Re: big distinct clause vs. group by - Mailing list pgsql-performance

From Claudio Freire
Subject Re: big distinct clause vs. group by
Date
Msg-id BANLkTikOUovTjsK8_9htTjrocHJxrGUv2g@mail.gmail.com
Whole thread Raw
In response to Re: big distinct clause vs. group by  (Robert Klemme <shortcutter@googlemail.com>)
List pgsql-performance
On Tue, Apr 19, 2011 at 11:07 AM, Robert Klemme
<shortcutter@googlemail.com> wrote:
> I find that slightly contradictory: either you do care about the
> values then your business requirements dictate the aggregate function.
>  If you only want to pick any value actually in the table but do not
> care about which one (e.g. MIN or MAX or any other) then you don't
> actually care about the value.  Because "SELECT a, MAX(b) ... GROUP BY
> a" and "SELECT a, MIN(b) ... GROUP BY a" are not equivalent.  And, if
> you do not care then there is probably no point in selecting them at
> all.  At best you could use a constant for any legal value then.

I know it sounds weird, but there are at times when you only want one
of the actual values - but don't care which one precisely.

It happened to me at least once.

So, it may sound as nonsense, but it is probably not. Just uncommon.

pgsql-performance by date:

Previous
From: Laurent Laborde
Date:
Subject: postgresql random io test with 2 SSD Kingston V+100 500GB in (software) Raid1
Next
From: Laurent Laborde
Date:
Subject: Re: postgresql random io test with 2 SSD Kingston V+100 500GB in (software) Raid1