Re: Problems with substring

Ducan,

> I am using postgres 7.1.3
> on redhat 7.2.

First, please consider upgrading.   7.2.1 has a lot of enhancements
over 7.1.3, and we're testing 7.3.0 now!

> drop view wm_ups_count;
> create view wm_ups_count as
> select
>  substring(phase from 1 for 1) as ups,
>  sys_key,
>  count(substring(phase from 1 for 1)) as phas_count
> from
>  power
> where
>  sys_key = 10
> group by
>  sys_key,
>  phase;
^^^^^^^^^^^^^

This is your problem.  It should be:
GROUP BY sys_key, substring(phase from 1 for 1);

If you GROUP BY something other than the columns you are displaying,
you will often see "duplicate" rows.

-Josh Berkus

______AGLIO DATABASE SOLUTIONS___________________________