Re: [foaf-protocols] owl:Restrictions in ACL - was Re: ACL Ontology and Discussion

On 21 Apr 2010, at 16:58, Joe Presbrey wrote:

>> Sadly, I'm going to have to implement the above in the short term though
>> as can't for the life of me see any other way of expressing:
>> 
>> if graph <Gx> holds the triple <group> <has_member> <webid> .
>>  where <Gx> is found by dereferencing <group>
>>  where <group> and <has_member> are defined in the ACL
>>  where <webid> is only known at runtime after checking foaf+ssl.
> 
> I'm not sure about the Gx dereference at this point.  I guess this is
> analogous to multiple groups being defined in the same graph and
> distinguished by fragment.  Are you planning to do this like:
> defrag(group)?  Along these lines I think Apache people like
> AuthGroupFile in .htaccess which explicitly declares what "graphs" are
> trusted for authorization.  Maybe you aren't counting this as another
> way of expressing, but in SPARQL I would say:
> 
> ASK {
>  GRAPH <acl> {
>    ?acl acl:accessTo <uri>; acl:mode acl:Read; acl:accessClass <group> .
>  }
>  GRAPH <group> {
>    <group> <has_member> <webid> .
>  }
> }

yes, a SPARQL query such as the above, if the engine knows to go and dereference the Graph ids, 
is close to the N3 rules I was thinking of.

The OWL inferencing is then done inside the graphs such as <group>.

Henry