W3C home > Mailing lists > Public > semantic-web@w3.org > June 2008

Re: Naive question on redirection.

From: Richard Cyganiak <richard@cyganiak.de>
Date: Sat, 7 Jun 2008 10:25:07 +0100
Cc: SWIG <semantic-web@w3.org>
Message-Id: <EFB69D5B-9451-484F-BF7A-109FEE5A1520@cyganiak.de>
To: Phil Archer <parcher@icra.org>

Phil,

This is a very interesting question, thanks for rising it.

On 6 Jun 2008, at 09:57, Phil Archer wrote:
> Suppose I have this triple
>
> <http://example.org/> ex:colour "red"
>
> and when I dereference the URI I get a 302 redirect to http://www.example.org/home.asp 
> .
>
> Do I know what colour http://www.example.org/home.asp is?

I don't want to dig too deeply into this, but here's an explanation  
that makes sense to me. This is an interpretation that is not fully  
supported by the wording of the specs, but to me it seems to be the  
most sane one.

An HTTP GET to a URI means “Please give me a representation of  
yourself.”

A 302 response means: “I can't or don't want to give you a  
representation of myself, but you can get a representation of myself  
over there.”

The important point in this interpretation: 302 from A to B means that  
you get a *representation of A* by doing a *GET on B*.

With this interpretation, we cannot deduce anything about B from the  
interaction, except that among its representations there is one that  
is also a representation of A. We cannot deduce any owl:sameAs or  
whatever else from the redirection. The only deduction we can make  
from an HTTP redirect is “A http:redirectsTo B”.

I strongly prefer this interpretation because it keeps the layering of  
the SemWeb *on top of* HTTP intact. Configuring the HTTP responses of  
a URI should *only* determine what representations clients can access,  
*not* what the resource is. (With the single additional constraint of  
httpRange-14, that only a special class of things called web documents  
can have representations, and having representations of other things  
like people is a form of unhealthy punning that is best avoided.)

So, in my eyes, you don't know the color of <http://example.com/ 
home.asp> unless the RDF you get back from there tells you the color.

Richard


>
>
> I'm pretty sure the answer's no, but has anyone else grappled with  
> the joys of redirects in this way?
>
> Phil.
>
> -- 
> Phil Archer
> Chief Technical Officer,
> Family Online Safety Institute
> w. http://www.fosi.org/people/philarcher/
>
>
>
>
Received on Saturday, 7 June 2008 09:25:44 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 1 March 2016 07:42:05 UTC