What do you not agree with? What I replied or with the way TABLEUPDATE() is working? This is the way it has been even before the CA. If you used SPT and issued UPDATE MYTABLE SET MYFIELD = 'myvalue' WHERE KEYID = 1 you would always get a successful return value as long as MYTABLE and MYFIELD existed even if there was no record where KEYID = 1.

>Mark,
>Sorry but I dont agree with this at all.
>
>In fact I see this is necessary if VFP generates the code and when we use DML code.
>The VFP generated code and most DML code doesnt provide us with a way to know if the
>code executed in the server afected an record, those codes are generaly simple Insert/Update/Delete
>statements and I like some kind of control about this situation.
>If we use SP in the server, the problem can be solved if we test @@Rowcount = 0 in the Stored Procedure, to return an error.
>I like to know if the record was afected and if not, I give the user an information about.
>
>Reading the url provide by Jim Nelson, it shows how to handle it when VFP generates the code.
>
>Peter