Derived Types and Subschemas: Towards Better Support for Logical Data Independence in Object-Oriented Data Models

Two concepts for the support of logical data independence in the context of the structurally object-oriented data model NO2 (New Object-Oriented data model) are presented: derived types and subschemas. In general, object types can be derived in two ways: derived types can restrict (the value set of) an existing base type, or new atrtibutes can be added to the tuple value set of an existing object type. One major design issue of the concept is to preserve object identity: an instance of a derived type has the same object identity as the corresponding instance of the base type. This enables us to permit updates to derived types even in cases where (say) the relational view mechanism could not permit them. The second concept, subschemas, supports application-specific views of schemas and the straightforward identification of those parts affected by schema modifications.