[tz] Belarus is listed in MSK timezone

[Guy Harris]

On Apr 3, 2015, at 2:28 PM, Dzmitry Kazimirchyk <dkazimirchyk at gmail.com> wrote:

> The key concept here is that "the rule itself doesn't change". I don't believe MSK can be considered a rule set

Does *anybody* here believe it can?  Not as far as I know.

The "europe" file has

Zone    Europe/Minsk    1:50:16 -       LMT     1880
                        1:50    -       MMT     1924 May  2 # Minsk Mean Time
                        2:00    -       EET     1930 Jun 21
                        3:00    -       MSK     1941 Jun 28
                        1:00    C-Eur   CE%sT   1944 Jul  3
                        3:00    Russia  MSK/MSD 1990
                        3:00    -       MSK     1991 Mar 31  2:00s
                        2:00    1:00    EEST    1991 Sep 29  2:00s
                        2:00    -       EET     1992 Mar 29  0:00s
                        2:00    1:00    EEST    1992 Sep 27  0:00s
                        2:00    Russia  EE%sT   2011 Mar 27  2:00s
                        3:00    -       FET     2014 Oct 26  1:00s
                        3:00    -       MSK

so the zone name is named "Europe/Minsk".

That zone uses various rule sets; it's currently not using *any* rule set, although it's used the "C-Eur" and "Russia" rule sets in the past.  (The zic man page speaks of "set[s] of rules" rather than "rule set[s]", but those are just different names for the same thing.)

You're talking about the *abbreviation* used when formatting times, not about the name for a rule set.  "MSK" - and "FET", and so on - are abbreviations used when formatting times, not names for rule sets; it would be as big a mistake to refer to "MSK" as the name of a rule set as it would be to refer to "EST" or "EDT" or "PST" or "PDT" or "CET" as the names of rule sets.

And, in fact, there's no "CET" rule set; there's a "C-Eur" rule set, and there are some zones that use the "C-Eur" rule set but *don't* use CET or CEDT or anything such as that as the abbreviation (Europe/Sofia and Europe/Bucharest used EE%sT with "C-Eur", and Europe/Luxembourg used WE%sT with "C-Eur", at one point, for example).

Rule sets and abbreviations are separate things.

So the question here is what should the abbreviation be.  I don't know where the notion of designating times as, say, "5PM EST", for 17:00 in the Eastern time zone when daylight savings time/summer time is not in effect, originated; in particular, I don't know whether it started in the English-speaking world and, if it did, what countries and cultures have adopted it.

It may be that Belarus never adopted it; Paul's mail shows some Belarusian sites showing times as, for example, "19h00 Minsk time".  The UN*X tradition, somewhat standardized in the Single UNIX Specification description of the TZ environment variable, has 3-or-more-character strings not containing spaces, so "Minsk time" wouldn't work for that.

That's why abbreviations such as "MSK" were invented.

It Might Be Nice if there were standard UN*X APIs to get more descriptive zone names.  On Windows, _tzname[] contains strings such as "Pacific Daylight Time", so it could presumably say "Minsk time" in Belarus (although it'd probably say it in Belarusian by default).  The CLDR appears to include descriptive zone names.

However, we probably still need to supply an abbreviation.

If abbreviations (rather than phrases such as "Minsk time") are used in Belarus, what abbreviation is used?  Our database isn't internationalized - that's the CLDR's job - so that would be "used in English-language text in Belarus".

If abbreviations *aren't* used in Belarus, are they used in any other countries when talking about Minsk time?  If so, what abbreviation is used in English-language text?

If they're not used *anywhere*, we have to pick one; what one would you recommend?