Re: Startup parameters timezone conversion

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: "David G(dot) Johnston" <david(dot)g(dot)johnston(at)gmail(dot)com>
Cc: Thomas Kellerer <shammat(at)gmx(dot)net>, "pgsql-jdbc(at)lists(dot)postgresql(dot)org" <pgsql-jdbc(at)lists(dot)postgresql(dot)org>
Subject: Re: Startup parameters timezone conversion
Date: 2022-01-19 15:07:59
Message-ID: 505670.1642604879@sss.pgh.pa.us
Views: Whole Thread | Raw Message | Download mbox | Resend email
Thread:
Lists: pgsql-jdbc

"David G. Johnston" <david(dot)g(dot)johnston(at)gmail(dot)com> writes:
> Indeed I presume it is. I wonder whether a carefully chosen timezone
> specification on the server would cause this to break since the server can
> be made to report the offset using either convention and so at least for
> some timezone specifications the flipping of the sign would not be required…

AFAIK, "reporting the offset" is always done with the ISO convention.
It's only when trying to interpret a time zone specification that
we consider the POSIX convention (and that's mostly because the
underlying tzdb code does so). This does lead to fun stuff like

postgres=# set timezone = 'GMT+2'; -- read as POSIX zone spec
SET
postgres=# select now();
now
-------------------------------
2022-01-19 13:03:36.000152-02 -- report as ISO
(1 row)

postgres=# set timezone = '+2'; -- read as numeric ISO offset
SET
postgres=# select now();
now
-------------------------------
2022-01-19 17:03:41.722767+02 -- report as ISO
(1 row)

IMO, all these cases are best-avoided legacy conventions.
In practice you should set the timezone using the tzdb zone name
for where you live, e.g. America/New_York.

regards, tom lane

In response to

Responses

Browse pgsql-jdbc by date

  From Date Subject
Next Message Dave Cramer 2022-01-19 15:08:45 Re: Startup parameters timezone conversion
Previous Message Tom Lane 2022-01-19 14:54:12 Re: Startup parameters timezone conversion