- From: Richard Cyganiak <richard@cyganiak.de>
- Date: Wed, 29 Aug 2007 20:49:19 +0200
- To: wangxiao@musc.edu
- Cc: "Booth, David (HP Software - Boston)" <dbooth@hp.com>, Tim Berners-Lee <timbl@w3.org>, Ed Davies <edavies@nildram.co.uk>, Technical Architecture Group WG <www-tag@w3.org>
On 29 Aug 2007, at 15:40, Xiaoshu Wang wrote: > Richard Cyganiak wrote: >> There's a lot of discussion currently about the distinction >> between IRs and non-IRs. Is it about “essential characteristics >> conveyable in a message”, about “can we attach an HTTP endpoint to >> it”, about “document-ness”? >> >> To me, this all misses the point. Even if we can nail down >> objective criteria to distinguish these buggers, this will *still* >> not tell us if we have to serve them up using 303/hash or 200. > I think to focus on what is information resource and what is not > missed the whole point of httpRange-14. We use the term IR vs. non- > IR to help the discussion of the topic because that is where causes > the ambiguity. But the point of httpRange-14 is not to distinguish > which resource is IR and which is not. It is about making our > statements unambiguous. > A machine couldn't careless if a URI identifies an IR or a non-IR. > And there is no reason for a machine NOT to believe that a person > shouldn't have a width/height/background or stylesheet or a > document can not have spouse and children. > In other words, I can think a person to be an information resource. > If you don't agree then I cannot communicate with you. But for me > and for people who agrees with my view, we should still get along > fine even if we didn't 303 or # redirect our personal URI. > > So, the issue is not about what is information resource and what is > not. It is about how we can convey our message in a way that > minimize confusion. The point of httpRange-14 is to tell us that > there is a potential cause for making ambiguous statements and > possible way to avoid it. Hmmm, interesting. Let me try to paraphrase what you said: “The distinction between IR and non-IR is not an inherent property of a resource. It is a modelling decision that web publishers have to make when they allocate a URI for a resource. There is no right or wrong way to make the decision. But communication is improved when everyone makes that decision in a consistent way, and thus we have a rule of thumb to guide us: It's an IR if and only if its essential characteristics can be conveyed in a message.” Is that approximately what you mean? This raises two questions for me: 1. What are the reasons against settling on IRs for everything, and ditching non-IRs completely? You mention the resulting confusion between the document and the document's topic. Anything else? 2. This boils down to a question wether to provide a representation of a resource, or instead provide an associated description of the resource (by means of a 303 redirect or hash truncation). What is the difference between a representation and a description? Is it purely a difference of delivery? Surely there must be a deeper difference when we think it is necessary to provide both in WebArch? 2b. For bonus points, explain why, in order to reduce ambiguity, web pages need representations but cars need descriptions. Cheers, Richard > > That's my two cents... > > Xiaoshu Wang > >
Received on Wednesday, 29 August 2007 18:50:20 UTC