It's a literal statement. We won't hide them. As always with the
> social contract, do not attempt to assume the inverse is true. Just
> because we won't hide them does not mean we're committed to going out
> of our way to make them well-known and easy to understand. It is not a
> commitment to some higher notion of transparency, but rather merely to
> avoid *obstructing* transparency.
>
> Complaining that you didn't know what the issues with the NM process
> were is precisely equivalent to complaining that you didn't know about
> some random bug which nobody had filed. Nobody was hiding anything,
> it's just that nobody bothered to document the problem; they're very
> different things.

I notice that you conveniently trimmed the portion of my statement that
went into detail about what I consider the core issue to be: what is meant
by "problems".

One could argue that failing to acknowledge, or do anything about, an
utter lack of transparency in our basic processes is, in fact, hiding
problems, by tacit acceptance and omission rather than deliberate
obfuscation. Complaining about a bug nobody has noticed is useless;
ignoring the existance of a "bug" and refusing to make it possible to even
see that there IS a "bug" when there appears to be one is a very different
thing. Our software is "open" - it can be reviewed, understood, and
potentially updated (even if that isn't accepted upstream in all cases);
our non-electoral processes appear to act, at least in practice and for the
most part, a great deal more like closed models.

However, it's a pointless debate if the majority of developers consider the
usage of "problem" in that clause to be strictly equivalent to "bug in the
software we ship". Certainly that meaning has been asserted more than once
in my experiences on the Debian mailing lists. I, personally, don't agree
with it, but there are very plausible arguments for both views.

*If* it means "problems" in the more general sense, then it's meaningful to
discuss what we think the problems are, and what we think that commitment
to "not hiding" them means in terms of practical application. But until the
question of what our Social Contract actually commits us to considering
as problems in the first place is clear, it won't go anywhere because it
*can't* - I, at least, have already seen enough evidence that I believe
that trying to change anything of the sort would simply get bogged down in
"But that isn't even something we promise to do", long before dealing with
the questions of "Is there an issue, and if so, what is that issue?" and
"How can we resolve the issue?"
--
Joel Aelwyn <fenton@debian.org> ,''`.
: :' :
`. `'
`-

Andrew Suffield

2005-01-26, 7:51 am

On Tue, Jan 25, 2005 at 06:01:26PM -0700, Joel Aelwyn wrote:
> On Wed, Jan 26, 2005 at 12:06:06AM +0000, Andrew Suffield wrote:
>
> I notice that you conveniently trimmed the portion of my statement that
> went into detail about what I consider the core issue to be: what is meant
> by "problems".

It's irrelevant.

> One could argue that failing to acknowledge, or do anything about, an
> utter lack of transparency in our basic processes is, in fact, hiding
> problems, by tacit acceptance and omission rather than deliberate
> obfuscation.

One could, but it would be stupid pointless word games. You might as
well make similar complaints about tagging bugs as wontfix and closing
them. I already told you once that this is *not* what it means.

--
.''`. ** Debian GNU/Linux ** | Andrew Suffield
: :' : http://www.debian.org/ |
`. `' |
`- -><- |

Joel Aelwyn

2005-01-26, 6:03 pm

On Wed, Jan 26, 2005 at 10:30:01AM +0000, Andrew Suffield wrote:
> On Tue, Jan 25, 2005 at 06:01:26PM -0700, Joel Aelwyn wrote:
>
> It's irrelevant.

That's funny. I like that. I really do. You should do stand-up.

Or maybe a mind-reading show, since you seem to know, better than I do,
what the topic of my message was.

>
> One could, but it would be stupid pointless word games. You might as
> well make similar complaints about tagging bugs as wontfix and closing
> them. I already told you once that this is *not* what it means.

In fact, the parts you have chosen to keep, and respond to, are the far
*less* relevant portions of what I wrote. They existed as a demonstration
only of one reason I consider it important for people to have some
agreement on what the usage of "problems" means in our Social Contract - so
that it can be decided whether those are *relevant* topics at all. Nothing
more.

Let's try this one more time:

1) I have seen people assert that "We won't hide problems" means "Our bug
database will be open", *and nothing more* - that the following phrase is a
delineation of exactly what that means, rather than an example of a minimum
expectation.

2) I have also seen people assert that this same phrase means we won't
hide any problems, real or perceived, that do not have an immediate and
overwhelming reason to be *temporarily* non-public (such as security
announcements where we would simply be cut off from any future announcement
if we publicized it too early).

(There have also been views that it should demand no hiding of security
issues either, but I find that an impractical and fairly useless suggestion
given the reality of security fixes today.)

These statements are plain facts. *I have seen both things asserted*.

Some set of other questions, which I will not repeat here because you'll
probably just trim the rest of the message again and argue about them
pointlessly if I do, depend on answering which of these two assertions is,
in fact, a correct summation of the project's stance on the topic. They are
mutually exclusive; it is not possible for both to be correct statements
of a single entity's stance simultaneously (it is possible that neither
of them is correct, in which case someone should propose an alternative
interpretation).

Perhaps fundamental and significant disagreements over what an entire
clause of our Social Contract means aren't important to you; for my part,
*I* would like to know what people believe they are agreeing with when they
agree to abide by the SC.
--
Joel Aelwyn <fenton@debian.org> ,''`.
: :' :
`. `'
`-

Andrew Suffield

2005-01-27, 7:54 am

On Wed, Jan 26, 2005 at 12:08:27PM -0700, Joel Aelwyn wrote:
> In fact, the parts you have chosen to keep, and respond to, are the far
> *less* relevant portions of what I wrote. They existed as a demonstration
> only of one reason I consider it important for people to have some
> agreement on what the usage of "problems" means in our Social Contract

Debating the definition of this word remains irrelevant, no matter how
much nonsense you write about it. See previous message.

Trying to reduce ethical issues to word games is, at best,
childish. Grow up.

--
.''`. ** Debian GNU/Linux ** | Andrew Suffield
: :' : http://www.debian.org/ |
`. `' |
`- -><- |