Walter -
Can you clarify the scenario a little? Are you seeing these anomalies when Entities are created on the RCP side, then streamed to the server to persist? You mentioned that you aren't able to recreate this behavior in your JUnit integration testing... are you serializing / deserializing your Entities in those tests? That would recreate streaming them over the wire... which might be where we're having problems.? That being said, nothing is coming immediately to mind as to what the issue might be. Thanks, Rick On Tue, Nov 15, 2011 at 5:55 AM, M. Walter <marc.wal...@sbb.ch> wrote: > Hi Rick, > > we found out another interesting thing. Under some conditions (i.e. using > entities on RCP client side) saving the "right" entity matters. We called > persist on the inverse side entity of the ManyToMany relationship which > lead > to only one UPDATE statement for the inverse entity. The expected INSERT > statement on the join table wasn't executed. As we switched the owning and > inverse side everything worked as expected. Two UPDATE and one INSERT > statements were executed. > > But again there is something strange: for the JUnit integration test this > does NOT matter apparently. Calling persist on the inverse side of the > relationship executed all necessary statements in order to create the new > ManyToMany relationship. Do you have an explanation for this behaviour? > > -- > View this message in context: > http://openjpa.208410.n2.nabble.com/Strange-behaviour-on-ManyToMany-relationship-using-RCP-client-tp6985422p6996020.html > Sent from the OpenJPA Users mailing list archive at Nabble.com. > -- *Rick Curtis*
