Michael Avila wrote:
> Interesting. I think I understand that. I have never worked with a SELECT
> within a SELECT (I think that is called a subquery). I am guessing that it
> works its way through the member status records until the latest date
> "floats" to the top (nothing is > than it).
>
> Will that be a problem performance-wise if there are thousands of records?
The actual execution of the query may be something completely different,
depending on what the optimizer decides. I doubt it will suck much,
because the optimizer is quite good. Try executing
EXPLAIN query;
to see how it would be executed, and
EXPLAIN ANALYZE query;
to see the above plus the real numbers that the executor got after
executing it (number of times each loop was executed, time spent at each
execution step, etc).
Read the ANALYZE page to see how are these things supposed to be read.
--
Alvaro Herrera http://www.CommandPrompt.com/
The PostgreSQL Company - Command Prompt, Inc.