An architecture for evolutionary database system design

We outline herein an approach to evolutionary database systems design and evolutionary database design which is intended to lower the total costs of maintaining, developing and executing application programs against a database. The premise is that it should be possible to have an evolutionary database system which executes efficiently while providing explicit support for changes in application usage. The key concepts are: hierarchical layering of a database specification, inclusion of static/dynamic characteristics in specification of the database, and a hybrid compile/interpret system for the execution phase of the data management system itself.