RE: argument from authority considered pointless

Thanks Roy.  I appreciate the response.  Maybe Sandro should answer this 
because I think the problem is the formalism requires the application 
to perform an action (identification) with the URI and that KR assumes 
a one to one mapping of name to identified object, not a potential n-set:

From the cited source (2396):

"...it allows the identifiers to be reused in many
    different contexts"

Apparently, not quite.  RDFers say it doesn't.  Again, it seems 
to depend on the use of a context.

"...An identifier is an object that can act as a reference to
    something that has identity."

I still believe the problem here is that entities don't 
have "identity".  They are as said in the next sentence, 
"identified" such that conceptual mapping and identification 
are synonymous.  I understand the use of the term though.

"...Having identified a resource, a system may perform a variety of
   operations on the resource..."

Ok except now

"The resource is the conceptual mapping to an entity or set of
 entities, not necessarily the entity which corresponds to that
 mapping at any particular instance in time.  Thus, a resource
 can remain constant even when its content---the entities to
 which it currently corresponds---changes over time, provided
 that the conceptual mapping is not changed in the process."

Sandro says (Re: URI Opacity Principle (was: Re: use of fragments as names is irresponsible):

"...that's not how KR languages are generally defined.  "An
interpretation must specify which object in the world is referred to
by each constant symbol."  [IAMA p186].  To say instead that "An
interpretation must specify which objects in the world are referred to
by each constant symbol," ... 

So the KR lanugage does not provide a conceptual mapping 
to select an entity or set of entities using the name/identifier?

So where RDF uses the syntax of the URI, they don't fulfill the 
obligation to map.  Is that it?

len

-----Original Message-----
From: Roy T. Fielding [mailto:fielding@apache.org]

> Ok, put the challenge back in the other
> court.  Roy, if the credential for discussing
> the issue is knowledge of or creation of libWWW,
> present the code in libWWW that declares and
> defines a "resource".

What was being discussed is whether the Web architecture is or is
not specific to HTTP.  The answer was that it is not, as evident in
the libwww and libwww-perl.  End of topic.  We already established
that 2396 defines resource and what that means for URIs, and any
further discussion of that issue in this forum has been deferred.

Received on Wednesday, 22 January 2003 18:19:29 UTC