On Thu, Oct 28, 2010 at 3:41 PM, Martin Pool <[email protected]> wrote: > We have something a bit like this in bug heat and "this bug affects me". > > I think it would be a bit hard for nonexpert users to estimate > difficulty or effort, though perhaps you could gather enough data to > be approximately correct. > > It seems to me the best thing to do through crowdsourcing is > de-duplication, which would both cut the absolute number of bugs, and > also give us good data about how many users are hitting something.
FWIW, crowd-sourcing this is dangerous. The Ubuntu project has long encouraged new contributors to help with bug triage. However, some of the most popular areas to file bugs on – X, network issues, sound – have bugs that look deceptively similar. Experts in these domains have asked us to turn our automatic duplicate filtering off, and have emphasized how *un*helpful de-duplication can be when done by novices. > I hope eventually to try to write a supervised machine learning system > to help with it. > Very cool. I hope to see it. :) jml _______________________________________________ Mailing list: https://launchpad.net/~launchpad-users Post to : [email protected] Unsubscribe : https://launchpad.net/~launchpad-users More help : https://help.launchpad.net/ListHelp
