Integrating programming languages and databases: What is the problem

Abstract

The problem of integrating databases and programming lan- guages has been open for nearly 45 years. During this time much progress has been made, in exploring specialized database programming languages, orthogonal persistence, object-oriented databases, transaction models, data access libraries, embedded queries, and object-relational mapping. While new solutions are proposed every year, none has yet proven fully satisfactory. One explanation for this situation is that the problem itself is not suciently well dened, so that partial solutions continue to be proposed and evaluated based upon incomplete metrics, making directed progress dicult. This paper is an attempt to clarify the problem, rather than propose a new solution.We review issues that arise on the boundary between programming languages and databases, including typing, opti- mization, and reuse. We develop specic criteria for evaluating solutions and apply these to the solution approaches mentioned above. The analy- sis shows that progress has been made, yet the key problem of meeting all the criteria simultaneously remains open.