RE: What do the ontologists want

On Sat, 19 May 2001, pat hayes wrote:

> >
> >: pat hayes:
> >:I am still trying to
> >: find out what 'resource' means, but Dan Connolly tells me that:
> >: the standard [definition of resource] is RFC2396:
> >
> >In RDF, a resource is something identified by a URI (that may have
> >anchor ids) as per rfc2396. That's all there is to it.
>
> Oh. The trouble is, the reason I wanted to find out what resources
> were was in part so that I could find out what URIs were. URLs I
> understand - they are a kind of global file-name -  but the W3C folks
> seem to think that URIs are something much more comprehensive than
> URLs: they *seem* to be saying that anything in the universe that can

"some W3C folks", please. There are a few of us, we all seem to be URI
enthusiasts but ascribing a common technical view to the lot of us is a
bit premature.

> be referred to by any language can be indicated by a URI, so that if
> I want to talk about the electron density of the Oort cloud, or a
> grain of sand on Pensacola beach, well then I just use a URI.  (I
> don't know quite HOW to do this, but I'm willing to learn.) So that
> means that *anything* can be identifed by a URI, so *anything* is a
> resource. People who died five centuries ago are resources, leptons
> are resources, sets of integers are resources, Unicorns are
> resources, Father Christmas is a resource.

Yes

> The only way I can make sense of this, I confess, is to think of URIs
> simply as names, and resources simply as entities. The trouble with
> *this* interpretation, however, is that it makes nonsense of the hype
> about the significance of URI's and how they are some new idea that
> needs an RFC standard to define them and will bring some new Web
> Power. They seem just like, well, common old names for things, like
> nouns in English or logical constants in logic. Never mind the 19th
> century: names in this sense probably go back to the cro-magnon era,
> or maybe before. Chimpanzees and gorillas can use names like this; so
> what is all the fuss about? Which makes me think that this can't
> really be what is meant by a URI. But I still don't know what else
> there is to the notion.
>

Yes, URIs are just names. Names in a certain syntax and with some controls
over who gets to dish them out. URLs are too, though URLs tend to encode
information about how to get at the named thingy using networked
computers. URLs aren't simply file names: they've long been used to name
Web services (eg CGI scripts etc), internet mailboxes, format and language
negotiated resouces and various other kinds of thing that aren't commonly
considered 'files'. URI (and URL) syntax is specified in RFC2396.

URIs, sadly, aren't magic at all. Just useful. URIs are handy
(particularly in electronic publishing, web services etc) since we've an
agreed convention for devolving names to organisations and individuals
for use. Their use makes it easier to do database joins, at least for
descriptions of things that have widely agreed URI names (like Web pages,
mailboxes etc). This in turn makes it a little bit easier to talk about
things that don't have widely understood Web names, eg. we can talk about
you in DAML+OIL through using various URIs (phone number, home page,
mailbox) to pick you out. The common URI syntax just reduces the number of
different ways of writing down your phone number etc, making it a bit
easier for databases to join up scattered descriptions of you.

> >I find it's
> >useful way to think when it comes to implementing code.  That may
> >seem a backways determination; if I create a URI do I create a
> >resource for it to identify? This is moot, the RDF machine can't
> >access a resource directly anyway, but it allows for the description
> >of say, unicorns.
>
> It does? How?

(process point: if we're going to have a discussion about identification
and description of non-existent or hypothetical entities, please start a
new thread.)

Dan

Received on Saturday, 19 May 2001 18:02:22 UTC