| Thread Previous • Date Previous • Date Next • Thread Next |
Hi Tim and all, I raised several bugs regarding the demographic model implementation. How does it work from now on? Who is responsible for confirming them, setting their importance etc? These is not clear for me. I committed several changes to the demographic model in the corresponding branch. I would like you to consider merging them to the trunk. Doing so I think that the majority of the bugs will be fixed. However I have some comments and doubts regarding the demographic model specification and the corresponding implementation: 1) I think that the reverse_Relationship_validity is not implementable at the current stage because it requires a demographic service that would return the respective partyRelationship from its partyRelationship reference. Analogous reasoning applies to source_Valid and Target_Valid in PartyRelationship 2) The function asString for PartyIdentity and Address is a bit strange for me. How do we know what it should return? The string representation for a party identity or an address should be specified in the archetypes, doesn't it? 3) The invariant Legal_identity_exists in Actor requires an identity with purpose "legal identity". I think this is misspecified for the following reason: the specification of what types of identity is allowed and what is required should be left out to the archetypes. In ISO22220 one of these types is legal name or registered name, not exactly legal identity. Even so it should not be mandatory to have one legal name. Imagine the situation where you have to register a newborn who has not a legal name yet. Even if things should be so, the way it is specified seems to be english oriented. It should use the ontology and use the local description of the term legal identity. In portuguese, for example, it should be something like "nome legal". While we don't settle this issue I preferred not to implement this invariant and returned true in the function hasLegaIdentity. Cheers, Sergio
| Thread Previous • Date Previous • Date Next • Thread Next |