9222926 Hull Towards a Toolkit for Active Databases This is the first year funding of a three-year continuing award. This is an interinstitutional collaborative research. Design, implementation and experimentation activities are conducted at both sites. Active databases present a novel and promising paradigm for programming databases. The Heraclitus project supported by this grant is focused on the development and application of database programming language constructs and techniques that can be used to specify and implement active databases. A central thesis of the project is to elevate "deltas", i.e., sets of proposed modifications to the database state, to be first-class citizens in database programming languages (DBPLs). While deltas have been used before in limited contexts and ad hoc ways, the Heraclitus languages permit programmers and users to freely create, manipulate and apply deltas. This offers the potential of developing novel mechanisms for supporting a variety of applications including active databases, database interoperation, hypothetical database access, concurrency control protocols, backup and recovery protocols, and resolving conflicts between updates. The research is centered around the development of a DBPL called Heraclitus Alg,C , an extension of C that incorporates the relational algebra and deltas, and persistence for relations and deltas. A variety of topics are being studied in connection with Heraclitus Alg,C , including language design, optimizations, parallel implementations, theoretical issues, and applications to active databases and database interoperation. The research will yield fundamental insights into active database technology, and also methodologies for using this technology so that diverse databases connected by the "information highway" can be used in a coherent manner. ***
