RE: What if an URI also is a URL

Oskar,

I am no authority on this, but I'll test my own understanding with the following explanation: 

When we speak of a URI refering to something, we must consider in what context we are speaking. 

In the HTTP context, i.e. what does a server return, it refers to a context.

In the RDF context, i.e. what is the meaning of the URI in an RDF statement, a URI is a string of characters which uniquely identifies some "thing", some concept, such as a person, a book, the color "green", or in some cases a website or resource on the web. 

This is my own non-normative explanation, no doubt someone will correct some of the details.

Regards,

James Lynn

HP Software
W3C Service Modeling Language WG
610 277 1896


-----Original Message-----
From: www-rdf-interest-request@w3.org [mailto:www-rdf-interest-request@w3.org] On Behalf Of Oskar Welzl
Sent: Monday, August 27, 2007 4:19 PM
To: www-rdf-interest@w3.org
Cc: Reto Bachmann-Gmür
Subject: Re: What if an URI also is a URL


Thank you for helping me on... the only point I don't get is:

> Oskar Welzl wrote:
> > c) Consequently, if I want to make a reference to my weblog (meaning 
> > the thing as such, a collection of all posts, pictures, the services 
> > offered etc.), I'd better not use http://oskar.twoday.net as this 
> > would only refer to one single HTML-document. In particular, the 
> > document served at http://oskar.twoday.net might have been 
> > dc:created yesterday, while the weblog as such started 2003.
> > I'd have to make up something like 
> > http://oskar.twoday.net/id/thisblog
> > and maybe state somewhere that http://oskar.twoday.net/id/thisblog 
> > has an indexDocument http://oskar.twoday.net
> >   
> I think there might be a confusion between the resource 
> <http://oskar.twoday.net/> and the representation returned by the 
> webserver. A Blog is an Information Resource which could be described 
> as an ordered collection of posts, the HTML returned by the webserver 
> is (or should be) a suitable representation of that thing. Depending 
> on the current time as well as the properties of the request the 
> server may deliver different representations of an invariable resource.

So while I thought <http://oskar.twoday.net/> shouldnt be used to refer to the Blog as "the collection of posts", you seem to say now that the HTML-document returned by the server might well be taken as a suitable representation of it. Therefore (if I understand you correctly), <http://oskar.twoday.net/> *can* be used in RDF to refer to the thing known as "Oskars Blog" in the real world.

Now isnt this a direct contradiction to the starting point, when TBL
wrote:

> The moment a server returns 200 OK for a request to the URI, it is 
> saying it identifies a document.

Identifiying a document is not the same as identifying a blog, even if the document is part of the blog. I can make statements about this one document that are not true for the blog and vice versa.

Whats the point I'm missing? Or am I being fussy here again? ;)

Thx
Oskar

Received on Monday, 27 August 2007 21:18:55 UTC