Supplementing the Well-Known Factory Pattern for Distributed Object-Oriented Systems

In distributed object-oriented systems where resource users and resource providers/managers live on different nodes in the network it is common design practice to implement the well-known Factory Pattern for enabling the client part of an application to create new instances of classes or components on the server side at runtime in order to use these objects subsequently. In this paper, we describe variants of and supplements for the Factory Pattern which address some missing, but very important aspects of the problem domain at hand, especially concerning the (automatic) destruction of remote objects. We discuss advantages and disadvantages of the different solutions and finally present a flexible compromise. Some thoughts about the implementation effort for this enhanced variant of the Factory Pattern and a short summary will conclude the paper.