Re: timestamp format bug - Mailing list pgsql-hackers

From Kevin Grittner
Subject Re: timestamp format bug
Date
Msg-id 47A1D15A.EE98.0025.0@wicourts.gov
Whole thread Raw
In response to Re: timestamp format bug  ("Roberts, Jon" <Jon.Roberts@asurion.com>)
Responses Re: timestamp format bug  ("Roberts, Jon" <Jon.Roberts@asurion.com>)
List pgsql-hackers
>>> On Thu, Jan 31, 2008 at 12:45 PM, in message
<1A6E6D554222284AB25ABE3229A92762715527@nrtexcus702.int.asurion.com>, "Roberts,
Jon" <Jon.Roberts@asurion.com> wrote:

> So on your db, run this query:
> select sub.t1, to_char(t1, 'yyyy-mm-dd hh24:mi:ss.us') as char_t1
> from
> (
> select timestamp'2008-01-31 12:31:40.500000' as t1
> ) sub
>
>
> I bet you get this:
> "2008-01-31 12:31:40.50";"2008-01-31 12:31:40.500000"          t1           |          char_t1
------------------------+----------------------------2008-01-31 12:31:40.50 | 2008-01-31 12:31:40.500000
(1 row)
> Don't you think it should have two identical columns?
No.  Why should the return value of a function influence the input?
What would you expect from this query?:
select sub.t1, substring(sub.t1 from 2 for 3) as substring_t1
from
(
select 'abcde'::text as t1
) sub
> Secondly, this link shows that ms should be 000-999 and us should be
> 000000-999999.
Yes.  That seems to me to work fine, as your examples show.
-Kevin




pgsql-hackers by date:

Previous
From: "Roberts, Jon"
Date:
Subject: Re: timestamp format bug
Next
From: "Kevin Grittner"
Date:
Subject: Re: timestamp format bug