The proposed system of stereotypes is confusing. It would be clearer if the concept of non-Entity (versioned-)identifiables would be dropped. The names of the remaining stereotypes could be shortened:
D2Class --> D2Entity --> D2Identifiable --> D2VersionedIdentifiable
Note that this is not an inheritance relationship. In order to respect the Liskov Substitution Principle, D2Class and D2Entitiy would both be specialised from UML Class; D2Entity --> D2Identifiable --> D2VersionedIdentifiable could in principle be realised as specialisation.
The definitions of D2Class and D2Entitiy could be improved:
D2Class – The class represents a complex data structure. Instances have no concept of identity and are not expected to be entry points in data sets.
D2Entity – The class represents a kind of domain object with a conceptual identity or an object which may be the entry point in a data set, because it contains enough information to be useful by itself in some contexts.
iancornwell1
commented
1 month ago
Bug description?
The proposed system of stereotypes is confusing. It would be clearer if the concept of non-Entity (versioned-)identifiables would be dropped. The names of the remaining stereotypes could be shortened: D2Class --> D2Entity --> D2Identifiable --> D2VersionedIdentifiable Note that this is not an inheritance relationship. In order to respect the Liskov Substitution Principle, D2Class and D2Entitiy would both be specialised from UML Class; D2Entity --> D2Identifiable --> D2VersionedIdentifiable could in principle be realised as specialisation. The definitions of D2Class and D2Entitiy could be improved: D2Class – The class represents a complex data structure. Instances have no concept of identity and are not expected to be entry points in data sets. D2Entity – The class represents a kind of domain object with a conceptual identity or an object which may be the entry point in a data set, because it contains enough information to be useful by itself in some contexts.
Version
4