I'm still wondering what's the best way to discuss design questions in open source projects. The usual process is to start a RT in the mailing list and then discuss it until a consensus is reached.

The past tell us that this often leads to the point where a lot has been discussed, no consensus has been reached and all possibilites to implement a feature are still open. So in the end, noone is able to decide who's wrong and who's right. The person initiating the RT is not able to tell if his RT is good and if he *is allowed* to implement it. Or if he will be banned from the community then :)

Currently, I think it's much easier to implement a RT before posting it and then discussing the feature by looking at the implementation. Because then there is something already working and one can see what's wrong with this approach. And there is one who is able to implement it.

10:50:57 AM