Justin, please use the jboss-user mailing list No our optimization doesn't violate this condition and enforces pass by copy on the optimization. If you want the raw speed (and many people want it) you can turn OFF the pass by copy thing so that you pass the native pointer. Yes it is non standard, but for those that know what they are doing and since 95% of the EJB calls will be in VM this is a huge improvement. The optimization was introduced in jboss 0.9 the ON-OFF feature for copy introduced in jBoss 2.0. It should really be a faq, can I ask you to try out the "faq o matic" from Micheal J? regards marc Justin Lee wrote: > > I have a question that's been nagging me for a few days and thought I'd > ask. I'm new to the list so please forgive if this is a repeat. In an > earlier message someone posted that they had submitted a patch that > would detect if the reference was local to the VM and would use the > local interface instead of the Remote interface. ( This is a rough > paraphrase. Sorry. ) Perhaps I'm misunderstanding, but here's my > question. According to the 1.1 spec, the container must ensure that no > references to beans within the container use the "native" java > interfaces, but go through the EJB architecture to obtain and use > references to other beans as if they were remote clients as well. So my > question is, does this patch violate that constraint? I'm fairly new to > EJB as well, so any clarification would help. Thanks. > > -- > Justin Lee | The land is never really owned until the deed is paid. > JEDI | And that is why the power lies wherein His body laid. > > -- > -------------------------------------------------------------- > To subscribe: [EMAIL PROTECTED] > To unsubscribe: [EMAIL PROTECTED] > Problems?: [EMAIL PROTECTED] -- -------------------------------------------------------------- To subscribe: [EMAIL PROTECTED] To unsubscribe: [EMAIL PROTECTED] Problems?: [EMAIL PROTECTED]
