RE: new text for Information Resource (section 3.1)

> -----Original Message-----
> From: public-webarch-comments-request@w3.org
> [mailto:public-webarch-comments-request@w3.org]On Behalf Of ext Sandro
> Hawke
> Sent: 08 September, 2004 22:57
> To: public-webarch-comments@w3.org
> Subject: new text for Information Resource (section 3.1)
> 
> 
> 
> 
> I found the definition of "Information Resource" in 3.1 [1] very
> confusing and discussion on www-tag suggests it's overly ambiguous.
> I'm willing to provide evidence of its flaws if necessary, but I'd
> rather skip to some replacement text.  I believe this text is
> compatible with the current text and merely clarifies matters, but
> people who interpret the current text differently may find this a
> significant change.
> 
>    An "Information Resource" is a collection of information
>    potentially transmittable via a computer network.  Digital forms of
>    creative works (such as documents and images) are Information
>    Resources, while certain conceptual entities (such as numbers and
>    RDF properties) are not.  This distinction is becoming useful as
>    people develop ways to use URIs to identify things which are not
>    Information Resources.
> 
>    Physical objects and phenomena (eg sound) are not Information
>    Resources, but they may be measured or otherwise used to produce
>    information which can form Information Resources.  A hand written
>    note is not an Information Resource, but a digital scan of it is.
>    The weather at a particular place (a physical phenomenon) is not an
>    Information Resource, but a collection of measurements or
>    predictions of the weather are.  In the strictest sense, a
>    computer-controlled "web-accessable" coffee-maker is not an
>    Information Resource, but its counters, timers, and the readings of
>    its sensors are.  The coffee-maker itself might have scratches,
>    stains, and be positioned in a particular way on the counter, but
>    it is the counters, timers, and sensor readings which are used to
>    generate the web page.

I'm OK up to this point -- though would suggest that the above definition
is not critical to AWWW and should simply be defined in an RDF Schema
so folks can classify resources as "information resources" as needed.

>     
>    Information Resources are the only kind of resources which can have
>    representations.  

Nope. I can't agree here.

A representation reflects the state of a resource. I see no reason why
a URI can't denote the actual coffee-maker (not its counters, timers, etc.)
and have that URI resolve to a representation of the coffee-maker, reflecting
its state (its counters, timers, etc.)

Whether a URI denotes the coffee-maker, or some data structure storing
information about the coffee-maker, shouldn't matter to a web client,
since the representation returned will likely be similar or the same.
But the semantic distinction between the coffee-maker and some data structure
storing information about the coffee-maker is certainly important to a
semantic web client, especially when e.g. the owner/creator of the coffee-maker
is likely different from the owner/creator of the data structure, etc.

>From a web perspective, the actual denotation of the URI is irrelevant. What
folks care about are the consistent representations accessible (an it likely
is impossible to determine what the actual denotation is based on the 
representations).

That's why the semantic web is so important. It allows folks to say what those
URIs actually denote, and describe the denoted resources accordingly.

But saying that only "information (bearing) resources" can have representations
is both (a) unnecessary, and (b) overly limiting -- forcing folks to posit
"information resources" in order to provide representations of dogs, planets,
colors, etc.

> The number 1, which is not an Information
>    Resource, might be said to be represented by the two-octet sequence
>    0x0001, but not in the sense of "representation" used in this
>    document.  A web-accessible control dial, set to "1", might respond
>    to HTTP GET requests with a representation of its state: 0x0001.
>    In this example, 0x0001 acts an identifier for the number 1 within
>    the data format being used.  An HTTP GET of a URI for the number 1
>    itself could meaningfully be met with an error or redirect, but not
>    with a representation.
> 
> I'm sorry for poking at a sensitive spot in the document; I hope this
> text is clear and precise enough to be helpful.

I'm all for clearer definitions. My proposal would be to simply rename
"information resource" as presently defined in AWWW with "web resource"
and leave the web architecture agnostic to the inherent nature of 
resources (apart from representations).

Patrick


> 
>       -- sandro
> 
> [1] 
http://www.w3.org/TR/2004/WD-webarch-20040816/#information-resource

Received on Thursday, 9 September 2004 09:15:27 UTC