Lists Home |
Date Index |
Dare Obasanjo wrote:
> I know this may sound like a chicken and egg question but is the reason peopl
> expect something to be at the URI because most examples of namespace usage by
> the W3C and others involve HTTP URLs?
My guess: a "URI" is (or was originally) a generalized term
for "URLs" (resource locators), "URNs" (resource names), and
"URCs" (resource characteristics). "URCs" never took off because
nobody could figure out what they should be, and there has
been resistance to using "URNs" because, ironically, they
can't be resolved. (This is ironic because the chief distinguishing
feature between Names and Locators is that the former are
not *meant* to be dereferenced.)
So common practice has been to use URLs even for things where URNs
would be more appropriate (XML namespace names, RDF resources, etc.).
Since the principle purpose of a URL is to *locate* a *resource*,
it's not suprising that people expect to find something at the
other end of one.
Opinions differ on this of course. In the Semantic Web world
there is no distinction between the resource itself and the
thing used to Identify it; the idea there seems to be that
everything important can be done simply by expressing relations
among resources without anybody ever actually retrieving and