Anthony Towns <aj@azure.humbug.org.au> writes:
> After that meeting [0], I'd assumed it was in Christoph and Marc's capable
> hands,
... without ever *asking* if that would be true. I assumed this idea to
be dead because last year's discussion on -newmaint showed that most DDs
were against that proposal.
> Some of them have had some progress including a bunch of accounts created
> during the DPL election, but, some didn't, such as Peter Samuelson's DAM
> processing, which I'd explicitly referred to as an example that should
> have been trivial for the DAMs to process, as any problems should have
> been resolved at AM stage since it was AMed by a FrontDesk member. Peter's
> still in the DAM queue at the time of writing.
As the background for that is private, it would have been nice to keep
it out of this discussion. Anyway, the reason why this problem ended up
in the DAMs hands is because *only* the DAMs were mailed with some
doubts, not the AM. And that only after the AM report was done...
Anyway, I'm really not going to tear apart the little bubble in which
you have placed yourself - I simply don't see how this would help
anyone.
Now to the actual matter at hand. I had this discussion in IRC yesterday
and it seems that I should explain my opinion about the GR a second
time: [1]
(i) You have added a policy for everything, but removal from the DM list
is still under-defined. This is a crappy idea. Imagine a Sven Luther
case in DM - someone who's technically capable and invests a lot of
time, but leads to regular flamewars. There is no question that
we would need to have some procedure to decide what should happen
in such cases. Now, back to the Sven Luther example: Imagine how
*that* flamewar would look if the procedure to kick him out would
have been hand-crafted just for his case?
So basically, I won't accept your proposal as remotely sane until
the initial policy includes some guidelines on removals from the DM
list.
(ii) Debian has a QA problem. Sponsorship did nothing to improve it. In
fact, I believe sponsorship to be one of the reasons for it. Let's
explain this a bit:
Sponsorship is one of the main factors that lead to the explosion
of the number of packages in the archive. The growth in the number
of packages is, in fact, much bigger than the growth in the number
of users. This means that the number of users per package has
fallen [2], directly translating to the fact that packages receive
less testing before they are released. This is equivalent to bugs
and packaging mistakes staying unnoticed for a longer time.
This problem is almost exclusive to packages that are priority:
optional|extra, ie packages likely to be maintained by newcomers to
Debian.
On the other hand, sponsorship is (besides the few cases were
people only want to maintain a few packages and not invest more
time in Debian) used as an education system. It's training people
on the job, allowing them to understand policies and procedures
when bugs are reported or their sponsor notices a problem while
uploading.
The shrinking user:pkg ratio has already shown it's effect:
Packages of seldomly used, specialized software are often of low
quality, ignore licensing problems and aren't integrated into the
rest of the Debian system as tight as they could be. The
ever-growing number of RC bugs in sid is a sign for that, a better
sign is the number of unfixed important bugs [3] and there is
always the simple test of taking 20 random packages from the
archive and checking them for obvious packaging problems [4].
I'm also believing that sponsoring is not as good as it should be -
people often sponsor software without doing the thorough checks
that *should* be done.
Now on to the actual matter: The proposed Debian Maintainers
concept is splitted into two parts:
(1) Someone needs to advocate a maintainer, some people need to
decide if that maintainer should get added to the keyring and
thus get upload permissions for all packages that carry him
as Maintainer/Uploaders and have the DM-Upload-Allowed field
set. Without spelling out how the approval by the DM
keyring maintainers should happen, I guess most people are
thinking about checking packages, looking at past work, bug
handling, ...
(2) As soon as someone is in the DM keyring, a DD can give him
upload rights for virtually every package by adding the DM to
the Uploaders field and adding the DM-Upload-Allowed field.
This concept is completely ignoring the problems that sponsoring
exposed - in fact, it makes these problems worse. The number of
checks done by DDs is reduced to one examination of an initial set
of packages by the DM keyring maintainers [5]. The set of packages
that can be uploaded by the DM is something completely different -
those are "only" checked by a random Debian Developer. Together
with the low user:pkg ratio, my distrust in the abilities of the
average and our experiences with sponsored packages, I believe that
your DM proposal will be a future source of QA problem if it is
implemented. [6]
Before ending this mail, I would like to point out how frustrated I am
that my concerns have been ignored until now (and probably will be
ignored in the future). I have (once!) explained my position here on
-vote [7]and got only a few answers, none of them actually refuting my
arguments. The current policy on Debian lists seems to be that you are
ignored unless you repeat your opinion as reply to every single mail in
a thread. I'm too busy and simply don't care enough about Debian at the
moment to present my opinion in that manner.
Thanks for listening,
Marc
Footnotes:
[1] BTW, I would like to thank the #debian.de people (including buxy,
who seems to try to get into the s3kr1t german cabal) to help me
understand which part of my argumentation isn't obvious.
[2] Yeah, I know that I ignore the fact that the number of packages per
user isn't stable but has grown itself, but I believe it to be
unimportant for this case.
[3] RC bugs get listed and people interested only in the release and
not in the package itself fix those bugs, while issues of lower
priority stay unfixed
[4] My last results for that were that 3/4 of the examined packages
were - in my eyes - too bad to be included in the archive.
[5] The current list for the keyring maints at least ensures that no
sloppy sponsoring is getting through...
[6] In fact, my original understanding of the whole idea was that a
small set of DDs (like the proposed DM keyring maintainers) would
check every package before a DM would be allowed to upload it on
its own. I thought that to be something very, very positive, as it
would ensure at least one thorough and proper check, instead of the
current tradition of minimal checking done by sponsors.
[7] <87bqf85pjd.fsf@pindar.marcbrockschmidt.de> and a more verbose
statement about my wish for clear removal procedures in
<877ipw5i1m.fsf@pindar.marcbrockschmidt.de>
Attachment:
pgpojCMA3UviH.pgp
Description: PGP signature