You could change them by multiplying the current value by -1 to ensure no duplicates there. However, you can only do this once per re-used ID number. Now you have to have another tracking/counting scheme to keep up with these meaningless values.

I think the best solution is just to undelete the found value and blank out the current data.

>You could change the key field of the record immediately before deleting
>REPLACE keyfield WITH a_meaningless_value
>DELETE
>That way your deleted records won't show when you use keymatch()
>
>
>>Hi
>>
>>Thanks for your reply
>>
>>I am having difficulty with duplicate records probably for the very reason that deleted records are not removed from the index.
>>
>>Does this mean that if I want to reuse a key tag that I have to :-
>>
>>1) look for the tag
>>2) then check to see if it was deleted and
>>3) if so then recall it and replace the details.
>>
>>Is there not a way to force removal of a deleted record without having exclusive rights???
>>
>>Regards
>>Geoff Scott