>Louis,
>
>While your answer was helpfull and fairly accurate, you did have one major inaccuracy in it. Both pessimistic and optimisitc prevent two users from changing the same record at the same time. They do their checking at different points in the process but neither will allow a logical data collision to occur.

You are right, I did not explain myself appropriate. What I meant was :
When you use optimistic buffering, "two users can change the
same record", but when you do TABLEUPDATE() for table buffering or
try to move the pointer for row buffering, only one will be successfull.
Please, Correct me if I am wrong.

Pardon me If this caused somebody confusion.