W3C home > Mailing lists > Public > www-rdf-interest@w3.org > June 2004

Re: Mistaken identity?

From: <tpassin@comcast.net>
Date: Tue, 22 Jun 2004 16:06:56 +0000
To: danny666@virgilio.it
Cc: Danny Ayers <danny666@virgilio.it>, www-rdf-interest@w3.org, distobj@acm.org, ken@bitsko.slc.ut.us, Norman.Walsh@Sun.COM
Message-Id: <062220041606.13172.40D8592000099E9100003374220073747802079C9C0E9F9B@comcast.net>

> 3. How can I ever assert:
> 
> <http://mydomain.org/a> owl:sameAs <http://yourdomain.org/a> .

Just because you use a particular URI as an rdf identifier doesn't automatically mean that the rdf resource is equal to the web page returned when you dereference the uri.  In fact, rdf's position about that notion is not really specified.  IMHO, you are better off considering your URIs to be abstract strings, and then hope that you get something cogent when you try to dereference them. If you do get something, so much the better.

So how could you make a statement about the actual thing returned by http://mydomain.org/a?  Make up a resource type, and assign the URL to it, something like this -

{danny:a rdf:type danny:WebPage}
{danny:a danny:URL 'http://mydomain.org/a'}

Get fancy and declare danny:URL to be inverse functional if you like.  

I know that some people don't like this kind of approach, but it removes the ambiguity and clearly says what you mean.  It gives you the chance to attach qualifiers, like the date for which the URL applies, and has other potential useful characteristics.  That makes up for the fact that you have to have one or two additional triples compared with your example.

Cheers,

Tom P

PS - Topic maps handle this kind of situation pretty much like the above sketch, and for the same reasons - to remove as much ambiguity about the URI idenitifier as possible.  With TM, the relevant predicates and resources are built in.
Received on Tuesday, 22 June 2004 12:07:29 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:52:07 GMT