Re: The 'resource' identified by a namespace name URI should be the namespace

On Fri, Jun 02, 2000 at 09:16:10AM +0100, David Carlisle wrote:
> > I still don't get the distinction. Why do you need "the URI of a namespace"
> > if the namespace is already named by a URI (assuming we semi-deprecate
> > relative URIs)? 
> 
> You don't for namespace processing (which only uses the namespace
> name) but some systems (notably rdf) require a URI for any resource
> that they want to reason about, and some people claim a need to reason
> about namespaces in this way.

But they don't need to resolve the URI in order to be able to 
reason about it. They just need the URI to act like a URI...

> namespaces with an absolute URI as namespace name have names that are
> unique (that is no other namespace should have the same name) but
> that doesn't mean that rdf can really use that URI to reason about the
> namespace as it may already be reasoning about the resource which is
> identified by the URI.

But that hasmore to do with the choice someone is making about the
URI they use to name the namespace than whether or not they should
use it at all...

> mailto:david@dcarlisle.demon.co.uk identifies the resource which is
> the mail to my home account.

And if you use that URI to name your namespace that is perfectly valid.
Your just saying that the namespace is named by something that
also identfies your mailbox. No problems there... A name doesn't
get to define how or why it gets used, just what it names and how long
you get to trust that. (plus I would claim that using an email
address as a namespace name violates the persistence requirement
that the Rec puts on those names).

> John's proposal was that there should be a URI that identifies the
> namespace used in the following document
> <x xmlns="mailto:david@dcarlisle.demon.co.uk"/>
> Note that the namespace name can not function in this role, even
> though it is an absolute URI, as the R it I isn't the namespace.

Ahh.... You think that just because the R happens to be a mailbox
that the I can't be used as a name. That isn't true. The R here
is the mailbox, sure. But that doesn't mean you can't say that
the namespace is named by the I. Now, if some application comes along
and wants to try and resolve that to something it will get teh
equivalent of "You have to send this guy email and ask him what 
this namespace means". Which, IMHO, is perfectly reasonable...

Not every namespace name has to be resolvable. It just has to name
according to the URI specification. If something wants to resolve
that URI to something then it can try. What has to happen is that
the namespace author has to decide whether or not he wants to
allow that application to succeed in its attempt to resolve it.
I think resolution is important so I'll use a URI that can
resolve to something useful. If you don't think that's important
then you can just use your email address (which isn't persistent).

-MM

-- 
--------------------------------------------------------------------------------
Michael Mealling	|      Vote Libertarian!       | www.rwhois.net/michael
Sr. Research Engineer   |   www.ga.lp.org/gwinnett     | ICQ#:         14198821
Network Solutions	|          www.lp.org          |  michaelm@netsol.com

Received on Friday, 2 June 2000 12:22:56 UTC