Hi Juan......

>Evan wrote very interesting data!

Yes he did.

>
>I am developing an accounting system, its design is simple (something like clients and orders). I handle accounts and items.
>
>I am testing my App on a 16 MB, 166 Mhz machine. The items.dbf file has 100.000 records and i think this is a small table on this machine. Am i right?
>
>My App doesn't select records matching a key very fast yet, because i use to build compound keys and i am not building optimizable expressions, but i will, of course.
>
>Also, there are different methods to process a set of records:
>
>Scan all for ...
>Select .... where ...
>Set Key Range exp1, Exp2 ...
>
>
>Any advice ?


The only general advice I can give is to avoid "FOR" expressions where Rushmore can't optimize the expression. As to the rest, I guess each situation will dictate different circumstances.