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

Re: Debian needs more buildds. It has offers. They aren't being accepted.

On Mon, Feb 16, 2004 at 11:41:28AM -0700, Jamin W. Collins wrote:
> On Mon, Feb 16, 2004 at 02:20:41AM +1000, Anthony Towns wrote:
> > On Sun, Feb 15, 2004 at 02:00:36AM -0700, Jamin W. Collins wrote:
> > > On Sun, Feb 15, 2004 at 03:13:53PM +1000, Anthony Towns wrote:
> > > > I've asked in the past, I've been told that they don't have
> > > > time. 
> > > So, they did respond.  Thus, two way communcation.
> > What communication, exactly, are you claiming was missing? 
> In this case, status updates, plain and simple.  No NM application
> should go for long (no more than three at most IMO) without an entry as
> to what is happening.  

Even if nothing's happening? Whatever for?

You also seem to be confusing the requirement that the applicant be
told what's going on and the website be updated so everyone knows what's
going on. The former's worth getting stressed about; the latter is not.
In neither case is it reasonable to be adding to the chorus of invective
that makes up this list.

> > Your email included the comment, eg, "First thing you should notice
> > from this list is that there's an applicant that's been waiting on DAM
> > approval with no comment for over 3 years (almost 3 1/2 years).  Yes,
> > 3+ years!  That's absolutely ludicrous!"
> Yes, no comment for 3 years (based on the information on the page) is
> just that, ludicrous.  

No, it's more than that: it's a wrong deduction. You made a mistake,
you then insulted people based on that, and you don't think you did
anything wrong?

> > Certainly there wasn't a comment on the webpage summary; but as tbm
> > has since described there was a fair degree of communication, and the
> > long pauses were mostly due to that particular applicant not
> > responding to requests from the frontdesk.
> In this particular case there was some communication, in others there
> was not.  If the DPL didn't feel there was a problem, I don't think he
> would stepped in to correct the non-problem.  The fact remains that the
> DPL *did* step in to improve communication with NM applicants.  

From what others have said, no tbm didn't step in as DPL, he stepped
in as n-m frontdesk. Which is, you know, the position responsible for
making sure n-m's and the rest of Debian are kept informed of what's
going on and ensuring that things are running as smoothly as possible.

> As a result the NM process seems to be better for it.

The issue is whether or not getting these improvements requires telling
people their actions are "ludicrous", or putting them on constant trial.

> > > > Can you see the difference between that and flaming people on
> > > > -devel?
> > > Sure, there was communication from both sides, not silence.
> > And this is different how, exactly?
> Simple communication.  If you don't get it by this point, I'm sorry.

For instance, the last time I can recall talking to the security team,
I did so on the -private list. There certainly isn't anywhere like the
nm status page you can look to find an update, and non-developers don't
really have any hope. If you want to count that as two-way communication,
you've got the same thing in your example: people get mailed, and mail
back for a while, then nothing happens for a while and no more mail
gets exchanged.

> > Your mail also says:
> > ] * Date: Sun, 13 Jul 2003 13:09:47 -0600
> > ] I spent the morning working up a short script to parse
> > ] http://nm.debian.org/nmlist.php and gather a list of all individuals
> > ] that are listed as waiting for DAM approval.
> > Did you at any point attempt to communicate this to the frontdesk
> > without including comments like "That's absolutely ludicrous!", "This
> > is absolutely crazy!" or just asking for more information instead of
> > demanding that "something" has to be done?
> No, I spoke with my sponsor concerning the delay.  I was not aware that
> contacting the front desk would be of any potential help.  

Then, you know, perhaps you should've stopped at the above, and not added
the gratuitous insults. Saying instead things like "Is there anything we
can do about improving this?" instead of "This situation is ridiculous"?

Given that your information was obviously lacking, that you had very
little experience of how Debian works, and at best only very one-sided
experience with how n-m works, why were you so confident of your
assumptions that you felt it appropriate to publically ridicule the job
the DAM was doing? Given that a bunch of your assumptions were in error,
don't you think you should re-evaluate your attitude and work out some
other way of dealing with things in future?

> > > > Did you at any point ask Martin, either as front desk or as DPL,
> > > > to look into this privately and in a friendly, non-accusatory
> > > > manner? 
> > > No.  Debian claims to be an open organization, why should this have
> > > been looked into privately?  
> > Because posting to lists causes a whole bunch of people who don't have
> > any idea what's going on to chime in with their two cents and demand a
> > long justification that doesn't benefit anyone at all.
> That's your opinion.  

...and should thus be treated with utter ignore, I take it.

Seriously, that's the primary cause of any reluctance on my part
towards discussing things on this list. If you're really interested in
communication, I think it'd be fairly valuable to work on addressing it,
rather than dismissing it out of hand.

> > If that weren't the case, I'd agree absolutely. Unfortunately it is
> > the case, and everytime someone makes the assumption that flaming
> > first is a good way to go about promoting beneficial change, it
> > becomes even more the case.
> As others have already stated, perhaps the frequency of the complaints
> indicates that there are indeed problems.

And as I've already responded to said others, perhaps some of the
responsibility of fixing that situation falls onto those doing the

> > > > Did you at any point offer any help (and follow through on that
> > > > offer)?
> > > Offer to help with DAM approval? No.  Didn't figure anyone would
> > > accept an offer of help with DAM approval from someone awaiting it.  
> > You've passed since, presumably. Have you done anything since then to
> > make James' job easier?
> No.  It was made very clear by others during the -devel discussion that my
> assistance would not be accepted.

That's awfully convenient for you; it's alright for you to complain
about people not doing enough for you, and you're not even _allowed_
to actually do anything to help others.

In any case, afaics the only person posting to that thread that was rejecting
offers of help was you:


The obvious way (afaics) to fix problems you see with the n-m process is
to first become an AM, so you have a properly informed perspective, and
spend enough time doing that, and doing it well, that you can legitimately
claimed to be experienced, then ask other AMs, and particularly the front
desk and DAM what problems they see, and work out how to make them less
of a problem. After demonstrating your competency and effectiveness like
that for a while, then either offer to be an assistant DAM (if you think
that's worth doing and you still think you're able), or start doing some
other things that you think'll be effective -- having done helpful stuff
for the n-m crew in the past, even if the idea seems stupid at first
glance, you'll have earned the benefit of the doubt that most people'll
be willing to at least try it out.

> > > It wasn't a first resort and by no means something jumped to.  And,
> > > it could have been avoided with a simple update on the NM status
> > > page.  The fact remains that communication was, and still appears to
> > > be, lacking where James is involved.
> > Personally, one of the things I used to quite like about Debian is
> > that you could join the project and be a gruff curmudgeon with no
> > interpersonal skills at all, and still be respected just because you
> > did good work. 
> This would depend on the position held.  There are some positions within
> Debian that require a certain amount of communication.  This is simple a
> fact of larger projects, communication is needed.

Well, it's nice making statements so general that they can never be wrong,
but it tends to result in ones that don't actually mean anything. I
suppose it's valid to say that a certain amount of communication is
required from the DWN editor, after all, if they don't post DWN (and
isn't that communicating?) they're not doing much of a job, are they?

The responsibility of the DAM (and hey, imagine that -- we're talking
about buildds for a while, and then suddenly we're back to requiring
the DAM's actions be justified lest he be deemed unsuitable for the
job, sweet) isn't "make everyone happy" or "accept people as quickly as
possible", it's "ensure people accepted into Debian as developers meet
our requirements". It'd be nice to have the former things happen as well,
but not at a cost to the latter.

> > Aiming for technical excellence, and having people just chip in and
> > solve problems when they see them rather than worrying about whose job
> > it is to do what made for, IMO, a fun and effective way of writing and
> > managing software. 
> Without communication, this can also lead to duplication of effort.

There's nothing wrong with duplication of effort. It can serve as
redundancy in case one of the doers stops halfway, it can serve as an
educational experience, and it can serve as a way of better servicing our
users' needs if the results happen to be different enough that they're
both valuable.

> Effort that could be spent doing other things.

The effort spent on communicating can often be spent doing other
things too.

> > Personally, I'd much rather work in that environment than one where
> > everyone feels they should be saying something, even when they don't
> > know what they're talking about, or one where people are constantly
> > being put on trial and made to justify their actions. The evidence
> > would suggest that that's not the way Debian works anymore, and I
> > think we're incredibly poorer for that change.
> Communication is a necessary part of any large project, and Debian is
> certainly a large project.

That's nice. It's a good thing communication's a part of Debian then,
isn't it? Or did you mean "Jamin W. Collins' preferred forms of
communication are a necessary part of any large project" ?

> > Personally, given the choice between two people, one of whom has more
> > technical skills and experience, and the other of whom is more
> > popular, a better manager, and a better communicator, I'd rather the
> > person with the technical skills to be the one hacking on my OS.
> Great, let him hack on the OS and let the individual with communication
> skills communicate with others.  At this point James is in at least one
> position that needs communication skills, skills that IMO he either
> doesn't seem to possess or at least use.

That's odd, I thought you said that after Martin's response to your
complaint last year your problems with the n-m stuff were resolved. And
this thread doesn't seem to be about a lack of communication, but rather
about communicating an answer the recipients don't like. But hey, whatever
it takes to take those damned Debian elites down a peg or two, right?


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

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

Attachment: signature.asc
Description: Digital signature

Reply to: