RE: of Duties

Dear Joshua,
I think a realistic obligation is one that is activated by:

  *   The satisfaction of a state of affairs: in ODRL this can be expressed with a Constraint (e.g. in the library you have to speak softly);
  *   The occurrence of an event or action (e.g. I enter the car park then I have to pay before leaving).

The first one can be represented in ODRL 2.2 with a Spatial Constraint.
The second with the time sequence of actions and the payment deadline I am afraid cannot be represented in the current version of ODRL.

In my opinion, the relationship between the obligation to do an action and the requirement that this action be permitted is at another level.
Best Regards
Nicoletta Fornara


From: Joshua Cornejo <josh@marketdata.md>
Sent: 22 July 2024 14:49
To: public-odrl@w3.org
Subject: Re: of Duties

Hola Victor,

There are potentially good use cases, just working on the chain of responsibilities.

As I am working on the evaluation, there is also the challenge for three types of users:

  *   those building policies (‘What does it mean an obligation with X action as I am ‘testing’ in my sand box?’),
  *   those evaluating the policies in production (‘What happens when P tries to execute action X that only is specified as an obligation?),
  *   those responsible for any ‘post mortem’ once a policy has been finalised (‘what actions were executed successfully, which remain outstanding, etc?).

Being easily extensible also creates downstream complexity, so having good guard rails that minimise misinterpretation is (sort of) the reason of my enquiry.

Regards,
___________________________________
Joshua Cornejo
marketdata<https://www.marketdata.md/>
embed open standards
across your supply chain

From: Víctor Rodríguez Doncel <vrodriguez@fi.upm.es<mailto:vrodriguez@fi.upm.es>>
Date: Monday 22 July 2024 at 13:30
To: <public-odrl@w3.org<mailto:public-odrl@w3.org>>
Subject: Re: of Duties
Resent-From: <public-odrl@w3.org<mailto:public-odrl@w3.org>>
Resent-Date: Mon, 22 Jul 2024 12:30:24 +0000

Hi Joshua,

In my opinion, having obligations without them being triggered by permissions makes the language more flexible and suitable for non-foreseen scenarios.

Regards,
Víctor
On 19/07/2024 16:27, Joshua Cornejo wrote:
Hello,

I am reading 2.6.4 Obligation property with a Policy<2.6.4Obligation%20property%20with%20a%20Policy> and trying to understand in which use case a Party will have to fulfil an obligation without it being triggered by odrl:Permission or by a odrl:consequence.

Example 20 implies that assignee person:44 has to pay €500 … but it is getting nothing (no Target nor Action) and has done nothing (no Action), so it doesn’t explain “why” it has to be fulfilled.

Example 21 expands a bit by having an action and a target. It is good to illustrate the structure, it shows how to articulate a Duty, but doesn’t explain what triggered the rule. This on actually looks like a separate dimension for an “audit” type role (obligation says delete – assignee didn’t execute the action “delete”, therefore must compensate) and can’t be evaluated.

Regards,
___________________________________
Joshua Cornejo
marketdata<https://www.marketdata.md/>
embed open standards
across your supply chain


--

Víctor Rodríguez-Doncel

✉️ D2110 - Ontology Engineering Group (OEG)

Departamento de Inteligencia Artificial

ETS de Ingenieros Informáticos

Universidad Politécnica de Madrid

📞 +34 910672914

🌐 http://cosasbuenas.es<http://cosasbuenas.es/>