Content-Type: text/plain;
  charset="us-ascii"
From: Josh Berkus <josh@agliodbs.com>
Reply-To: josh@agliodbs.com
Organization: Aglio Database Solutions
To: pgsql-sql@postgresql.org
Subject: Bug with Daylight Savings Time & Interval
Date: Mon, 20 May 2002 15:34:32 -0700
Cc: pgsql-bugs@postgresql.org
MIME-Version: 1.0
Content-Transfer-Encoding: quoted-printable
Message-Id: <200205201534.32937.josh@agliodbs.com>

Folks,

Found this interesting bug:

jwnet=3D> select version();
                            version
---------------------------------------------------------------
 PostgreSQL 7.2.1 on i686-pc-linux-gnu, compiled by GCC 2.95.3
(1 row)

jwnet=3D> select ('2001-07-31 10:00:00 PST'::TIMESTAMP) + ('248 days'::INTE=
RVAL)=20
;
        ?column?
------------------------
 2002-04-05 10:00:00-08
(1 row)

jwnet=3D> select ('2001-07-31 10:00:00 PST'::TIMESTAMP) + ('249 days'::INTE=
RVAL)=20
;
        ?column?
------------------------
 2002-04-06 10:00:00-08
(1 row)

jwnet=3D> select ('2001-07-31 10:00:00 PST'::TIMESTAMP) + ('250 days'::INTE=
RVAL)=20
;
        ?column?
------------------------
 2002-04-07 11:00:00-07

jwnet=3D> select ('2001-04-01 10:00:00 PST'::TIMESTAMP) + ('100 days'::INTE=
RVAL)=20
;
        ?column?
------------------------
 2001-07-10 11:00:00-07


It appears that Spring Daylight Savings Time causes PostgreSQL to change my=
=20
time zone.  Only the spring, mind you, and not the fall.   This is=20
potentially catastrophic for the application I'm developing; what can I do =
to=20
see that it's fixed?  Or am I misunderstanding the behavior, here?

--=20
-Josh Berkus

P.S. I'm posting this here instead of the online bug form because I know th=
at=20
Bruce is on vacation.