Re: The 'missing protocol part'?

Stasinos Konstantopoulos wrote:
> On Wed Apr 23 06:50:58 2008 Phil Archer said:
> 
>> A first pass at this leads me to suggest that for a URI, u, the basic 
>> function is
>>
>> describe(u)
>>
>> which would return RDF triples like
>>
>> <rdf:Description rdf:about="u">
>>   <ex:color>red</ex:color>
>>   <ex:shape>square</ex:shape>
>> </rdf:Description>
>>
>> (or the same thing in other serialisations)
> 
> Isn't POWDER also about resource discovery? A function
> 
>  described( d, h )
> 
> with the same POWDER doc and/or creator parameters as describe(),
> returning a set of resources that (a) are described by d and (b) are
> subsumed by host or hosts h.

Yes, you're right (and I'd make h optional) - but what would it would 
return? The relevant XML fragment?

<iriset>
   <includehosts>example.org</includehosts>
</iriset>

Suppose you want all the resources that are red and square, you'd have a 
query like

?&ex;color=red&amp;&ex;shape=square

meaning 'tell me all the things you know about that are red and square'

And then we probably want to be able (again, optionally) to restrict the 
source of the descriptions, so we could ask 'tell me all the things that 
maker X says are red and square'

A PP is going to have to do a lot of processing

And are we in danger of re-inventing SPARQL?

How would OWL processors handle a request such this? returning the 
property restrictions for all sub classes of classes that match the 
query parameters?

P

Received on Wednesday, 23 April 2008 08:27:04 UTC