I’m in sunny and windy Berlin (Germany) again this weekend to attend Wikimedia’s “Hackathon” (Wikimedia is the organization behind Wikipedia and some other projects and websites, in case you didn’t know). I’d like to thank Wikimedia Foundation for the invitation.
Apart from discussing bug management with people I spent most of the day looking at the situation in Wikimedia’s bugtracker a bit – checking the growth rates in the most popular products (where to triage reports more aggressively), gathering some statistical data, trying out funky queries to identify rotting reports (no NEEDINFO state or tag? Sigh, Mozilla Bugzilla makes the same mistake), adapting one of my Greasemonkey scripts to work, looking at existing Wikimedia bug management documentation and identifying missing and redundant information, and in the coffee breaks dreaming of a Natural Language Processing hook when submitting new tickets to analyze the quality of a report on the fly and make the reporter improve it (I know that there’s scientific papers about this topic but I still haven’t seen any implementation so far). Errm, yeah, that was a long sentence.