| Abstract: | Name management -- how a computing system allows names to be established for objects, permits objects to be accessed using names, and controls the availability and meaning of names at any point in time -- is a major challenge for developers and users of any complex software system. Experience indicates that name management is particularly problematic in convergent computing systems. These are systems that combine elements from several disparate computing domains into a single, synergistic whole; object-oriented databases, world-wide-web applications and image understanding environments are all examples of this increasingly important class of systems. Despite its importance, historically name management has received relatively little attention from researchers in computer science. As a result, computing systems typically rely on ad hoc and primitive approaches to name management, making systems confusing to use and difficult to maintain. The results of this dissertation include the development of novel modeling techniques and mechanisms with both theoretical and practical implications for name management in convergent computing systems. In particular, some broadly applicable models are defined to help reason about existing approaches to name management, especially their context formation and control capabilities. Example applications of the models illustrate how they can contribute to improving our overall understanding of the use of names, as well as some of the difficulties often encountered, in various approaches. Using the models as a basis, prototype mechanisms are then developed, implemented and evaluated for a representative instance of a convergent computing system. Specifically, a consistent and uniform name management mechanism is defined and demonstrated for both C++ and CLOS application programming interfaces to an object-oriented database (OODB) system. The resulting mechanisms represent a substantial improvement in the area of name management in this domain of convergent computing systems. Finally, based on these models and mechanisms, a new, seamless approach to achieving polylingual interoperability is defined and demonstrated. This approach permits applications to uniformly process data objects from an OODB, despite the fact that those objects were originally defined, created and persistently stored using the capabilities provided by several distinct programming languages. |
