Krzysztof Kowalczyk wrote:
I think the first step should be: put up a list of wanted features, period, and advertise it, without the bounty system. Bounty can always be added later. [...] Yes, I know there's a bug system and I know there are wiki page that list a lot of ideas on what can be done. But those things are chaotic and unprioritized i.e. looking at them I don't really know which bugs/features are regarded as highly desirable and which ones are just blue sky ideas that wouldn't even be accepted into CVS when done.
Of course, if we had BugZilla instead, this would be trivial. With BugZilla, you can give each bug a priority, and then you can directly link to a list of all bugs with a given priority. This is just an example; there are hundreds of thousands of ways of cataloguing/categorising bugs in BugZilla.
Maybe instead of the Bounty system, I should volunteer to help install BugZilla on Wikimedia instead?
Currently, as I percieve it, one of the barriers to entry for new developers is not knowing what to work on.
That's certainly a part of it, too, yes. If we set up BugZilla and find that it was a miracle cure, there will no longer be a need for payments and rewards.
What is missing is a farily short (say <10 items at any given time), prioritized list of bugs/features that are known to by highly desirable/high priority but not done. At least I'm not aware of such list.
Of course, it would be easy for the People In Charge™ (Board of Trustees, say) to make sure that the number of bugs with top priority is kept at a certain maximum, and the direct link I mentioned above will always be valid.
Timwi