I'm trying to set [now(), 2049-12-31 00:00:00) as the default for a
tsrange column (Postgres 9.2), but can't figure out how to do it. I'm
either getting syntax errors or now() is being evaluated, so that the
default becomes something like [2012-07-14 14:04:35, 2049-12-31
00:00:00), which is not
Chris Bartlett writes:
I'm trying to set [now(), 2049-12-31 00:00:00) as the default for a
tsrange column (Postgres 9.2), but can't figure out how to do it. I'm
either getting syntax errors or now() is being evaluated, so that the
default becomes something like [2012-07-14 14
At 8:35 PM +0100 16/7/12, Daniele Varrazzo wrote:
Even without this quirk, the problem of mapping timestamps to other
languages data types could be an even stronger design factor. I've
personally settled for -12-31 which is python's datetime.max, maps
ok to doubles and won't create problems f
I am not sure if this is expected behaviour or a bug.
Using PG 9.2 beta 2 and PGAdmin3 1.16 beta 2.
Connect as bob (superuser)
In public schema:
create table people (cols...)
create view people_view as select * from people
Create schema bob
create table bob.people (cols...)
create view
At 7:37 PM -0700 25/7/12, Adrian Klaver wrote:
I am guessing if you do show search_path; from psql you will see
that the public schema is before the bob schema. The SELECT for the
unqualified people table in CREATE VIEW bob.people_view will find
public.people first in that case.
I don't thin
At 7:51 PM -0700 25/7/12, Adrian Klaver wrote:
On 07/25/2012 07:47 PM, Chris Bartlett wrote:
At 7:37 PM -0700 25/7/12, Adrian Klaver wrote:
I am guessing if you do show search_path; from psql you will see that
the public schema is before the bob schema. The SELECT for the
unqualified people
Date and time functions like current_time return the client machine's
time. Is there a way of getting the database server's time? I have a
situation that requires comparison of a date stamp on records with
"today", but I need to avoid the possibility of a user changing their
computer's clock ti
