[tz] [PROPOSED] Treat 1931 changes in Mexico as DST

[Paul Eggert]

On 9/29/22 11:25, Michael H Deckers wrote:
>     the (incorrect) offset of -06:40:00 h presently used for
>     local mean time at Matamoros.

Thanks for reporting that bug. It should be -06:30:00; apparently I 
transcribed the number incorrectly from Shanks & Pottenger.

While reviewing this I noticed another bug: the 1922 transition in 
Tijuana was off by an hour due to my mistake in subtracting timestamps. 
The first attached patch fixes both of these problems.

For a transition out of LMT to a whole-number standard time, it's better 
to use UT for the transition time, so I installed the second attached 
patch to do this more systematically in US timestamps (also, one 
transition in Spanish Morocco). This second patch doesn't affect the 
generated timestamps.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-Circa-1922-Mexico-fixes.patch
Type: text/x-patch
Size: 5193 bytes
Desc: not available
URL: <http://mm.icann.org/pipermail/tz/attachments/20220930/5a7d836a/attachment.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0002-Prefer-UT-for-whole-hour-UT-transitions.patch
Type: text/x-patch
Size: 9577 bytes
Desc: not available
URL: <http://mm.icann.org/pipermail/tz/attachments/20220930/5a7d836a/attachment-0001.bin>