getting rows you may not use

[gerardg]

I'm thinking about splitting one query into a count query, and a query that gets everything else. The original query is ugly - temporary tables, filesorts, ...

First I'd get the counts, then do:

while ($result_count <= $needed_results && $row = mysql_fetch_object($result)) {
if($row->Max > $prev_counts_array($row->ID)) {
// do stuff with row
$result_count++;
}
}

So, say the second query would retrieve 100 rows, but you only need 4. Do the other 96 records use resources - or only when they're fetched?

[van__]

Well being that a query result IS a resource, yes... they do take up resources in the fact that the query resource (RAM troll) is bigger and fatter with each row it contains in it's greedy lil paws.

Of course, 100 rows of data isn't going to hurt you. But, when you're dealing with 1000+ rows that aren't needed, ya that could slow you down. But 100? Let the RAM troll be fat and happy.

Of course, it's generally best to query as verbosely as possible to ensure the fewest rows returned.