Relational schema evolution for program independency

The database schema is assumed to be stable enough to remain valid even as the modeled environment changes. However, in practice, data models are not nearly as stable as commonly assumed by the database designers. Even though a rich set of schema change operations is provided in current database systems, the users suffer from the problem that schema change usually impacts existing application programs that have been written against the schema. In this paper, we are exploring the possible solutions to overcome this problem of impacts on the application programs. We believe that for continued support of the existing programs on the old schema, the old schema should continue to allow updates and queries, as before. Furthermore, its associated data has to be kept up-to-date. We call this the program independency property of schema change tools. For this property, we devise so-called program independency schema evolution (PISE) methodology. For each of the set of schema change operations in the relational schemas, the overview of the additional code blocks due to the PISE compliance is presented in order to prove the comprehensiveness and soundness of our PISE methodology.