[tz] Definition of "timezone"

Dale Ghent daleg at elemental.org
Thu Dec 21 23:12:27 UTC 2023

On 2023-12-21 07:08, ApoY2k wrote:
> do you (or your
> peers on the tz project) consider a "timezone" to be inherently tied to the
> real-world geographical features they are used for, or is the word
> "timezone" in this context more broader understood as "any 'Zone' entry in
> the tzdb" - which would e.g. also denote "Etc/UTC" as a "real timezone".


Yeah I've always carried the nagging feeling that the terminology can be confusing. If you google "time zone map", many of the links you get will display a map like the one here:


Obviously the time zones depicted on that map and others like it are far more generalized than the time zones that the TZDB defines. The IANA TZDB reflects the geopolitical differences in accounting for the number of hours offset from GMT that a completely arbitrary place has. This place can be an entire country or a lone Antarctic research station; its's always(?) some subset of the large swaths of land depicted on those "standard time zone" maps. Because of this, it has always felt a little strange to me to refer to the TZDB as the "time zone database" because its more detailed contents reflect nothing like the "standard time zone map" depicted at the above URL, and vice versa.

Because of those innately human and arbitrary differences, I've always wondered if the TZDB should be thought of more as a "time locale database." This is borrowing a concept from the internationalization and localization (i18n, i10n) systems in computing that exist to address another innately human thing - language. I feel like use of "locale" to describe these entries in the TZDB (TLDB?) better-conveys the local and morphing nature of a given area's sense of time, kind of like how language is thought of, I guess?

For example, the standard time zone known as GMT-6 contains central Canada, the central United States, and much of Mexico and Central America. Looking at the above reference map, you might think that all people in this time zone follow the same rules and are always on the same clock. We know that's not the case - the US and most of Canada still observe DST rules which periodically alter this time zone's effective GMT offset in those areas, but some parts of Canada, all of Mexico, and all other counties in that time zone do not observe DST even though they are geographically within the time zone known as GMT-6. I haven't done an exhaustive search, but I don't doubt that there are two or more countries or areas within the same "standard time one" that all observe DST, but at different start and end times. These break that one large GMT-6 time zone - and any standard time zone for that matter - up in to what I think should be referred to as "time locales".

I don't know, I'm new around here and am bracing myself for possible incoming fire because of that thought. :D

As for UTC, I feel that it being in the database is for convenience. We all know UTC is not a time zone, but rather a time coordinate. It just happens to coincide with GMT, so why not? I maintain an OS *and* have deep involvement in astronomy and astrometry so the whole UTC/GMT/UT1 association here amuses me. It doesn't bother me.


More information about the tz mailing list