Re: Do resources have representations?

>    <http://web4.w3.org/> :mirrors <http://www.w3.org/>.
> 
> might be a good way to say (the true fact) that the browser experience
> is meant to be the same, modulo network issues, using those two URIs.
> But for that triple to make any sense in RDF, we have to consider the
> URIs as identifying something like ResponsePoints [1].   If you go any
> more abstract (like to "documents") then the :mirrors relationship is
> meaningless. 
> 
>      -- sandro
> 
> [1] http://esw.w3.org/topic/ResponsePoint

Agreed -- as vocabularies stand, URIs denote reponse points.

a vocabulary that allows one to state:

<http://web4.w3.org/> :mirrors <http://www.w3.org/>
<http://www.w3.org/> :hasRepresentation <blanknode>
<blanknode>	:mimeType	"text/html"
<blanknode>	:revision	"2"

might be able to let one infer enough information about the
representations at each responsepoint and their equivalence.  That's
useful, I think, and such statements should be very easy to make as a
bootstrap for the semantic web -- it will talk about existing resources
in a meaningful and less abstract but clearly defined way.

Ari

Received on Tuesday, 29 July 2003 13:53:30 UTC