On 2/6/12 10:22 AM, TXVanguard wrote:
Rick Hillegas-3 wrote:
This means that more than 1 row is returned by the subquery. That, in
turn, suggests that the query, which succeeds for you on another
database, may not be behaving in a predictable way there. Since more
than one row qualifies to drive the update, which row do you want? Is
there some other restriction which you can put in the subquery to
guarantee that you get one, predictable row?
Hope this helps,
-Rick
I wonder why this doesn't cause a problem for Access?
The Access update could change the target row twice. No one will notice
if both matching rows from T1 have the same value in column B. Or...not
many people will notice if the rows from the join always come back in
the same order so that the same value of B always wins. It might be
useful to run the join outside the update and see what it returns on
Access and Derby.
Hope this helps,
-Rick
