| From: | Greg Stark <gsstark(at)mit(dot)edu> |
|---|---|
| To: | Kevin Grittner <Kevin(dot)Grittner(at)wicourts(dot)gov> |
| Cc: | Alvaro Herrera <alvherre(at)commandprompt(dot)com>, "David E(dot) Wheeler" <david(at)kineticode(dot)com>, Pg Hackers <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: storing TZ along timestamps |
| Date: | 2011-05-27 22:54:05 |
| Message-ID: | BANLkTik3Sugo5xwsL--XG7R1hM=+jq1g5A@mail.gmail.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On Fri, May 27, 2011 at 2:32 PM, Kevin Grittner
<Kevin(dot)Grittner(at)wicourts(dot)gov> wrote:
> I think the key thing is that the timestamp portion of it would be
> identical to our current TIMESTAMP WITH TIME ZONE -- always store it
> in the value UTC zone.
Fwiw our timestamp with time zone stores seconds since the epoch. This
is a quantity which is independent of timezones entirely. Excluding
relativistic effects there have been the same number of time zones
since that point in time regardless of where you stand relative to the
sun.
My question for Alvarro is whether he really wants the text label for
the time zone at all, or just the offset which was used to enter it.
That is, if I enter "12:00pm" with my current time zone set to GMT and
later update the tzdata on the machine to start summer time on a
earlier date should the data type now show "1:00pm BST" or should it
still display "12:00pm +000" and leave it up to the reader to decide
whether why I entered it in a weird time zone for that time of year?
--
greg
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Steve Crawford | 2011-05-27 23:13:12 | Re: storing TZ along timestamps |
| Previous Message | Greg Stark | 2011-05-27 22:44:38 | Re: dblink crash on PPC |