On Sun, 24 Oct 2021 at 21:25, Philip Paeps via tz <
tz@iana.org> wrote:
- NEWS: Mention this.
- asia (Palestine): Guess Fri>=23, not lastFri, for fallback
transitions in the future. Fri>=23 is more consistent with past
practice than lastFri. The first differing prediction is a
fallback on 2025-10-24, not 2025-10-31.
This seems reasonable. It's still only a prediction but improving the likelihood of not having to fix the transition date with fewer than two weeks notice for a couple of years would be a win.
It's also similar to how we've been recently handling Palestine's March transition. Recently, both transitions seem to be more-or-less pinned to the last Sunday of the month (à la EU rules) so things don't get too badly out-of-sync, but with the actual change taking place on a more-locally-favorable day of that same weekend (which is the part that has been less consistent, of late).
I've installed the attached patch to the commentary since it was left unchanged despite the improvement to the prediction.