> Agreed. But when Tom pointed out the problem in your query you were
> quite sure you couldn't be wrong. When I asked you to run explain to
> see what kind of row estimate you got, I got no answer. This was a
> query problem not a hardware tuning problem.
>
> The best step for getting a good answer is forming a good question.
>
> So, has changing the query made it faster? Did explain show what I
> expect, a huge cartesian product? Just guessing.
>
Yes changing the query fixed the problem. When I replied to Tom I
was under the impression that my redaction of the query was causing a
confusion.
In my defense the query I posted works fine in mssql server and mysql.
I just presumed it would work the same in postgres. I guess that was
my mistake.