My book says:
When a distributed framework is used, a client makes a call to what appears to be the interface of a business object. What it actually calls, however, is a stub that mimics the interface of that business object. This layer between clients and business objects is added because it is more practical to place stubs in the remote and distributed locations of clients than to place complete copies in the location of business objects.
In a distributed framework, the client calls a business method on a stub as if it were the real object. The stub then communicates this request to a tie. The tie calls the method on the real business object. A result is returned to the stub and the client.
I don't understand how this distributed framework concept relates to EJBs and the
J2EE architecture in the 'regular' case in which the client tier presents HTML pages.
I can understand the distributed framework concept when the client tier is an application, but I feel that this is the exception.
So, over all, when the client tier is HTML, I think the EJB framework is just overly complex.
Any thoughts?
Thanks,
Dan