The extension of data abstraction to database management

The long-term goal of the User Software Engineering (USE) project at the University of California, San Francisco, is to provide an integrated homogeneous programming environment for the design and development of interactive information systems. Realization of this goal involves the development of new software tools, their integration with existing tools, and the creation of an information system development methodology in which these tools are systematically used [1,2]. The successful construction of interactive information systems requires the utilization of principles of user-centered design [3,4,5], combined with features traditionally associated with the separate areas of programming languages, operating systems, and data base management [6]. It has become increasingly clear that the key to being able to provide such a unified view lies in providing a unified view of data [7]. The potential benefits of such a unification are considerable, including: 1) conceptual simplification of the system structure permitting, for example, joint design of data structures and data bases 2) the elimination of duplication or inconsistencies among diverse software components 3) the ability to achieve greater reliability in systems because of reduced dependence upon multiple software systems