Re: [hibernate-dev] session-per-application in a rich client an antipattern?