Michael Monnerie writes: > On Mittwoch, 5. April 2006 19:22 Justin Mason wrote: > > (For what it's worth, 0830 UTC was chosen as an hour that's not > > during working hours for most of the committers; 0000 UTC isn't good > > on that count, since it's 1600 PST.) > > I don't understand why my results made a problem then - process start > was 0303 and ended around 0343 MEST, which should be 0103-0143 UTC. If > the switch was done at 0830 UTC, that shouldn't have raised a bug.
the problem is: - the mass-check must take place after the 0830 UTC tag time; - *and* must take place in the same UTC "day" period as the other mass-checks. So your mass-check was satisfying #1, but not #2. However that has given me an idea ;) If I change the scripts so that each "day" period isn't considered as running from 0000-2359 UTC, and instead is considered as 0800-0759 UTC, that may be a good workaround... --j.