This blog is developed to make notes on the development of DFDF at http://dfdf.inesc-id.pt

Thursday, November 15, 2007

URI Identity and Web Architecture Revisited

Out of the discussions on W3C's TAG mailing list, I have written down my thoughts on URI's identity issue and my personal viewpoint on the architecture of the web. The article is published at here.

Here is a brief summary.

(1) Current definition of resource ignores the nature of URI as an interface to the web. There are three different kind of resources, the one that we cared about should be defined as "abstract entities that have a dereferencible URI in the web".
(2) The current definition of "information resource" on the AWWW document is not well thought. The debate about what information resource is do not solve any real issues.
(3) TAG's httpRange-14 is incorrectly phrased. HTTP response code should indicate if a URI is informational but not resource.

4 comments:

Kingsley Idehen said...

What about simply describing an Information Resource and basic payload unit for the Web Information BUS? That is to say: Data+Representation is the basic of exchange on the Web Information BUS.

Data without Representation is a Resource that isn't directly Web Information BUS palatable.

I am not a Web Page, Weblog, Bookmark, Photo, or anything else, but all of these Resource Types combine data and representation which make them Web Information BUS palatable, and can be used triangulate the entity *Me* via the Web :-)

In our Physical Web, we do the same thing using a different Information BUS that accepts DNA based payloads that our DNA input devices comprehend. But even in this realm, you can only at best, triangulate the entity *Me* from what you perceive via DNA based input and output devices :-)


Kingsley

Xiaoshu Wang said...

I do not know what do you mean "Data + Representation"? Representation is or contains data, isn't it?

Ed Summers said...

This document has meant a lot to me over the past few years in thinking about the Linked Data program. I wish that your link to the paper didn't 404. I did manage to find a copy here if anyone else is interested.

Ed Summers said...

And Xiaoshu just uploaded it to a more permanent location at arXiv http://arxiv.org/abs/1212.6372 ; thanks Xiaoshu!