The use of RDA allows for the server objects to connect to the database, no matter on what machine the database is located.
We now have two levels of independence in the demonstration project. The server objects do not depend on the location of the database, and the database access procedures. The interface to the database handles all necessary connections, such as locating the database on the network, and supplying commands and returning data.
The other level of independence is client to server. The server objects can be running on any machine on the network (presuming the existence of an object request broker). Client programs send messages to the server objects, and the underlying object request broker (ORB) delivers the messages to the appropriate server implementation. Furthermore, if there is no server implementation running, the ORB can be directed to start one. To the client, the only visible interface is that of method calls. There is no need for the client to know the machine name, port numbers, or other communication level information. In this way, the methods are more than remote procedure calls, and provide a higher abstraction for the clients.