As systems consume CVIs from different sources there can be certificate number collisions. For these to be treated as unique identifiers, receiving systems need a way to identify the source of the Certificate Number. When these are received directly from the source system, this can be inferred from the source, but if the data are stored, forwarded, etc. the identity of the generating system is lost.
I would suggest adding a Certificate Number Namespace attribute. This would ideally be a universally unique ID such as an OID, or we could just have an agreed upon list that we maintain.
Alternatively, if we could get consensus on a prefix pattern say paper always start with State code, mCVI start with mcvi, eCVI always start with ecvi. This would mean all electronic CVI vendors update their numbering schemes. I don't like this much. Just a thought.
As systems consume CVIs from different sources there can be certificate number collisions. For these to be treated as unique identifiers, receiving systems need a way to identify the source of the Certificate Number. When these are received directly from the source system, this can be inferred from the source, but if the data are stored, forwarded, etc. the identity of the generating system is lost.
I would suggest adding a Certificate Number Namespace attribute. This would ideally be a universally unique ID such as an OID, or we could just have an agreed upon list that we maintain.
Alternatively, if we could get consensus on a prefix pattern say paper always start with State code, mCVI start with mcvi, eCVI always start with ecvi. This would mean all electronic CVI vendors update their numbering schemes. I don't like this much. Just a thought.