Oops, you are right. The weirdness starts in 2072:

Asia/Hebron  Fri Mar 25 23:59:59 2072 UT = Sat Mar 26 01:59:59 2072 EET isdst=0 gmtoff=7200
Asia/Hebron  Sat Mar 26 00:00:00 2072 UT = Sat Mar 26 03:00:00 2072 EEST isdst=1 gmtoff=10800
Asia/Hebron  Fri Sep  9 22:59:59 2072 UT = Sat Sep 10 01:59:59 2072 EEST isdst=1 gmtoff=10800
Asia/Hebron  Fri Sep  9 23:00:00 2072 UT = Sat Sep 10 01:00:00 2072 EET isdst=0 gmtoff=7200
Asia/Hebron  Fri Oct 14 23:59:59 2072 UT = Sat Oct 15 01:59:59 2072 EET isdst=0 gmtoff=7200
Asia/Hebron  Sat Oct 15 00:00:00 2072 UT = Sat Oct 15 03:00:00 2072 EEST isdst=1 gmtoff=10800
Asia/Hebron  Fri Oct 28 22:59:59 2072 UT = Sat Oct 29 01:59:59 2072 EEST isdst=1 gmtoff=10800
Asia/Hebron  Fri Oct 28 23:00:00 2072 UT = Sat Oct 29 01:00:00 2072 EET isdst=0 gmtoff=7200
Asia/Hebron  Fri Mar 24 23:59:59 2073 UT = Sat Mar 25 01:59:59 2073 EET isdst=0 gmtoff=7200
Asia/Hebron  Sat Mar 25 00:00:00 2073 UT = Sat Mar 25 03:00:00 2073 EEST isdst=1 gmtoff=10800
Asia/Hebron  Fri Sep  1 22:59:59 2073 UT = Sat Sep  2 01:59:59 2073 EEST isdst=1 gmtoff=10800
Asia/Hebron  Fri Sep  1 23:00:00 2073 UT = Sat Sep  2 01:00:00 2073 EET isdst=0 gmtoff=7200
Asia/Hebron  Fri Oct  6 23:59:59 2073 UT = Sat Oct  7 01:59:59 2073 EET isdst=0 gmtoff=7200
Asia/Hebron  Sat Oct  7 00:00:00 2073 UT = Sat Oct  7 03:00:00 2073 EEST isdst=1 gmtoff=10800
Asia/Hebron  Fri Oct 27 22:59:59 2073 UT = Sat Oct 28 01:59:59 2073 EEST isdst=1 gmtoff=10800
Asia/Hebron  Fri Oct 27 23:00:00 2073 UT = Sat Oct 28 01:00:00 2073 EET isdst=0 gmtoff=7200

By my reading, that's 4 transitions per year.

Cheers,

Dave Rolsky


On Sat, Dec 30, 2023 at 2:08 PM Dave Rolsky <autarch@urth.org> wrote:
Hi all,

With the latest release (2023d) I'm seeing some surprising output from zdump for the Asia/Gaza & Asia/Hebron zones:

$> zdump -v Asia/Hebron
...
Asia/Hebron  Fri Mar 23 23:59:59 2068 UT = Sat Mar 24 01:59:59 2068 EET isdst=0 gmtoff=7200
Asia/Hebron  Sat Mar 24 00:00:00 2068 UT = Sat Mar 24 03:00:00 2068 EEST isdst=1 gmtoff=10800
Asia/Hebron  Fri Oct 19 22:59:59 2068 UT = Sat Oct 20 01:59:59 2068 EEST isdst=1 gmtoff=10800
Asia/Hebron  Fri Oct 19 23:00:00 2068 UT = Sat Oct 20 01:00:00 2068 EET isdst=0 gmtoff=7200
Asia/Hebron  Fri Mar 29 23:59:59 2069 UT = Sat Mar 30 01:59:59 2069 EET isdst=0 gmtoff=7200
Asia/Hebron  Sat Mar 30 00:00:00 2069 UT = Sat Mar 30 03:00:00 2069 EEST isdst=1 gmtoff=10800
Asia/Hebron  Fri Oct 11 22:59:59 2069 UT = Sat Oct 12 01:59:59 2069 EEST isdst=1 gmtoff=10800
Asia/Hebron  Fri Oct 11 23:00:00 2069 UT = Sat Oct 12 01:00:00 2069 EET isdst=0 gmtoff=7200

This pattern of >2 transitions per year continues from that year.

Is this correct? It seems awfully weird.


Cheers,

Dave Rolsky