RE: Message to `tz' mailing list
I am confused about the use of the UTC/local indicators and the standard/wall indicators in a tzfile. According to tzfile(5), they "are used when a time zone file is used in handling POSIX-style time zone environment variables". Does this mean that these indicators are irrelevant when computing local time using only a tzfile (without looking at the environment).
That's correct: the ttisgmt and ttisstd indicators can be ignored if you're simply trying to figure out what wall clock time is for the zone that's associated with the file. (The indicators are there for the case when the time zone file is used as a "template" for the DST rules to be used when handling POSIX-style TZ environment variables; you need to know whether the transition time is defined in terms of UTC, local wall clock time, or local standard time to be able to know when translating the transition time for use in another zone.) --ado
At 2006-02-16T09:30:07-05:00, Olson, Arthur David (NIH/NCI) [E] wrote:
the ttisgmt and ttisstd indicators can be ignored if you're simply trying to figure out what wall clock time is for the zone that's associated with the file.
Thank you very much for the clarification. Raghavendra. -- N. Raghavendra <raghu@mri.ernet.in> | See message headers for contact Harish-Chandra Research Institute | and OpenPGP details.
This is not ticking in on my computer screen, e.g. I don't understand the argument that IsGmt and IsStd are irrelevant when you only want to figure out wall clock time making a program reading the tz file. However, I may have misunderstood Raghavendra or Arthur David Olson. In my view the program has to take it into account explicitly, so that 2:00w should correspond to wall clock time, but 2:00u and 2:00s should be converted to wall clock time base on GMT offset and daylight saving offset. Let's look at Mexico City. It has GMT offset -6 and daylight saving offset +1 between First Sunday of April at 2:00 and Last Sunday of October at 2:00. Both of these could be stated as 2:00w instead if desired. But if we use u or s it changes the meaning: Last Sunday of October 2:00u means that it was 21:00w the day before in Mexico because in October before DST ends there is 5 hour difference to GMT (-6+1). Last Sunday of October 1:00s means that it was 2:00w because we have a daylight saving offset of +1 just before DST ends. But the next second we will have 1:00:01s and 1:00:01w because now the daylight saving offset is no longer in effect. And Last Sunday of October 2:00s would therefore be 2:00w for the second time that day. I don't know if this was what you were discussing? Jesper Nørgaard Welen -----Original Message----- From: Olson, Arthur David (NIH/NCI) [E] [mailto:olsona@dc37a.nci.nih.gov] Sent: Thursday, February 16, 2006 08:30 To: tz@lecserver.nci.nih.gov Subject: RE: Message to `tz' mailing list
I am confused about the use of the UTC/local indicators and the standard/wall indicators in a tzfile. According to tzfile(5), they "are used when a time zone file is used in handling POSIX-style time zone environment variables". Does this mean that these indicators are irrelevant when computing local time using only a tzfile (without looking at the environment).
That's correct: the ttisgmt and ttisstd indicators can be ignored if you're simply trying to figure out what wall clock time is for the zone that's associated with the file. (The indicators are there for the case when the time zone file is used as a "template" for the DST rules to be used when handling POSIX-style TZ environment variables; you need to know whether the transition time is defined in terms of UTC, local wall clock time, or local standard time to be able to know when translating the transition time for use in another zone.) --ado
participants (3)
-
Jesper Norgaard Welen -
N. Raghavendra -
Olson, Arthur David (NIH/NCI) [E]