Leadership in Debian (was Re: [all candidates] DPL salary)
On 2013-03-14 13:10, Enrico Zini wrote:
We've definitely come to expect too much from a DPL, and we need to
break that up. [...]
Thanks. Your message explains better what I've mentioned, that (even
ignoring the associated problems) I don't see it as healthy for us to
push for a DPL with more and more time, but rather to fix the job to be
more possible on available time.
I don't say that because of my own situation, but because I want us to
continue to be able to have a good set of DPL candidates to choose from.
The more people see an expectation of a "full time" role, the fewer
people will be willing to run -- in my view, this would be true for the
most appropriate candidates even if we paid the DPL a salary.
I believe that Debian is at its best when it is a flat organisation
where different groups and individual contributors work together
directly as needed. The DPL and others can help by following progress,
speaking to delegates, suggesting help where it is needed, and so on.
But in each case they should be aiming to nurture healthy teams that
function well without intervention, not to make themselves continually
indispensable in every area.
I think this type of leadership can be tricky for many of us, partly
due to tendencies in wider geek culture. When we see something
non-ideal, we tend to be quick to think of solutions that seem better to
us, and to want to share them, and it tends to be hard for us to leave
things alone to be implemented once we have made some input that might
be forgotten or misunderstood. We tend to think in terms of the
elegance of a correct solution, and be suspicious of lessons on social
leadership for being too often expressed in pop psychology.
I used to take more of a "do everything" approach myself as a student,
but I learnt from seeing an society where I'd taken on most of the work
run into immediate problems when I had to step back from it. In recent
years in DebConf, since I know that I have limited time, I have tried
not to over-commit myself with too many specific tasks, but to be a good
coordinator by following overall progress and allocating my time to the
specific needs at each moment, and finding people to help look at
problems where needed rather than trying to push my own solutions. If I
am elected DPL, I will continue the same approach in that position.
The same factors encourage us to hold on as long as possible to other
roles in Debian, in the fear that if we give them up someone might mess
up all the work we've put in or neglect important tasks. However, the
best way to ensure that our work carries on well is to train up
successors and pass on tasks to them early, and to make sure that there
is a real team of people working on a task rather than only one
indispensable person. For both the DPL and other Debian roles, we need
turnover of people to bring in new ideas, and teams which pool ideas
rather than just listen to top-down leaders.
 Of course, there is also some academic literature on leadership
that's at least as rigorous as non-theoretical computer science.