Closed benedictws closed 3 years ago
The duty to evidence controls looks so:
:O3 a odrl:Duty ;
nl:creditor <https://permid.org/1-4295899615> ; # CME
odrl:action [ a md:Evidence ;
md:actionScope md:Controls ;
odrl:count "1"^^xsd:int
] ;
odrl:duty :D10 .
Where the duty :D10 is the activating request:
:D10 a odrl:Duty ;
nl:debtor <https://permid.org/1-4295899615> ; # CME
odrl:action [ a md:Notify ;
odrl:count "1"^^xsd:int
] .
Above I'm using the action "evidence" - i.e. provide evidence of the controls applied. This wording can be seen in the CME license. But is this really different to "report" - i.e. report on the controls applied?
We already have the concept "report" - see Issue 7 . Do we need this additional concept: "evidence"? Is there a useful distinction here to be made between the two?
Sometimes a party to an agreement must provide information about their activities to the other party. The example that comes up in market data licenses is evidence of controls - i.e. what access control strategies are being applied to the data.
The difficulty in modeling this is that this duty always holds - it can never be fulfilled and finished with. The assigner can at any moment request this evidence of the assignee. Fulfilling the request does not discharge the duty; the assigner can always ask again.
So it's probably best modeled as a duty that is activated by a notification. The assigner can notify the assignee as often as they wish, and the assigner must provide the evidence every time. The assigner's duty is activated by the request.
This is a pattern we see in consent duties (requires notification) and payment duties (requires invoicing).