Re: Reflections about the questions for the candidates
* Enrico Zini <email@example.com> [2006-03-05 14:20]:
> On Sun, Mar 05, 2006 at 02:27:36AM +0000, Martin Michlmayr wrote:
> > > Actually, in a recent chatting one of the past DPLs told me that he
> > > tried at some point, but the feedback he got was roughly "who cares?".
> > Since we talked recently, I'm wondering if you're referring to me. If
> > so, I didn't express myself clearly. Let me know if you're referring
> > to me and I'll try to elaborate what I meant.
> Yup, that was from my memories of our conversation at FOSDEM: I'm sorry
> if I have misuderstood or misrepresented you.
Sorry for the delay but I've been busy filing 500 or so bug reports. ;-)
When I talked to you at FOSDEM, I didn't want to suggest that nobody
cares about reports from the DPL. Quite the opposite! I think it's
very important to keep people up to date, and that doesn't just apply
to the DPL but to everyone. We could do much better actually writing
up new developments and announcing them. (See for example the comment
by bkoz on http://lwn.net/Articles/177353/#Comments to see what kind
of great feedback you can get.)
What I tried to express was that I encountered two problems while
sending updates during the time I acted as DPL:
- Most work done by the DPL is work to keep Debian going but it's
not necessarily any sexy work you can announce. When I got asked
at conferences what my job entails then I would often answer
that a big chunk of my time is making sure that the project actually
keeps running. You usually don't see this work - you'd only see it
if it didn't get done, because then we'd have all kinds of problems.
Many things the DPL does aren't sexy at all but they are necessary
tasks to keep the project running smoothly. This can involve work
like mediating between people, or obtaining some hardware, sorting
out hosting, dealing with legal stuff, etc, etc. There are many
small things, but they all add up.
So the obvious problem is that much of the work the DPL does is
very important, but it's not really debian-devel-announce material.
And that's a problem I had as DPL: I didn't want to make monthly
postings just for the sake of it, even if there wasn't anything
to *announce*. I also felt that people treated me as the Debian
News Summarizer rather than the DPL.
As DPL, you work together with delegates and basically make sure
that they can do their work. What this means is that they do the
cool stuff and that *they* should post to d-d-a. In the past I'd
summarize what delegates did, but then I increasingly encouraged
them to post their own summaries. Again, that's important work -
but people don't see that the DPL was actually involved. They only
see that a delegate has posted something, give them credit, and go
on complaining that the DPL doesn't do anything. (Oh, and I'm *not*
saying here that the DPL is behind every posting on d-d-a; but in
some cases, they have been involved behind the scenes.)
So, what I'm saying is not that reports are useless but that most
stuff is not important enough to warrant a d-d-a posting. If I
were DPL again, what I'd do is to send small summaries to
debian-project (where it *is* okay to send a two-line summary of
something, and do that for each item), collect these summaries on
a web page (with an RSS feed) and then put them into d-d-a postings
every 3 or 6 months.
- The second problem was that if you do post something (see the
report I sent after 6 months of being DPL) people only complain
that you're doing self-promotion... but that's just the typical
problem of Debian and free software, i.e. that no matter what you
do someone will complain. What I did, and looking back this was
a stupid thing, was to say "fuck them all, then I'll just not post
what I've been up to". The right thing of course is to ignore such
people and do post because the majority appreciates it. But don't
assume that just because many people want reports that you'll only
get positive responses...
In summary: updates are completely crucial, but not everything the DPL
does is d-d-a material. Instead, smaller reports should be posted to
-project and then summarizes for d-d-a every once in a while.
I think some of the things I've just said also came up in the DPL
debates this year, but I only skimmed them when it became clear that it
was mostly a venue to complain that the DPL doesn't do anything and to
ask candidates to justify themselves and explain what they'll do...
[Yes, a posting on the tasks the DPL does for my journal is
forthcoming. I 'just' need to find the time to dig through my DPL
archives to come up with some examples of typical tasks.]