OK, thanks. That seems to make sense.
regards
Iain
- Original Message -
From: "Tom Lane" <[EMAIL PROTECTED]>
To: "Iain" <[EMAIL PROTECTED]>
Cc: <[EMAIL PROTECTED]>
Sent: Thursday, December 02, 2004 11:54 AM
Subject: Re:
"Iain" <[EMAIL PROTECTED]> writes:
> Just a quick question out of curiosity, I was just wondering if this is
> supposed to be valid sql:
> select count(*) as cnt
> from sometable
> group by somecolumn
> having cnt > 1
No. The HAVING clause logically executes before the SELECT output list
does,
Hi,
Just a quick question out of curiosity, I was just wondering if this is
supposed to be valid sql:
select count(*) as cnt
from sometable
group by somecolumn
having cnt > 1
This isn't valid in pg (7.4.6), but this is:
select count(*)
from sometable
g
