>1) You don't want to use indexes in your select (errors in the index may give 
>wrong results), so try:
>           select
>             T.F1+0,
>             T.F2+0,
>             count(T.ID+0)
>           from
>             T
>           group by
>             1, 2
>           having
>             count(T.ID+0) > 1

Oops, I forgot index problems could lead to there being NULL or duplicates in 
the PK field, so change the above query to:

           select
             T.F1+0,
             T.F2+0,
             count(*)
           from
             T
           group by
             1, 2
           having
             count(*) > 1

Set

Reply via email to