Re: Uniform access to descriptions

Xiaoshu Wang wrote:
> >
> > Client dereferences [a], preferring application/rdf+xml.  Response
> > to [a] is 200 OK, Content-Location [d].  Thus, it may be inferred
> > that [d] is also a variant of the resource identified by [a],
> > "topic".  This works perfectly well, if in fact [d] is an RDF
> > representation of "topic" and not a description of resource [a] as
> > "a wiki page". However, what if topic.rdf contains only the
> > assertion that [a] "is a" "wiki page", which *is* a description of
> > resource [a]? 
> All of them are /descriptions/.  I failed to follow what is the 
> difference between /representation/ vs. /description/. In fact, I 
> wouldn't prefer to give [b,c,d] a different URI.  I.e., there is no 
> Content-Location.  And I do have a concrete use case described at 
> http://dfdf.inesc-id.pt/tr/web-arch, where alternative location is
> not preferred. 

You can omit Content-Location if you want, my example uses that header
to clarify what's going on.  It is a best-practice to include Content-
Location, otherwise you are disabling caching.  In such a case, your
variant representations are not awww:resources or even Information
Resources, though, so it's definitely a fringe case.  If I were stuck
in the position you describe, I would re-factor my architecture to
conform with mainstream Web usage.

>
> I don't follow here.  If you are the resource owner of [a-d], don't
> you know that [b-d] describes [a]?  And why do you waste another
> round trip to 303 (or 400)?  Sure, it doesn't break anything. But
> what do you gain?  (I am talking in strict sense, I understand that
> you may want to do that due to saving your development effort).
> 

There's nothing about my development effort that needs "saving", I have
no ulterior motive here.  No, I do not know that [b]-[d] describes [a],
nothing about [b]-[d] indicates that [a] is "a wiki page".  Unless [d]
is an RDF document which asserts that [a] is "a wiki page", which is
clearly a description of [a] because [a] is not a representation of the
concept of "a wiki page", it is a representation of the concept of
"topic".  Why bother with 303 or 400?  Because if [d] is *not* a
representation of [a], then a 200 OK response returning [d] as the
result of content negotiation would be completely misleading and wrong.

-Eric

Received on Sunday, 13 April 2008 23:38:49 UTC