Re: [poe] whats's the rationale behind uids?

> I don't have a problem with blank nodes - but all means use them....but the ODRL Info Model says you must uniquely identify the party/asset. Hence, the blank node requires the uid property.
> 
Yeah well… then indeed we have a problem. This means the model includes a restriction that we may have a problem with…

I guess a MUST can be replaced by a SHOULD.
> You say "... then the implicit identification of the node via the usage of that predicate is also perfectly fine."
> I would then ask, please explain how the blank node (with foaf:name Bob) implicitly does that?
> 

If a predicate is declared to be inverse functional (which is not the case for foaf:name but just suppose this for a moment) then if 

```
A foaf:name "bob".
B foaf:name "bob".
```

then an OWL reasoner would deduce that 

```
A owl:sameAs B
```

You are right this is not identification in terms of assigning it an explicit `id`. But for the underlying goal that, I presume, this is mentioned in the model, this may be just enough.

-- 
GitHub Notification of comment by iherman
Please view or discuss this issue at https://github.com/w3c/poe/issues/174#issuecomment-306688018 using your GitHub account

Received on Wednesday, 7 June 2017 05:08:46 UTC