> Which to me leaves the question: is our usage described at [1] an > intended use case, that's here to stay? or simply a "fortunate > unspecified behavior", and we should change our code to this new > append() feature?
For now, I'm going to call it a fortunate unspecified behavior. At some point in the future when I can get time allocated to add robust polymorphism support to RequestFactory, that behavior might go away if it helps promote uniformity between proxy->entity and context->code type hierarchy mappings. -- Bob Vawter Google Web Toolkit Team -- http://groups.google.com/group/Google-Web-Toolkit-Contributors
