Luis,
TABLEUPDATE(.T.) will update all records changed and since you are table buffering, I assume the record pointer moves to check for all the records to change. So, IOW, how does it know what's the last record to update unless all have been checked?

If you pass it 0, then the flag is to just update the current record. In which case the record pointer shouldn't have to move.

As for the diff between 0, 1 as the first parameter and passing 2 is whether or not an error array is created with the errors or whether an error in the update occured generates an error. Note that this doesn't necessarily mean that the tableupdate was successful.

This is my understanding of the sit. HTH.

>I am updating all records with TABLEUPDATE(.T.). this code was originally
>in 3.0b,now is running in 5.0a, as you know, the sintax for TABLEUPDATE changed in 5.0, but it recognizes TABLEUPDATE(.T.) from 3.0b.
>I guess that is the way it works and it seems to me that the
>explanation in the help file if not very clear.
>
>It says :
>"TABLEUPDATE( ) returns false (.F.). If you specify 0 or 1 for nRow, the record pointer remains on the record where changes could not be committed and can issue AERROR( ) to determine why the changes could not be commited."
>
>It doesn't say that the pointer remains in the last record updated,
>see what I mean ?