Shianmiin wrote:
>
> I thought the Atomic should be at statement level, could anyone tell me
> why PostgreSQL behaves differently?
>
re-state for clarification purpose.
Since SQL is a set language, there is no concept of row order. I thought the
checking should be on a per set operation (i.e. per statement), not per
micro operation basis (e.g. in this case, the order of operations could
result in success/failure). I don't know how strict the standard defines
here and it could be implementation specific. Still wondering why PostgreSQL
does it differently from other major DBMS?
--
View this message in context:
http://postgresql.1045698.n5.nabble.com/Another-unexpected-behaviour-tp4610242p4612763.html
Sent from the PostgreSQL - general mailing list archive at Nabble.com.