>Just curious what folks think out there regarding what's the best way to generate key field values when adding new records to tables.
>
>When my users click the ADD button, I've been running code to use the table in the key field order, go bottom, find the last value and add 1 to it for the new record's key field value.
>
>It seems to be working fine, though I realized I had to play with my SET DELETED ON/OFF commands so my next value was unique and not the same as a deleted record's.
>
>I know some folks keep separate tables to track the next available key value for their tables. I'm assuming it looks like a 2 field table, one for the table's name, one for the next value. You'd then check there for the next value you want to assign.
>
>Any way better than any other??

In order to avoid problems with DELETED key values, I have been using expressiones like this for my primary indices:

IIF(DELETED(), -RECNO(), <>)

(Of course, my primary key field is never negative.)

However, this solution isn´t very good for it hampers Rushmore, so I add a *regular* index on <> just for SQL and Rushmore.

HTH!