Michael Hunger wrote: > Where do yo get the server from? I didn't mention it and the others not > either.
Sorry, I just read an interview with Gosling about service vs server (interface vs implementation), which is why. I should have explained that. > Sure the running implementation of the service is not needed to be locally > available, there was never a question > regarding this. Well, actually there is. If I understood Niclas correctly, the implementation of a Service should always be a ServiceComposite. For local services this is not a problem, but if we want the ServiceComposite to live in a remote VM, and then be imported and invoced locally, then the Service mechanism MUST allow objects to not be ServiceComposites. > Why must we ensure that the injected Service may be no ServiceComposite? If > it is a ServiceComposite with a remote proxy > Mixin in the current Module configuration this is perfectly ok, the qi4j > application on the server then has the Mixin > which actually implements the service and exports it that the proxy can > access it ? > > Or do I get sth wrong? The main question is: must the Service mechanism depend on the implementation to be a ServiceComposite? If so, how do we deal with remote services? /Rickard _______________________________________________ qi4j-dev mailing list [email protected] http://lists.ops4j.org/mailman/listinfo/qi4j-dev
