[Jeremy Hylton] > + # The _p_atime field is seconds since the start of the day. > + # When we start a new day, we'll expect to see most of the > + # _p_atime values be less than now. We actually expect them to be greater than now, right? At the start of a new day, now will be close to 0. Also wonder if this is still sorting in the intended order (not obvious).