ResponsePoint

From W3C Wiki

A ResponsePoint is a communications end point potentially waiting for a new message or connection. Servers answer clients at one or more ResponsePoint's. In some cases, with suitable coordination, multiple servers can share a single ResponsePoint, such as with round-robin DNS.

There is an argument (and this is why this is relevant on this wiki) that if you get a 2xx successful response from an HTTP URI, then that URI identifies (refers to, denotes) a functioning ResponsePoint. (This is issue tag:httpRange-14, and ties in with HashURI and SlashRedirection.) While strings which are syntactically HTTP URIs may be used in RDF to denote other things, doing so can be seen as conflicting with the basic ontology or design of the web.

Consider all the things one might mean by the term `http://www.w3.org/Consortium/`:

<http://www.w3.org/Consortium/> a r:ResponsePoint.


<http://www.w3.org/Consortium/> a r:ResponsePoint.
<http://www.w3.org/Consortium/> r:service :s1.


  • It could be the computer system `cs1` on which `s1` runs
<http://www.w3.org/Consortium/> a r:ResponsePoint.
<http://www.w3.org/Consortium/> r:service :s1.
:cs1 r:runs :s1.


  • It could be the document `d1` which is provided by `s1`
<http://www.w3.org/Consortium/> a r:ResponsePoint.
<http://www.w3.org/Consortium/> r:service :s1.
:s1 r:providesDocument :d1.


  • It could be a representation `r1` of the document `d1
<http://www.w3.org/Consortium/> a r:ResponsePoint.
<http://www.w3.org/Consortium/> r:service :s1.
:s1 r:providesDocument :d1.
:d1 r:representation :r1.


  This service.providesDocument.representation chain is lumped
  together in `log:content`, and taken one step farther in being
  parsed into an RDF graph/N3 formula by `log:semantics`.
  • It could be the primary subject of `d1`
<http://www.w3.org/Consortium/> a r:ResponsePoint.
<http://www.w3.org/Consortium/> r:service :s1.
:s1 r:providesDocument :d1.
:d1 r:primarySubject :subj.


  • It could be some user's experience of `r1`
<http://www.w3.org/Consortium/> a r:ResponsePoint.
<http://www.w3.org/Consortium/> r:service :s1.
:s1 r:providesDocument :d1.
:d1 r:representation :r1.
[ a r:userExperience; r:user <someUser>; r:repr :r1 ] 


  • ... etc.

Lots of options, but they build out from the URI as a logical symbol denoting a ResponsePoint to the services available at that response point, to what the services might do and who or what might provide them.

Documents and services can move, being offered at different URIs, but the connection between a URI and a ResponsePoint is static and syntactic. In many cases the URI and the ResponsePoint can be used interchangably, but of course they are different things.

Do they map 1-1? No, because of things like case folding.


Some e-mail comments from PatHayes, 2003-07-21