W3C home > Mailing lists > Public > semantic-web@w3.org > September 2006

Re: A URI for your Favourite Pub: httpRange-14 Question

From: Chris Bizer <bizer@zedat.fu-berlin.de>
Date: Thu, 21 Sep 2006 21:31:14 +0200
Message-ID: <000e01c6ddb4$807fd1b0$83ec2da0@named4gc1asnuj>
To: "T.Heath" <T.Heath@open.ac.uk>, <semantic-web@w3.org>

Hi Tom,

as I understand it, TAG really doesn't want you to mix up the pub's homepage 
with the pub itself.
You can see it this way, the owner of the pub has assigned a URI to the 
homepage of the pub but hasn't assigned a URI to the pub itself yet. Until 
he does, it is best to create a new URI in your namespace (like 
http://mydomain.com/things/theredlion) to talk about the pub. If you say 
intersting stuff about the pub, other people might pick up your URI and also 
use it for talking about the pub.

If this URI gets dereferenced, your server should respond with a 303 
pointing to an RDF document which contains your information about 
http://mydomain.com/things/theredlion.

If, in the future, the guy assigns a URI to his pub, you should use his URI 
instead of your 'temporary' URI.
Or alternatively, if your URI is already used by other people, add a 
<yourURI> owl:sameAs <hisURI> statement to your information.

This all this sounds complicated, but makes sense as
1. a pub and its homepage are definitively two different things
2. you shouldn't define URIs in the namespace of somebody else, as with 
http://www.theredlion.co.uk/#uri.

So TAG has voted for a clear architecture in the long-run and against nice 
hacks in the short run ;-)

Cheers (and see you at ISWC?)

Chris

----- Original Message ----- 
From: "T.Heath" <T.Heath@open.ac.uk>
To: <semantic-web@w3.org>
Sent: Thursday, September 21, 2006 7:05 PM
Subject: A URI for your Favourite Pub: httpRange-14 Question



Hi all,

I've got an httpRange-14 related question, and was hoping someone might
be able to help; apologies if this is going over old ground, I've read
and read and am still no clearer.

Ok, so here's the situation: you want to make some RDF statements about
your favourite pub (lets call it "The Red Lion"), and so you need a URI
to identify it.

- Option 1: you mint your own URI, something like [1], even though this
URI is only really used to identify that pub by you, and you're
therefore guilty of not avoiding URI aliases [2].

- Option 2: the pub has a website [3], so you use that URL as your URI.
Then when the URI is dereferenced some information about the pub is
returned. The pub hasn't quite got around to providing an RDF
description of itself on the site, but at least a human with a browser
can read the pub menu, and you haven't created unnecessary URI aliases.

Clearly Option 2 seems best, right? Well, not if you read the TAG advice
about minting URIs [4]. At least according to my reading of [4] then if
a GET request to the pub website [3] returns an HTTP 2xx, then the URI
refers to an "information resource" (i.e. the pub homepage in this
case), NOT to the pub itself. So, when you make all your RDF statements
with that URI as your subject, then you're actually making statements
about the pub homepage, not the pub itself. Or have I missed something??

Does anyone have an elegant solution to this problem? DualUseUris [5]
sound great if we were talking about my own site, but in this case it
would require me to persuade the pub landlord to set up content
negotiation on their server(!). You can imagine the response: "Content
what? You're havin' a laugh mate!".

Am I best off just minting my own, and ignoring the consequences? What
about adding a fragment to the existing URL to distinguish the URL from
the URI, even though it will have no effect when the URI is dereferenced
(again, unless I get technical with the landlord)? What about minting my
own like [7] and configuring the server at mydomain.com to reply to
dereference attempts with HTTP 303 and the URL of the pub?

Any help very greatly appreciated,

Thanks,

Tom.

Supplementary Question:

1. Does an RDF Document count as an information resource? I.e. if you
dereference a URI and get HTTP2xx and some RDF back, then does that
imply that statements about that URI are statements about the RDF
Document, rather than the thing described in the RDF?

[1] http://mydomain.com/things/theredlion
[2] http://www.w3.org/TR/webarch/#uri-aliases
[3] http://www.theredlion.co.uk/
[4] http://www.w3.org/2001/tag/issues.html#httpRange-14
[5] http://esw.w3.org/topic/DualUseUri
[6] http://www.theredlion.co.uk/#uri
[1] http://mydomain.com/things/http//www/theredlion/co/uk/

-- 
Tom Heath
PhD Student
Knowledge Media Institute
The Open University
Walton Hall
Milton Keynes
MK7 6AA
United Kingdom

Tel: +44 (0)1908 653565
Fax: +44 (0)1908 653169
Web: http://kmi.open.ac.uk/people/tom
Email: t.heath@open.ac.uk
Jabber: t.heath%open.ac.uk@buddyspace.org
Received on Thursday, 21 September 2006 19:49:08 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 1 March 2016 07:41:53 UTC