Server papercuts: a Lucid retrospective
Last week we froze the Ubuntu archives in preparation for the 10.04 LTS Beta2 milestone release. With this Beta2Freeze came the end of the Lucid Server papercuts effort, so now is a great time to analyze what went well and what went wrong.
Nomination: We got 43 bugs nominated, which is not too much, not too little. We had good community participation on that phase of the project. One common misunderstanding was that we couldn’t really introduce new features or a change in behavior, given that we started the effort past-FeatureFreeze.
Bugfix: We fixed 9 bugs for beta1 and 10 bugs for beta2, leaving only two of them unfixed at the end (the others were invalidated as a papercut by the bugfixer). 6 of the papercuts were fixed by community members outside of Canonical. The Freeze dates left us about 10 days for each round. Since those were relatively-easy-to-fix bugs, I’d have loved to have new contributors involved in fixing them, but unfortunately that wasn’t the case. All the papercuts were fixed by tried-and-true Ubuntu Server members.
I also was disappointed we couldn’t fix more true papercuts, the things that are just painful and never solved otherwise. Most of the bugs we solved were mildly annoying issues. Does that mean there is no major annoyance left in Ubuntu Server ? Or did we miss them in the nomination phase ? Or we couldn’t fix them because that would have involved a new feature and we were past-FeatureFreeze ?
What should we do in the future ? Should we always have a Server papercuts effort, and not just for LTS releases ? Should we start fixing them earlier, so that we can also work on easy-to-implement, missing features ? Feel free to leave your comment here ! We should have a session about specific Server LTS QA efforts (papercuts, bugzapping) and how much we should pursue them for Maverick at UDS in Brussels next month. Come and join us !