No reply ..sigh!
There is some logic that is part of microflow within a process. I feel this ought to be implemented in stateless session beans. However, I would like to make a distinction between process logic versus logic that belongs in a domain model. Therefore, I would like my model classes to be functionally rich and containing all logic pertaining to domain model. This works great if you are handcoding all your model classes and then making them persistent by doing all that annotation business. However, if I have a process to auto-generate model classes from database tables, I would like subclasses to contain afore mentioned logic. In Hibernate, this was not an issue (I simply mention the subclass as the entity). If I use annotations, implementing this becomes more of a challenge.
Clarification on code generating base model classes:- In parallel development happening with multiple teams, often times, the database schema may evolve as the project progresses. In this case, it makes sense to have a single owner to manage base classes, while teams make contributions to the sub-classes.
However, if I use @EmbeddedSuperClass, the base class cannot be the target of a persistent relationship, right?