A New Rule? (was McMurdo/South Pole)
A partly-baked idea to solve the more general problem of when zones appear and disappear: would it make sense to have another kind of entry in the zone RULES column to indicate that we should now follow another zone's practice? How about the keyword, Zone, a space, and the name of the other zone, with the next column following whitespace, if present, being the [UNTIL] column?
On Sun, 22 Sep 2013, Bill Seymour wrote:
A partly-baked idea to solve the more general problem of when zones appear and disappear: would it make sense to have another kind of entry in the zone RULES column to indicate that we should now follow another zone's practice? How about the keyword, Zone, a space, and the name of the other zone, with the next column following whitespace, if present, being the [UNTIL] column?
I liked the syntax proposed by Andy Lipscomb on 5 Sep 2013, with "=" as a magic character in the GMTOFF, RULES, and FORMAT fields of a Zone record. See <http://mm.icann.org/pipermail/tz/2013-September/019952.html>. He gave this example: Zone America/North_Dakota/New_Salem -6:45:39 - LMT 1883 Nov 18 12:14:21 =America/Denver = = 2003 Oct 26 02:00 =America/Chicago = = which would mean: America/North/Dakota/New_Salem used LMT until a specified date, then followed the same rules as America/Denver until a specified date, and now follows the same rules as America/Chicago into the future. --apb (Alan Barrett)
On Sun, Sep 22, 2013 at 6:04 AM, Alan Barrett <apb@cequrux.com> wrote:
I liked the syntax proposed by Andy Lipscomb on 5 Sep 2013, with "=" as a magic character in the GMTOFF, RULES, and FORMAT fields of a Zone record. See <http://mm.icann.org/pipermail/tz/2013-September/019952.html>. He gave this example:
Zone America/North_Dakota/New_Salem -6:45:39 - LMT 1883 Nov 18 12:14:21 =America/Denver = = 2003 Oct 26 02:00 =America/Chicago = =
I seem to be behind the times. 8-) I agree that that's better. My only thought is that maybe the first '=' should be the whole GMTOFF, and the other zone be the RULES. That eliminates one '=' at the end and maybe makes the line easier to parse. Or maybe not. I need to read the cited paper.
On Sun, Sep 22, 2013 at 6:33 AM, I wrote:
On Sun, Sep 22, 2013 at 6:04 AM, Alan Barrett <apb@cequrux.com> wrote:
I liked the syntax proposed by Andy Lipscomb ...
Zone America/North_Dakota/New_Salem -6:45:39 - LMT 1883 Nov 18 12:14:21 =America/Denver = = 2003 Oct 26 02:00 =America/Chicago = =
My only thought is that maybe the first '=' should be the whole GMTOFF, and the other zone be the RULES. That eliminates one '=' at the end and maybe makes the line easier to parse. Or maybe not. I need to read the cited paper.
Well, that was an easy read. 8-) Can anyone think of a use case for "use this zone but with a different set of rules"? If not, then I would suggest making the other zone name be the RULES as I said.
On Sun, 22 Sep 2013, Bill Seymour wrote:
I liked the syntax proposed by Andy Lipscomb ...
Zone America/North_Dakota/New_Salem -6:45:39 - LMT 1883 Nov 18 12:14:21 =America/Denver = = 2003 Oct 26 02:00 =America/Chicago = =
Can anyone think of a use case for "use this zone but with a different set of rules"?
Yes, I can easily imagine wanting to say "Use the same GMTOFF as this other zone, but without DST rules." It may be unlikely to be needed, but I wouldn't want the syntax to forbid that possibility. --apb (Alan Barrett)
participants (2)
-
Alan Barrett -
Bill Seymour