Marc, that's a technology answer. It's a good answer in that respect, but
you haven't addressed the underlying question, which is more relational that
EJB oriented.
The question is actually one of persistent object identifiers, also known as
surrogate keys.
This argument comes and goes with tedious regularity on all platforms with
all technologies. Comes and goes is an understatement. It's more like Night
of the Living Dead. It seems to me that we have a good opportunity to cut
off its head, fill its mouth with garlic, drive a stake through its heart
and burn it. I would like it to _stay_ dead.
Why don't we begin by scoping the problem broadly and then defining the
semantics of surrogate key operations in a distributed environment.
Then we can devise a model and interfaces that have consistent semantics in
both distributed and standalone environments.
I feel that this should be handled by the container.
Really it should be defined in the J2EE spec, but if we do a good job it
probably will.
Your thoughts, gentlemen?
Peter Routtier-Wone
---------------------------------------
One day, all this too shall parse.
_______________________________________________
JBoss-user mailing list
[EMAIL PROTECTED]
http://lists.sourceforge.net/lists/listinfo/jboss-user