oops this is repeated

this parameter is no longer used and should disappear right?

this seems right in the case where end was not explicitly passed in and is intended to represent the current time - but if it was passed in, end is (I believe) intended to be an end partition time, not an end execution time.

(i.e. if I do daily_schedule(start=today, end=2 days from now, day_offset=1)

The expected behavior is that it run for 3 days starting tomorrow (not 2 days)

Otherwise this makes a ton of sense to me, in the common case where you set a start time but no end time, no point in returning a partition that hasn't executed yet (at least where we have the cron scheduler which is always looking at the last partition)

This diff has made me realize that the cron scheduler might just not be working at all when you set an explicit end_date, incidentally... probably not the end of the world since we are about to remove the cron scheduler entirely

my feedback above is making me realize that 'a schedule with an end_date explicitly set' is conspiciously absent from these tests :/ maybe include one here just for daily? And as per feedback above, that would want to go through the end_date, not partition_days_offset days before the end_date (i.e. that test should have existed before and this change as currently written probably should have broken it)