GH-33 First version of port of AbstractEntity from DDDBITS

[hschwentner]

A base class for entities with convenience implementations for equals(), hashCode(), toString() based on the entities identity.

Belongs to issue #33

Questions:

• is AbstractEntity the right name? Or would it be merged with interface Entity?

[odrotbohm]

I'm not sure I like adding implementation classes to jMolecules, even if abstract ones. All these base class are likely to encode some assumptions that are highly likely to be invalidated by a concrete project's requirements. Just the decision whether you'd want to allow an entity to live with a null identifier for a bit of time is taken away here.

Lifting those support classes up jMolecules rubber stamps them as the way to implement those interfaces when I'd argue there is no such one way. Is there anything wrong to keep those implementation helper libraries around and just move them to use jMolecules interfaces and annotations as needed?

Kind of what I already mentioned here.

[hschwentner]

I'm not sure myself, if jMolecules is the right place for that. Just wanted to open the discussion again. I agree, there's not the way to implement it right and we don't want to pollute jMolecules with specific implementations. On the other hand every project that does DDD tactical design has to deal with implementing the identity. So every project that uses jMolecules has to deal with that.

Hm, maybe we should talk in person again :-)