Re: Identity of URIRefs / Resources

Reto,

> If I understand things correctly an RDF graph is a *set* of triples, so
> duplicates are never part of model even if a serialization may contain
> the same triple multiple time.  If the resource with URIRef
> "http://gmuer.ch/%C3%BC" would be necessarily the same as the one with
> URIRef "http://gmuer.ch/ü" the model would contain only one statement. 

You're correct.  I was wrong.  RDF/XML and much software treat duplicates as
insignifiant [1,2,3].  Dupicates have the same value as well [4], so they add
no additional meaning to the document [5,6].  However, it can be expensive to
remove duplicates in practice [7], so watch out!

> > > Does it make sense that "Two RDF URI references are equal if and only if 
> > > they compare as equal, character by character, as Unicode strings.", 
> > 
> > Yes. That is correct.
> > 
> > > wouldn't it cause less problems to say "Two RDF URI references are equal
> > > if and only if the resolve to the same URI".
> > 
> > The problem is that two RDF URI references don't have to "resolve" to 
> > anything! They could point to a resource that is not network 
> > retrievable, for example. For this reason, I think the tag URI scheme 
> > is a good idea for most URIs. See <http://taguri.org> for more info 
> > on that scheme.
> 
> My use of the word "resolve" was misleading. I didn't mean to talk about
> dereferencability of the resource but about the valid URI that is
> produced with the method describe at
> http://www.w3.org/TR/rdf-concepts/#dfn-URI-reference and which is the
> same for the two URIRefs in my example.

I'm a little uncertain by what you mean.  The only way for two URI strings to
be equal is for them to have EXACTLY the same character sequences.  They only
consist of a subset of ASCII characters too.

Note that <http://gmuer.ch/aü> is not a valid URI can can never be a resource
name.  The string <http://gmuer.ch/a%C3%BC> is a valid URI and so can be a
resource name.  So is <http://gmuer.ch/%61%C3%BC> even though that form is
depreciated.  Each URI is also distinct, so these URI references identify
different resources (without OWL):

<http://gmuer.ch/a%C3%BC>
<http://gmuer.ch/a%c3%bc>
<http://gmuer.ch/%61%C3%BC>
<http://GMUER.CH/A%C3%BC>

Case is signifiant!  Yes, that is annoying IMHO.

> > > I'm asking because I'm implementing and RDF based CMS [1] where GET and
> > > MGET requests are answered according of the properties the requested
> > > resource has in the model and I have no way to find out whether the user
> > > requested http://gmuer.ch/%C3%BC or http://example.org/��;.
> > 
> > Take a look at the SPARQL protocol, which provides a standard way of 
> > using HTTP to get RDF information.  See 
> > <http://www.w3.org/TR/rdf-sparql-protocol/> for more information.
>  
> I'm not sure how I would have to express in a SPARQL-Get query, that I'm
> talking about the resource with URIRef http://gmuer.ch/%C3%BC and not
> about the one with URIRef http://gmuer.ch/ü.  The same problem applies
> for URIQA.

I'm a little lost.  Can you go though an example of how you are using HTTP?  A
typical usage pattern?

--
Jimmy Cerra

[1] http://lists.w3.org/Archives/Public/www-rdf-interest/2004Oct/0073.html
[2] http://lists.w3.org/Archives/Public/www-rdf-interest/2004Oct/0090.html
[3] http://lists.w3.org/Archives/Public/www-rdf-interest/2004Oct/0091.html
[4] http://www.w3.org/TR/rdf-concepts/#section-Graph-syntax
[5] http://lists.w3.org/Archives/Public/www-rdf-interest/2002Nov/0410
[6] http://lists.w3.org/Archives/Public/www-rdf-interest/2002Nov/0414.html
[7] http://lists.w3.org/Archives/Public/www-rdf-interest/2004Mar/0105



		
__________________________________ 
Discover Yahoo! 
Find restaurants, movies, travel and more fun for the weekend. Check it out! 
http://discover.yahoo.com/weekend.html

Received on Sunday, 5 June 2005 17:49:18 UTC