Hi Dave!

If there are no duplicate records on the same date, the only problem I see
is that it could be a rather slow select.

I think

SELECT *
FROM CONTACTHISTORY A
WHERE NOT EXISTS (SELECT 1
FROM CONTACTHISTORY B
WHERE B.CONTACTTYPEID = A.CONTACTTYPEID
AND B.CONTACTDATE > A.CONTACTDATE)

could be considerably faster. Try for yourself. If it is a large table,
make sure you've got an index on (CONTACTTYPEID, CONTACTDATE DESC).

Set

>I forgot to say I'm using FB 1.0
>This seems to work...
>
>select * from contacthistory A where
> A.contactdate in (select max(contactdate) from contacthistory b where
>a.contacttypeid = b.contacttypeid)
>
>Can any one spot any problems?