revise the "rollback()" part of the above message.  after some thought
im less sure if i want to go with it; it would involve keeping track
of two separate copies of each object's attributes; one, what was
loaded from the database at the very beginning, or after the last
commit(), and the second, whatever was present before each flush()
operation (since the subsequent flush() uses that difference to
calculate).  also if the session is wrapped in an enclosing
transaction, its still not going to really "rollback" all the way.
The alternative way to "rollback()" is to just expire() every
instance...which means every instance incurs a load operation when
next accessed...still kind of expensive.  not to mention im concerned
about packing in too many new things into one release; each new
feature is going to require documentation and support going out the
door.


--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"sqlalchemy" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at 
http://groups.google.com/group/sqlalchemy?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to