On Mon, 18 Oct 2021 at 12:45, Paul Eggert via tz <
tz@iana.org> wrote:
This announcement is unclear, as it could mean
that the transition is:
(A) from Thursday 10-28 24:00 to Thursday 10-28 23:00.
And it could also mean that the transition is:
(B) from Friday 10-29 01:00 to Friday 10-29 00:00.
For Palestine's transition to winter time, tzdb currently uses method
(A) for 2014 and 2019, and method (B) for 2015-2018 and for 2020-2021.
Do you know which is the correct method for the year 2021? If method (B)
is correct, tzdb is already OK; if not, we'll need to update tzdb.
While this distinction is apt for the time-of-day (transition from 00:00/24:00 to 23:00 or 01:00 to 00:00), there is also the matter of day-of-week. In 2014 and 2015 we had the transition occurring on a Friday morning (Thursday night). Since then, our predictions have been on a Saturday morning (Friday night).
So either way the recent announcement is interpreted, it's looking like our current prediction (from Saturday 10-30 01:00 to Saturday 10-30 00:00) will be wrong, and would change the clocks 24 or 25 hours too late.