W3C home > Mailing lists > Public > public-sweo-ig@w3.org > June 2007

[Fwd: Re: What if an URI also is a URL]

From: Dan Brickley <danbri@danbri.org>
Date: Wed, 06 Jun 2007 17:52:45 +0200
Message-ID: <4666D84D.5030606@danbri.org>
To: SWD Working Group <public-swd-wg@w3.org>, 'W3C SWEO IG' <public-sweo-ig@w3.org>
CC: Tim Berners-Lee <timbl@w3.org>
SWD, SWEO,

With a SW Interest Group chair's hat on .... "help!"

This perma-thread (the question of URIs for people, HTTP content 
negotiation, meaning of #blah IDs etc) has been rumbling along for 
years. It is getting really embarrassing that this stuff keeps coming 
up. Can we do something here? A Faq or TAG finding or wiki page or ... 
... something, anything, to help the mailing lists move along to more 
fruitful topics of discussion?

Apologies for the x-post, this problem is split a little across groups.

cheers,

Dan

attached mail follows:



Tim Berners-Lee wrote:
> This is the web architecture.   The client strips off the '#me'  and 
> acceses the dcoument
> <http://www.example.com/mophor> (if it hasn't already for some other id 
> in the same document).
> The server sends back a document telling it about 
> <http://www.example.com/mophor#me> and maybe other things.

When a client is looking at what it received from 
http://www.example.com/mophor it doesn't know that's the only possible 
view upon http://www.example.com/mophor (indeed it might know the 
opposite if there is a Vary header)

When the client is interpretting #me in a way that is based on the 
content-type of what it received from http://www.example.com/mophor it 
is more tightly concerned with a particular document.

Since the mechanism for a fragment identifier is based on the document 
type, which seems if anything even more tied to documents than 
http://www.example.com/mophor
Received on Wednesday, 6 June 2007 15:53:11 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:17:38 GMT