On Thursday 2013-07-11 00:14 -0700, Robert O'Callahan wrote: > We can't have a rigid rule about 24 hours. Someone requesting a review from > me on Thursday PDT probably won't get a response until Monday if neither of > us work during the weekend. > > But I think it's reasonable to expect developers to process outstanding > review requests (and needinfos) at least once every regular work day. > Processing includes leaving a comment with an ETA.
So, partly, I'm really bad at figuring out ETAs for small tasks, since I find the priority order of small tasks to be relatively dynamic, given how often small things come up. For example, reading and responding to this thread (something I've spent at least 15 minutes on so far this morning, and it'll probably end up being more than 30, which is probably 10% of my non-meeting working day today). Should I have prioritized that above doing code reviews, or should I come back to this thread and give you my thoughts in 3 weeks (as I'm currently doing on the prefixing policy thread, which I feel requires more thought)? I spend a pretty big portion of my time on things that come up at the last minute: questions from colleagues, discussions on lists (Mozilla lists and standards lists, etc.). (Another interesting question: should I prioritize questions / needinfos from people in the *middle* of writing a patch over code reviews which are at the *end* of writing a patch? Right now I think I sometimes do, and sometimes treat them equally.) Like Boris, I feel guilty about not getting to reviews, and I feel like I'm bad at figuring out how to prioritize them. I suppose what leaving an ETA would do is force me to try to stick to what I've promised, which in turn means doing code reviews rather than doing things like reading email or responding to this thread. -David -- 𝄞 L. David Baron http://dbaron.org/ 𝄂 𝄢 Mozilla http://www.mozilla.org/ 𝄂 _______________________________________________ dev-platform mailing list dev-platform@lists.mozilla.org https://lists.mozilla.org/listinfo/dev-platform
