Please tell me more about design and coding benefits of using ODMG vs PB directly. From the tutorials, you still have to get a broker, ask the broker to begin transaction, perform the operation (insert, update or delete) and commit or rollback the transaction in the UseCase classes.
I know ODMG provides locking. I'm using PostgreSQL with optimistic locking, so I would think PostgreSQL should handling concurrency conflicts and my commit would fail, then I would retry with a refreshed transaction. Does ODMG offers automatic retry?
I have used object databases with OQL capabilities before, and to me OQL not as robust as SQL in constructing complicated queries. Also in the tutorials, QBE seems more elegant than OQL since we are passing objects in QBE and building strings in OQL. Please enlighten me. I may have developed a brain leak in my old age.
Thanks! Tina
--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
