W3C home > Mailing lists > Public > w3c-rdfcore-wg@w3.org > July 2001

RE: literals as resources

From: pat hayes <phayes@ai.uwf.edu>
Date: Fri, 6 Jul 2001 12:42:59 -0700
Message-Id: <v0421010cb76bc2dfb37f@[130.107.66.237]>
To: <prichard@brience.com>
Cc: w3c-rdfcore-wg@w3.org
>To clarify my position.
>- A RDF-literal is a resource, in the common sense of the term "resource".

Well, you know, there IS no common sense of this term. It was used 
for the first time by Tim B-L about 5 years ago  and has never been 
formally defined. But after extensive enquiries, I have been told 
that the recieved view is that 'resource' is pretty much the same as 
'entity' or 'thing', ie anything whatsoever than can possibly be 
referred to.
So with that understanding:

>- It is not a RDF-resource as defined by the M&S, because a RDF-literal
>has no URI.

that is irrelevant; the issue is whether it could possibly be given a 
URI; and the answer to that question is, yes, it could, since 
anything could.

>Seen in other computing worlds:
>- A java String is an Object, a java char[] is a primitve data type.
>- An <!ENTITY> can be internal (a literal), external: SYSTEM, PUBLIC,
>or even NDATA with a Notation attached
>
>Java or XML could have get rid of (respectively) char[], internal entities.
>They didn't.  And this brings me to ask the Question and my own answer:
>
>Q: does RDF need "literal" e.g. a primitive (non-RDF)resource type?
>(My) A: YES.  Because prooved by Computer History (aka Axiom.)
>If one needs something more elaborate, one is able to use a data URI
>scheme, or some appropriate protocol.  RDF does say:
>"The object of a statement can be another resource or a literal"
>I can do what I want.  I'm happy.
>
>The term "resource" is just too imprecise.  Should it be RDF-resource,
>U-Resource (Universal resource, accessible by the Universe) or
>resource in italics?
>In brief, I don't see anything wrong in the current M&S that precisely
>defines what a RDF-resource is.  The term is just confusing.

Amen to that.

>Or shall we say "a literal is a primitive resource, and, in this sense,
>doesn't belong to what RDF names "resource."
>
>Again, the real question is: do we have the need for "vanilia strings"?

I don't think anyone (? certainly not me) wants to eliminate literals 
from RDF. They are obviously useful. But I don't think that was the 
issue we were discussing. Similarly, I don't want to translate or 
automagically morph literals into some kind of URI. And again, I 
don't think we were discussing that. The point was only that there is 
seems to be a contradiction in the spec. On the one hand, it seems to 
say, or strongly imply, that literals are not resources; but 
everything is a resource; which is a contradiction. Our primary goal 
is to clean up the spec, and this seems like an obvious thing that we 
can clean up. While we are at it, we can do a bit of spit and polish 
on 'resource' as well.

Pat Hayes

---------------------------------------------------------------------
IHMC					(850)434 8903   home
40 South Alcaniz St.			(850)202 4416   office
Pensacola,  FL 32501			(850)202 4440   fax
phayes@ai.uwf.edu 
http://www.coginst.uwf.edu/~phayes
Received on Friday, 6 July 2001 15:42:59 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 3 September 2003 09:38:07 EDT