I was chatting with a colleague on a feature that is loooooooong overdue and proposed to kill a meeting and use the time saved to write the damn thing.
He replied to me
Believe me, I would like to drop a meeting, not sure which one.
To which I replied
That's easy, anyone of them :)
and gave my thoughts on meetings.
If there is a need for bi-weekly meetings to integrate XXX and XXX, we've got a problem that meetings can't solve.
I have a radical take on meetings, especially regularly scheduled meetings:
- assuming n persons in the meeting you waste most of the time n-2 people's time (and if n>10, it's likely n-1 people's time)
- people tend to not prepare meetings. They instead think about the issue at stake while in the meeting and thus wasting n-1 people's time. Force people to write ideas in a (somewhat short) email, and that will force them to think about the issue more deeply and synthesize.
- a need for a regular scheduled meeting is a sign of lack of trust, lack of natural communication and/or lack of proper task isolation: in any case, better treat the problem at the source than patching with a meeting.
The key to open source success is multiple but one big component is extreme resource/time stress. This constraint leads to:
- very focused teams
- limited need for sync-up style communication (hence the usual small core team)
- proper separation of tasks to limit waste
I am not against communication, I am against communication wasting time (the asymptotic version being pure noise). I favor 1-1 communication personally as the most efficient brain-picking strategy.