Namespace URI address resources

[Dan Brickley]

On Thu, 17 Jun 1999, Jonathan Borden wrote:

> Murray Maloney wrote:
> >At 02:46 PM 6/16/99 +0200, Ronald Bourret wrote:
> >>My apologies for the sarcasm, but although I find the use of namespace
> URIs
> >>to find schemas a wonderful theoretical idea, I'm having more than a
> little
> >>trouble seeing how it could possibly work in practice.
> >>
> >Come work with Commerce One and we'll show you.
> >
>     There is no doubt in my mind that using the namespace URI can work in
> many circumstances to locate a schema. There is also no doubt in my mind
> that a PI containing an href can also be used to locate a schema. The issue
> ought not be what has been done in pre-spec implementations rather what is
> the best way to do this for the future. The problem with using the namespace
> URI is that
> 
> 1) it identifies a single resource

...which might have multiple content-negotiable manifestations (each of 
which is a resource too), much as normal Web resources can manifest
themselves in HTML,PDF,GIF,PNG (or, closer to the point) in different
lanuages (fr,en etc) as a normal part of the dereferencing business.

> 2) the location and protocol used to access the resource is hardcoded into

'a' not 'the'; possiblity of other access mechanisms for getting at 
(content/language negotiated bytestream manifestations of the) resource
isn't ruled out. And is only hardcoded with certain URI schemes.

> the namespace and hence cannot be changed without causing fundamental
> changes to the structure of the entire document.



> 3) mechanisms to resolve URIs which are location and protocol indenpendent
> (e.g. "urn:xxx") are not widely available and hence cannot be used in
> practice.

This is just more (meta)data, and XML is as good a mechanism as any for
shipping about such stuff. We don't need to wait for global
infrastructure before exchanging such modest facts as 'uri:xxx:doc444'
-manifestedAt-> 'ftp://docs.org/doc444'. 

eg:
<abc:CV xmlns:abc="urn:x-xxx:schemata/abc">
[...instance data using the 'abc' vocab here...]
</abc:CV>

[accompanied perhaps by data about resolving the URI naming 'abc'...]

	<uri:resolveinfo abstract="urn:x-xxx:schemeta/abc" 
			 uri="http://yetanotherindustryregistry.org/abc.xml"/> 
 


Sure, the terminology to exchange this resolution needs to be nailed
down, but the point is that what we lack is not a URN resolution
_mechanism_ being widely deployed, but a bit of terminology for thinking
how to exchange this URN2URL-ish information. Which by contrast doesn't
seem so hard...

Dan


xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev at ic.ac.uk
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/ and on CD-ROM/ISBN 981-02-3594-1
To (un)subscribe, mailto:majordomo at ic.ac.uk the following message;
(un)subscribe xml-dev
To subscribe to the digests, mailto:majordomo at ic.ac.uk the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa at ic.ac.uk)