Up until now the id of the first and last program entry was based on the
filtered range. Instead, we want these to be based on the first timeslot
that was scheduled before/after the specified range. This way we ensure
that the ids of these entries are stable as long as the schedule doesn’t
change.