Saturday, July 5, 2008

Re: [HACKERS] time_stamp type

"Stephen R. van den Berg" <srb@cuci.nl> writes:
> What's the deal with this type?

It's a domain over timestamptz, as required by the SQL spec definition
of the information_schema.

postgres=# \dD information_schema.time_stamp
List of domains
Schema | Name | Type | Modifier | Check
--------------------+------------+-----------------------------+----------------------------------------------------+-------
information_schema | time_stamp | timestamp(2) with time zone | default ('now'::text)::timestamp(2) with time zone |
(1 row)

[ re-reads spec... ] Hm, actually the spec is self-contradictory here:
SQL99 20.7 saith

CREATE DOMAIN TIME_STAMP AS TIMESTAMP (2)
DEFAULT CURRENT_TIMESTAMP(2);

which appears to imply that TIME_STAMP is a domain over timestamp
*without* time zone ... but that is contradicted by the specification
that the default is CURRENT_TIMESTAMP, which yields a value *with*
time zone. (LOCALTIMESTAMP is the function that should have been
mentioned if they really meant without time zone.)

[ pokes further... ] Hmm, last year's SQL200n draft saith

CREATE DOMAIN TIME_STAMP AS TIMESTAMP(2) WITH TIME ZONE;

with no mention of a default. I do wish these people could make
up their minds.

regards, tom lane

--
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

No comments: