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

RE: Literals as Resources (was RE: draft partitioning of the issues)

From: Graham Klyne <Graham.Klyne@Baltimore.com>
Date: Tue, 03 Jul 2001 14:15:22 +0100
Message-Id: <5.0.2.1.2.20010703131252.03518440@joy.songbird.com>
To: "Ron Daniel" <rdaniel@interwoven.com>
Cc: "Aaron Swartz" <me@aaronsw.com>, "rdf core" <w3c-rdfcore-wg@w3.org>
Having some sympathy with both sides of this, I'm going to play devil's 
advocate for now:

At 01:34 AM 6/29/01 -0700, Ron Daniel wrote:
>I have already cited section 2.1 of the spec, which says:
>
>     The object of a statement (i.e., the property value) can
>     be another resource or it can be a literal; [...]
>
>This seems pretty clear.

It does suggest that literals have a special place in the scheme of things.

I think that even if all literals were to be regarded as data: URIs, they's 
still have a special place, as they are the only way to attach some 
specified data to an RDF graph without reference to some external data.

>But if you disagree, is there a
>particular standard of proof for supporting evidence you
>would have to see before acknowledging that the 1.0 M&S
>considers the two to be distinct?  How about:
>
>1) A member of the original working groups (me) has stated
>    that their recollection was that the two were considered
>    distinct. (That is a pretty strong recollection BTW).

That does support your view.

>2) The spec has many examples of simple quoted strings as
>    values, or URI references as values. But it has no examples
>    of data URLs or any examples showing how a literal is
>    mapped into a Resource or vice versa.
>
>3) All RDF parsers and APIs I am familiar with treat literals
>    as distinct from resources, which is a pretty solid
>    verdict from implementation experience.
>
>4) Art Barstow asked the WG if anyone was familiar with
>    implementations that did not treat them as distinct.
>    Nobody has replied in the affirmative, although Jan
>    Grant said that he could change his code to make it so.
>    But the point is that this would be a change.

I think all of these can be interpreted as recognizing that literals, in 
some sense, have a distinguished place in and representation in RDF without 
actually saying that a literal is not a resource.

Thus, considering Brian's discussion of scope:

[Brian McBride:]
>....  The question is not whether a
>literal is a resource, for to answer that we need to resolve a bunch
>of difficult issues around what resources are.  The question is
>whether the abstract model described in m&s treats literals specially.
>If it does, then so must we if we are to avoid 'reformulating' the model.

It seems clear to me that there is some special treatment, but it is not 
clear Literals cannot be interpreted as a subset of Resources.

#g
--


------------------------------------------------------------
Graham Klyne                    Baltimore Technologies
Strategic Research              Content Security Group
<Graham.Klyne@Baltimore.com>    <http://www.mimesweeper.com>
                                 <http://www.baltimore.com>
------------------------------------------------------------
Received on Tuesday, 3 July 2001 09:34:43 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 3 September 2003 09:38:06 EDT