Re: "information resource"

On Wed, 2004-09-01 at 07:42, Patrick.Stickler@nokia.com wrote:
> Regarding the August 16 version of "Architecture of the World Wide Web"
> 
> http://www.w3.org/TR/2004/WD-webarch-20040816/#dereference-uri
> 
> In section 3.1 it states:
> 
> "The term Information Resource refers to resources that convey information.
>  Any resource that has a representation is an information resource."
> 
> While I understand the desire to introduce a term which enables people
> to speak directly about resources which are web-accessible, it seems to
> me that this particular term will provide more confusion than utility.

It is incumbent on others to use the term as defined by the document,
not by how they think they would define it or what they think those
particular words might mean. Just because it might cause someone some
discomfort doesn't mean its wrong or badly named.

> Since *any* resource *can* (potentially) have a representation, the
> membership of the class of "information resources" is a reflection of
> the management, over time, of those resources, not any intrinsic 
> characteristic of the resources themselves.
> 
> I may have a dog, which is denoted by a URI, and if I choose to publish
> representations of that dog via that URI, that in no way changes the
> nature of that dog. And I have a hard time thinking of that dog as
> an "information resource", just because someone can dereference its
> URI to get some representation of the dog.

Better get used to it because every physical instance of every product
in the world is about to get a URI. That box of Gillette razors is going
to have a URI that denotes that exact physical thing. Dogs are going to
have RFID chips embedded in them and there are standards for encoding
the binary identifier in that RFID chip as a URN and the exact intent is
that this URN denotes that particular dog. It is an information
resource. It produces information. There are network layer methods for
transferring the information produced by that dog from the physical
world to the electronic (the concept of a boundary between the two
becoming more and more useless every day).

> Likewise, membership in this class of "information resources" will
> be transient. At one time, there may be a URI denoted resource that
> has no representation. Then it does, at which time it becomes an
> "information resource". Later, the representations are no longer
> accessible, at which time it ceases to be an "information resource";
> insofar as the definition provided is concerned.

It ceases to be an information resource from your perspective. Based on
policies associated with the identifier (be they from the scheme used,
communicated to you through previous interactions, etc) you may be able
to expect that it a) may become an information resource again and b) may
still be an information resource to someone else but right now you can't
prove that it is one.

> I expect that most folks will percieve membership in the class of
> "information resources" to reflect an intrisic characteristic of
> the resource itself, rather than simply a condition of its management
> (or management of representations of that resource, depending on 
> how you look at it).

This is the first time I've heard of anyone making that kind of
distinction....

-MM

Received on Wednesday, 1 September 2004 13:15:33 UTC