Re: #rdfms-difference-between-ID-and-about (every document is in the Web)

On Fri, 15 Jun 2001, Dan Connolly wrote:

> Brian McBride wrote:
> >
> > Dan Connolly wrote:
> >
> > > You trust that I have a birthday even though you don't
> > > know it, right? By the same token, it seems easy
> > > enough to accept that resources have URIs even though
> > > those URIs aren't always specified.
> >
> > Forgive me butting in.  Is the set of real numbers
> > a subset of the set of resources?  After all, they have identity.
> > Note that URI's are countable and real numbers are not.
>
> That's the subtlety I meant to set aside when I wrote:
>
> | I think you've slightly overstated the case there,
> | but the argument holds even the way you've phrased it, so...
>  -- Fri, 15 Jun 2001 09:20:55 -0500
>
> The way he phrased it, every resource has a URI-name.
> The way I see it, any particular resource can be named
> with a URI; that doesn't mean they all have (unambiguous) names,
> as you point out: all real numbers are resources;
> only denumerably many of them can be named with URIs.
>
> Another subtlety that isn't relevant to the ID/about
> issue is whether 'resource' is used in the general
> sense of 'anything in the domain of discourse'
> (i.e. things you can refer to using absolute-URIs-with-fragments)
> or in the stricter sense of, roughly, 'something you can
> get at via the network' (i.e. things you
> can refer to using absolute-URIs-without-fragments).

I've never seen a clear definition of this 'stricter' sense. The phrase
'something you can get at via the network' doesn't seem even to be a rough
approximation: we have URIs for all kinds of things, isbn:, tel:, urn:,
java:, etc. I'm not sure why we would want to define our notion of
'resource' with special reference to a computer network. Is the intuition
is something like 'those things in the domain that can be named with RFC2396
identifier strings'? Is there anything more we can say about those things?


> I'd like for the revised RDF specs to use 'resource'
> in the stricter sense, but in speaking of real

Can you propose a definition? I agree that there is a need to
make some distinction here, or at least to couch the specs in a way that
more carefully distinguishes an object from its name(s).

> numbers as resources, of course we're using it
> in the general sense.
>
> The traditional logic-literature
> term for the more general concept is 'object'; I'd
> be happy to start using that in these WG discussions.

'Object Description Framework'? Catchy :)

--danbri

Received on Friday, 15 June 2001 15:15:38 UTC