Henry explains his view on the decision process at Jakarta. An interesting read, and I share his opinion except that what he names "meritocracy" should better be named "doocracy". Meritocracy is the recognition people gain from what they've done in the past, and not from what they're doing right now.

Doocracy is good if it doesn't turn into despotism or imposed decisions, which can happen if later community-driven refactoring is not welcomed. The Iwilldoitocracy ("I think XYZ should happen, if I don't hear a -1 by XXXDay, I'll go ahead and do it.") is a better choice as it triggers discussion with interested people, enriching the solution with their opinions and leading to a better result, known by a larger group of people.

That's how it usually happens in Cocoon. Most ideas and proposals posted to the cocoon-dev list end up with "WDYT?" ( what do you think), thus seeking opinions and discussions.

Of course, it happens sometimes that opinions are so different or diverging that consensus is difficult to reach, in which case votes or doocracy allow the decision to happen.

But most often consensus is quickly reached, and the one who writes down the code, even if the initiator of the change, also becomes the scribe of the collective thinking of the development team.

What makes the difference regarding the meritocracy is both people's talent and the time they can devote to opensource development. You can't change much to the first one, and you would like so much to have more of the second one :-)