Re: performance on select query

[lucy smiles]

out put of explain


table | type| possible | keys | key | key_len| ref | rows|Extra

tbl1 | range | PRIMARY | id | NewIndex | 8 | NULL | 147491 | Using where; Using filesort

tbl2 |ref |tbl2id |tbl2id |5 | tbl1.id | 1 | Using where

i understand when doing a blank search it has to do a table scan and then sort it alphabetically. Is there any other way where the result can be sorted alphabetically with less time and then limit the result set to 10/20 records?

at the moment we are looking up with alphabet A, and this seems to be one that doesn't have any answers?

thanks
lucy