"Joshua D. Drake" writes:
> This looks like a NULL vs '' issue. Am I wrong?
No, it's a NULL vs NOT IN issue. Specifically, if the subquery yields
any NULLs and the comparison operator is strict (which nearly all are)
then it's impossible to get a TRUE result from the NOT IN --- the only
possibil
>>> "Hoover, Jeffrey" wrote:
> In Sybase?
Sybase ASE defaults to the same behavior as Microsoft SQL Server, but
they have a configuration option to yield standards compliant behavior
in this regard (SET ANSINULL ON).
-Kevin
--
Sent via pgsql-admin mailing list (pgsql-admin@postgresql.org)
T
wanted. :)
-Original Message-
From: Joshua D. Drake [mailto:j...@commandprompt.com]
Sent: Wednesday, January 28, 2009 12:47 PM
To: Scott Whitney
Cc: 'Hoover, Jeffrey'; 'Kevin Grittner'; pgsql-admin@postgresql.org
Subject: Re: [ADMIN] Strange query problem...
On Wed, 2009
a2;
> 2|two
> 4|four
> 0|
>
> sqlite> select * from a1 where i not in (select i from a2);
> one
> three
> five
> sqlite>
>
> -Original Message-
> From: Scott Whitney [mailto:swhit...@journyx.com]
> Sent: Wednesday, January 28, 2009 1:22 PM
day, January 28, 2009 12:31 PM
To: Scott Whitney; Kevin Grittner; pgsql-admin@postgresql.org
Subject: RE: [ADMIN] Strange query problem...
How do other databases handle this? I tried it in SQLite and I get
different behavior (see below). Can someone try it in Oracle? In MySQL?
In Sybase
So, you're sayin' I ain't crazy? :)
-Original Message-
From: Hoover, Jeffrey [mailto:jhoo...@jcvi.org]
Sent: Wednesday, January 28, 2009 12:18 PM
To: Kevin Grittner; Scott Whitney; pgsql-admin@postgresql.org
Subject: RE: [ADMIN] Strange query problem...
Wow! I would neve
>>> "Scott Whitney" wrote:
> Um. How is this possible?
> mydb=# select * from time_recs where id_time_rec not in (select
> id_time_rec from punch_time_recs);
> (0 rows)
>Table "public.punch_time_recs"
> Column | Type | Modifier
