W3C home > Mailing lists > Public > www-tag@w3.org > September 2007

RE: Dereferencing HTTP URIs (redux?)

From: Booth, David (HP Software - Boston) <dbooth@hp.com>
Date: Wed, 12 Sep 2007 22:29:17 -0400
Message-ID: <EBBD956B8A9002479B0C9CE9FE14A6C203292A69@tayexc19.americas.cpqcorp.net>
To: "Alan Ruttenberg" <alanruttenberg@gmail.com>
Cc: "John Cowan" <cowan@ccil.org>, "Pat Hayes" <phayes@ihmc.us>, "Williams, Stuart (HP Labs, Bristol)" <skw@hp.com>, "Mark Nottingham" <mnot@yahoo-inc.com>, "W3C-TAG" <www-tag@w3.org>

> From: Alan Ruttenberg
> 
> I'm not sure thing-described-by help. For one thing, the URI 
> that you  
> create using that service is not the one that you started with, so  
> the relationship between u and what you get is not direct. 

I'm assuming you mean that the host name of 
http://thing-described-by.org?http://aunt-tillie.example/foo
is different from the host name in
http://aunt-tillie.example/foo .  Yes, that is true, hence it won't help
if you want them to be the same.  But bear in mind that even though it
may look pretty to have them use the same host name:

 - Two URIs with different host names may in fact have the same owner.

 - Two URIs with the same host names may in fact have different owners,
due to delegation of a portion of the URI space.

> Second,  
> from the point of view of the english sense of the thing, the thing  
> that describes something isn't the thing itself.

I don't know what you mean by this.  
http://aunt-tillie.example/foo can be a document that describes a
non-information resource identified by
http://thing-described-by.org?http://aunt-tillie.example/foo .

> 
> I suppose that this would be OK if it was the result of a temporary  
> redirection from u, but then
> a) Setting up that redirection may be an equally complicated 
> task for Aunt Tillie

I don't know what you mean.  Aunt Tillie did not have to set up
*anything* to use the thing-described-by.org redirect service.  Unlike
purl.org, her page is *not* registered with thing-described-by.org .
Thing-described-by.org always 303-redirects to whatever URl you pass as
a query string.  See the explanation at http://thing-described-by.org .

> b) How does one compute the effective result code through a 
> redirect?  
> Is it the first, the second, the last? Can't be the last, 
> because the  
> 303 redirects to something that returns a 200. Can't be the first  
> because such a redirection would be a 302.

I'm not sure what you mean by "effective result code".  If the *first*
response code is 200, then the httpRange-14 decision says that the URI
identifies an information resource.  If the *first* response code is
303, then that decision says the URI may identify any resource, and if
the last response is 200, then the representation returned *might* tell
you what resource the original URI identifies.

My personal opinion is that if the first response code is 302, then the
second response can be treated as the first response.  I think this is
the intpretation in the current draft of the httpRange-14 finding:
http://www.w3.org/2001/tag/doc/httpRange-14/2007-08-31/HttpRange-14.html
#sec-redirection-response-codes


David Booth, Ph.D.
HP Software
+1 617 629 8881 office  |  dbooth@hp.com
http://www.hp.com/go/software

Opinions expressed herein are those of the author and do not represent
the official views of HP unless explicitly stated otherwise.
 
Received on Thursday, 13 September 2007 02:29:40 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 22:56:17 UTC