RDDL does not help identify; it helps describe and in a highly
context specific way
Peter
From: Christopher B Ferris
[mailto:chrisfer@us.ibm.com]
Sent: 18 June 2007 17:42
To: Peter F Brown
Cc: oasis-member-discuss@lists.oasis-open.org;
robin.cover@oasis-open.org
Subject: RE: [oasis-member-discuss] OASIS namespace documents for URN
namespaces
Peter, you
wrote:
> I would argue, always have argued, that
you shouldn't find anything
> at the end of a URI that purports to be an identifier except, well,
> something that helps identify the artefact being referenced.
If a RDDL isn't
someting that helps identify the artefact identified, when that artefact is a
namespace,
then I don't
know what it is.
A namespace is
an abstraction. It has an identifier. A namespace is NOT a schema document. I
might
have multiple
'schema' that each are legitimate means of validating a document that includes
elements and
attributes for a given namespace (e.g. XML Schema, RNG, Schematron, DTD, etc.)
They may be
isomorphic or complimentary in nature.
Having a
namespace document (RDDL) available as the resource retrieved when you
dereference
the namespace
URI helps 'identify the artefact' that is referenced by that URI. In the case
of the
namespace
documents that have been designed for OASIS specs, the informative value is
increased by
virtue of the
fact that the specifications which define the semantics of the namespace
components
are also linked
and the namespace versioning policy is provided.
Cheers,
Christopher
Ferris
STSM, Software Group Standards Strategy
email: chrisfer@us.ibm.com
blog: http://www.ibm.com/developerworks/blogs/page/chrisferris
phone: +1 508 234 2986
"Peter F Brown"
<peter@pensive.eu> wrote on 06/18/2007 10:46:11 AM:
> Of course it all depends on what you want to use the identifier for.
> I would argue, always have argued, that you shouldn't find anything
> at the end of a URI that purports to be an identifier except, well,
> something that helps identify the artefact being referenced. An RDDL
> document is not an identifier - it is a resource descriptor; it is
> intended to provide information about the identified artifact but, I
> would argue, is not the only thing that could be relevant at such an
> end point. As such there *should* be a distinction between the
> namespace urn and any mechanism to dereference it in a particular
> context for a particular purpose - whether OASIS provides it os not,
> is another question.
>
> The Topic Map community understands this intimately in their
> distinction between a "subject identifier" and a
"subject indicator"
> and didn't get lost down the W3C blind alley of confusing labels and
> resources à la RDF... sorry, shouldn't have dragged that up ;-)
>
> Peter
>
> ==================
> From: Christopher B Ferris [mailto:chrisfer@us.ibm.com]
> Sent: 18 June 2007 16:03
> To: Peter F Brown
> Cc: oasis-member-discuss@lists.oasis-open.org;
robin.cover@oasis-open.org
> Subject: RE: [oasis-member-discuss] OASIS namespace documents for
> URN namespaces
>
>
> The OASIS TAB has published guidelines for namespace URIs (amongst
> other subjects related to
> metadata) here:
> http://docs.oasis-open.
> org/specGuidelines/namingGuidelines/resourceNaming.html
>
> These guidelines include recommendation that namespace URIs be HTTP
> scheme URIs and that there
> be a published RDDL document at the namespace URI location. URNs are
> permitted for use, but HTTP
> scheme URIs are preferred/recommended since they can be resolved
> without additional machinery. Note
> that OASIS hs no plans to implement such a resolution mechanism:
> http://docs.oasis-open.
> org/specGuidelines/namingGuidelines/resourceNamingCommentaryV07.
> html#urnResolution
>
> An example of these guidelines in action can be found with the WS-RX
> TC's WS-RM namespace URI:
>
http://docs.oasis-open.org/ws-rx/wsrm/200702
>
> Frankly, I don't personally see any material benefit to use of a URN
> scheme namespace. I suspect that
> some believe that by using a URN, they have created a 'permanent
> identifier'. Well, frankly, there is no
> reason why an HTTP scheme URI cannot be 'permanent'. It isn't as if
> OASIS is going to drop off the face of
> the earth anytime soon.
>
> Cheers,
>
> Christopher Ferris
> STSM, Software Group Standards Strategy
> email: chrisfer@us.ibm.com
> blog: http://www.ibm.com/developerworks/blogs/page/chrisferris
> phone: +1 508 234 2986
>
> "Peter F Brown" <peter@pensive.eu> wrote on 06/18/2007
09:43:24 AM:
>
> > Paul:
> > I think the objective is laudable but I don't see the point of
the
> > "http://www.oasis-open.org/namespace?ns=" part of the
statement. Surely,
> > if some application wants to do something with the
"pure" (and on its
> > own non-deferenceable) urn, then surely it will ba equipped to
make that
> > call using http get or whatever.
> >
> > By effectively tying a namespace to a derefereceable network
endpoint
> > and to specific and prescribed dereferencing mechanism, is there
not a
> > danger that you:
> > - either limit what can be legitimately found at the network
endpoint to
> > an RDDL file only, cutting out other possible uses of the
namespace;
> > - or creating false expectations and possible architectural
problems
> > when unpredictable results come from finding something other than
a RDDL
> > document at the end of the pipe.
> >
> > On the other hand, I'd be in favour of a mechanism by which RDDL
> > documents, when available, can be accessed using a prescribed
OASIS
> > managed URL to which you would tack the namespace urn.
> >
> > Regards,
> >
> > Peter
> >
> > -------------
> > Peter F Brown
> > Chair, CEN eGovernment Focus Group
> > Founder, Pensive.eu
> > Co-Editor, OASIS SOA Reference Model
> > Lecturer at XML Summer School
> > ---
> > Personal:
> > +43 676 610 0250
> > http://public.xdi.org/=Peter.Brown
> > www.XMLbyStealth.net
> > www.xmlsummerschool.com
> >
> >
> > -----Original Message-----
> > From: Paul Denning [mailto:pauld@mitre.org]
> > Sent: 30 May 2007 18:45
> > To: oasis-member-discuss@lists.oasis-open.org
> > Subject: [oasis-member-discuss] OASIS namespace documents for URN
> > namespaces
> >
> > Some OASIS TC's use an HTTP URI when they define an XML
> > Namespace. Some TC's have started using RDDL as a namespace
document
> > when these namespace URI's are dereferenced using HTTP GET.
> >
> > Other OASIS TC's use URN's to identify their XML Namespace.
> >
> > I suggest that OASIS provide a mechanism to resolve a namespace
> > document (RDDL or other) using HTTP and passing in the namespace
URN
> > in a query parameter.
> >
> > For example,
> >
> >
http://www.oasis-open.org/namespace?ns=urn:oasis:names:tc:emergency:EDXL
> > :DE:1.0
> >
> > would return a RDDL document associated with the EDXL-DE
namespace.
> >
> >
> > Paul
> >
> >