Re: namespace as resource

[Devon Smith]

>
> an example of the potential problem:
>
> <http://namespace.com> dc:creator "devon" .
> <http://namespace.com> dc:creator "mike" .
>
> i created the namespace. mike created the page.
> no one could tell from those two statements, however.
>

That's because you have supplied incomplete information, but it's so
human-understandable that we don't notice that it's incomplete.  These two
statements are really shorthand for something like this:

subject        predicate                           object
ns1          urlSpec                   "http://namespace.com"
ns2          pageAtURL             "http://namespace.com"
ns1         dc:creator                          "devon"
ns2         dc:creator                          "mike"

Devon (and Sean Palmer) is right on target talking about the distinction
between the url and the page, but the issue here is really about how you
specify or constrain or identify resources, and how much information you can
omit.  It's easy, when you hand-craft examples, to make them look like they
are machine-understandable (e.g., by using rdf) while in reality they are
the same old human-only kinds of statements we all (usually) understand but
machines won't.

Isn't this what makes it interesting?

Cheers,

Tom P

Received on Saturday, 18 August 2001 01:52:53 UTC