How do you generate your primary key? I use a surrogate primary key which use a reference table for the next key # when adding a new record. That way, I never get uniqueness problem when adding a record.

What do you mean by "it screws up all the connections to forms"?

Some folks use a primary key with a FOR NOT Deleted() clause to avoid this problem. If you go for this method, you will need to change the key field before recalling a deleted record. Also, Rushmore don't use filtered indexes, so you will need a separate index on you primary field.

HTH

>I have a volatile data set and need to delete records quite often. I then replace them with new data and get violations of uniqueness on various indexes. I don't want to pack too often because it screws up all the connections to forms. So my question is whether there is a way to get the automatically generated RI code to honor the deleted records when testing for uniqueness? I'm also interested in other approaches people might use. TIA, Bob Alei