Oscar van Vlijmen <ovv@hetnet.nl> writes:
Some LMT (local mean time) values in the tz database seem to be incorrect. Below are the values I believe to be correct.
Alaska America/Juneau -8:57:41 Alaska America/Anchorage -9:59:36 Alaska America/Nome -11:01:38 Alaska America/Adak -11:46:38 Philippines Asia/Manila 8:04:00 Guam Pacific/Guam 9:39:00 Northern Mariana Islands Pacific/Saipan 9:43:00 Samoa Pacific/Apia -11:26:56 American Samoa Pacific/Pago_Pago -11:22:48
These LMT offsets agree with the current tz database. I would guess that your procedure got confused by the fact that these locations started out on the opposite side of the International Date Line, compared to where they are now.
Alaska America/Yakutat -9:14:55
Are you sure about that one? <http://www.welcometoalaska.com/yakutat.htm> says Yakutat is 139.72722 degrees west, which corresponds to -9:18:55, which is what is in the current tz database.
Note that West Samoa is no longer the correct name; it is Samoa.
Thanks; I'll update australasia and iso3166.tab accordingly.
No LMT information was given for:
BIOT Indian/Chagos 4:49:40 Cocos Islands Indian/Cocos 6:27:40
This was because we don't know when standard time was introduced in these locations. Might as well put something in rather than nothing, so I'll guess 1907 and 1900, respectively. 1907 because Chagos was part of Mauritius at the time, and we have Mauritius introducing standard time in 1907. 1900 because all we know is that it must have occurred while the Ross family was ruling Cocos (1830-1978) and we might as well guess 1900 as any other time.
Some time ago I looked the latitude/longitude up from various sources, some more reliable than others. Locations missing in the tz database are:
Thanks; I'll update the base info accordingly. (One minor point: I don't bother to keep coordinates in "antartica" if they're already in zone.tab, or if they are for large regions like the South Orkneys.)