The DAIDA project has made an attempt of formally defining and relating declarative and computational entities considered relevant to the process of database application development. Such entities range from object-oriented specifications to executable modules of database programs. To bridge the gap between semantics and computation, they also include abstract machine-based formal specifications and transformational theories. In a second contribution, selected characteristics of such entities and relationships are modeled uniformly in a software information system. Emphasis is placed on those properties that may become relevant when applications have to be modified or adjusted. Besides discussing the interaction of these aspects of the DAIDA methodology, the paper outlines an operational project prototype and reports first experiences.

Database applications must be capable of dynamically creating large quantities of names based on a scheme that can be communicated to and shared by some user community. The paper concentrates on this particular aspect of the set- and predicate-based Data Base Programming Language DBPL. It re-interprets the key and view mechanisms of relational database systems as naming schemes by introducing the concept of a selector for user-defined associative naming of partitions of large data sets. The DBPL approach results in a novel interpretation of relations as scopes with variable extents of named and typed data objects. The consequences of associative naming schemes on typing and binding mechanisms are discussed and various extensions required by data-intensive applications are introduced. The paper also describes the system support that the DBPL implementation provides for efficient name space management.

Practice has proven that databases are the keystones for nearly all application systems with a wider functionality, utilization and availabilty. As a consequence, next generation database systems will have to provide their services with a degree of interoperability that has to be substantially improved over existing solutions. In this paper we argue that this objective can be achieved only through full exploitation of current developments in computer language technology.

We claim that the merits of modern computer languages' naming schemes, typing systems and binding mechanisms are vital also for database management systems if they want to improve the quality of their interaction with application languages and programming environments as well as their own functionality and extensibility.

This paper studies the technological basis of modern computer languages and presents DBPL as a working example of a highly interoperable database programming language that exploits such technology.

Requirements of future data systems are discussed by emphasizing the abstraction principles considered helpful for the adequate design and organization of data-intensive applications and for the modularization, localization and, finally, the implementation of data-based systems. We conclude by relating the potential of advanced language technology to such specific demands of the next generation of post-relational data systems.

Key words: database programming languages, query languages, interoperability, open systems, language design principles, naming schemes, type systems, binding mechanisms, DBPL.