From: Tom Lane <tgl@sss.pgh.pa.us>
To: Sbob <sbob@quadratum-braccas.com>
cc: pgsql-admin@lists.postgresql.org
Subject: Re: time data type question
In-reply-to: <b28e1928-7822-46aa-a0c8-ad4e985d721f@quadratum-braccas.com>
References: <b28e1928-7822-46aa-a0c8-ad4e985d721f@quadratum-braccas.com>
Comments: In-reply-to Sbob <sbob@quadratum-braccas.com>
	message dated "Fri, 06 Dec 2024 12:32:50 -0700"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <2847788.1733514522.1@sss.pgh.pa.us>
Date: Fri, 06 Dec 2024 14:48:42 -0500
Message-ID: <2847789.1733514522@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/2847789.1733514522%40sss.pgh.pa.us>
Precedence: bulk

Sbob <sbob@quadratum-braccas.com> writes:
> I thought I would see the time columns shift to mountain time as well. 
> am I doing something wrong?

timetz acts completely differently from timestamptz.  timetz
actually stores two fields, a time-of-day (microseconds since
midnight I think) and a timezone expressed as numeric offset
from UTC.  Once stored, the value does not react to changes
in the timezone setting.

By and large we consider timetz deprecated.  It's there
because the SQL spec requires it, but it doesn't behave in
a way that most people find useful.

			regards, tom lane