See also: IRC log
<Stuart> Hello /me just dialing (having checked in an early comment on the vocab page).
<jar> We should finish the HTTP in RDF review: http://esw.w3.org/topic/AwwswHttpVocabularyInRdfComments
Stuart: What do you mean by "subject matter"?
jar: I mean the RDF that we want to write. I'll fix the sentence. Sorry for the confusion.
<scribe> ACTION: Jar to clear up confusion about what he means by "subject matter" [recorded in http://www.w3.org/2009/01/06-awwsw-minutes.html#action01]
<trackbot> Created ACTION-12 - Clear up confusion about what he means by \"subject matter\" [on Jonathan Rees - due 2009-01-13].
jar: Rules of engagement okay?
dbooth: Fine to start with.
jar: Also, when you introduce a class you need to have a relation to that class, to help understand what the members are.
<scribe> ACTION: jar to add additional rule about having a relation when you add a class [recorded in http://www.w3.org/2009/01/06-awwsw-minutes.html#action02]
<trackbot> Created ACTION-13 - Add additional rule about having a relation when you add a class [on Jonathan Rees - due 2009-01-13].
jar: rfc2616:Entity okay?
dbooth: Looks good to me.
harry: If you don't have a URI, how could you post or put something there?
jar: Through some other protocol or API. You can also do a GET using an IP address and use that connection.
harry: If rfc2616:Resource doesn't require a URI, then we may need another class that does have a URI, so that people know they can access it with HTTP, etc.
<Stuart> Hmmm Re: "For example, suppose I connect to an HTTP server that has no DNS host name, and specify a relative URI." that you have established a connection at least gives you an IP address that could be used in the authority field.
harry: I'm okay with removing the URI requirement for the rfc2616:Resource definition, but we may want something like that later.
<Stuart> In Fieldings model of the world the mapping goes from URI to resource and from resource (modelled a "time varying membership function") to representations or URI (which I take to be redirection or the like).
<jar> JAR requests use cases from Harry that would benefit from modeling a has-URI/has-no-URI distinction.
<Stuart> ie. by Fieldings model the identifying URI is not intrinsic to the resource itself.
<jar> (TimBL: about separation of resource from its name (URI).)
<Stuart> I used to have a "...these are not the resources that you are looking for..." in the sense that the resource could simply be portrayed as the things that all the respresentations obtained from the resource are representations of... ie. if it's character changes - a weather page becomes a gambling page is just a strange resource and "...not the resource you were looking for...".
<Stuart> However, even though it may be poor practice allowing the mapping between URI and resources to change then the in general URI are not/cannotbe intrinsic to the resources they denote.
jar: What about "comes from"?
<hhalpin> So, for timbl: http:rfc2616=awww:Representation
timbl: Shouldn't this be "has representation", as in AWWW?
jar: Wanted to start with 2616 notion first, not AWWW.
<Stuart> From 2616:
<Stuart> An entity included with a response that is subject to content
<Stuart> negotiation, as described in section 12. There may exist multiple
<Stuart> representations associated with a particular response status.
<Stuart> FWIW: IMO rfc2616:Representations are admissable as entity bodies on requests and respinses.
timbl: It's pointing out there may be multiple representations for a given resource -- not that conneg is a key characteristic of representation.
harry: I think it means it *might* be subject to conneg. Otherwise we have too many defs of representation around.
timbl: There's no way to know if conneg happened.
dbooth: The ability to have conneg is a by-product, not the distinguishing characteristic.
<Stuart> Ahah... TimBL has just hit upon an essential facet of a class defn that we may have missed - identity criteria, ie a test for determining sameness.
dbooth: I'm arguing that the domain of the "comes from" property is actually rfc2616:Representation: i.e., there does not exist an rfc2616:Entity X and rfc2616:Resource R such that "X comes from R" and X is not an rfc2616:Representation.
jar: Talk about HTTP in RDF doc next time?
<Stuart> Ok... I think that there is pun going on around representation - in the sense of a relation between and entity and a resource in the sense that the given entity is said to webarch:represent the rsource and a syntactic entity that has a mime type and bits.
<hhalpin> I agree Stuart.
<hhalpin> I think the punning is actually three ways.
<jar> hhalpin: Let's figure out how to plug 'web resource' and 'entity' into http-in-rdf vocabulary.
<Stuart> awww use representation to refer to such an entity, but I can see that that an entity in fact awww:represents a resource can only be objecivly determined by obtaining said entity in a response.
This is scribe.perl Revision: 1.133 of Date: 2008/01/18 18:48:51 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Succeeded: s/peopel/people/ Succeeded: s/on/in/ Succeeded: s/it/ie/ No ScribeNick specified. Guessing ScribeNick: dbooth Inferring Scribes: dbooth Default Present: DBooth, TimBL, Jonathan_Rees, Stuart, hhalpin Present: TimBL DBooth JonathanRees StuartWilliams HarryHalpin Got date from IRC log name: 06 Jan 2009 Guessing minutes URL: http://www.w3.org/2009/01/06-awwsw-minutes.html People with action items: jar[End of scribe.perl diagnostic output]