OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

chairs message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: Re: [chairs] Specification Namespaces? Standard XML Schema Validation Service


Rex,

I'm not convinced.  As we've learned with both the CAM and RELAX NG work
such validation with XSD is tenuous at best.  If we were considering
something -
a semantic service based on registry would offer far more robust support -
including
significant version control.

Furthermore - using the technology that CAM provides frees you from being a
slave to namespaces - and thereby shackling the easy reuse of content in
their XML
instances by your customers.  Not to mention that many parsing engines and
systems
handle multiple namespace declarations in unpredictable ways causing
interoperability
issues.

The unfettered use of namespaces across XSDs of transaction sets being
developed
by OASIS would IMHO be a disasterous mistake that will cripple long term
possibilities of providing interoperable dictionaries that operate at the
noun
semantic level - let alone being able to leverage coming technology - such
as
OWL + semantic registry - as was very very ably shown yesterday by the
presentation from Ankara University on its work.

 See PPT here:

  http://www.srdc.metu.edu.tr/~asuman/RIDE04_Talk.ppt

Namespace I would urge should be used with extreme caution and in a
minimilistic way.  Paradoxically CAM itself makes one single use of
namespace to differentiate embedded attribute and predicate usage,
that needs to be filtered separate from the rest of an XML instance itself.
Otherwise - judicious use of XPath can easily differentiate the context
and usage of fields in all but the most arcane syntax instances.

BTW - we are using this declaration pattern for the namespace which
I would suggest is adequate - if you absolutely must have a namespace
for your purposes:

  xmlns:as="http://www.oasis-open.org/committees/cam";

Thanks, DW.

----- Original Message ----- 
From: "Rex Brooks" <rexb@starbourne.com>
To: <jamie.clark@oasis-open.org>; <karl.best@oasis-open.org>;
<chairs@lists.oasis-open.org>
Cc: <chiusano_joseph@bah.com>
Sent: Wednesday, March 31, 2004 9:11 AM
Subject: [chairs] Specification Namespaces? Standard XML Schema Validation
Service


> Hi Jamie, Karl, fellow chairs,
>
> I just reviewed the TC Process and TC Guide and discovered that there
> is now no mention of OASIS providing a namespace for standards and
> specifications. I do not recall seeing in the brief notes about
> previous changes to the TC Process that this was removed. I admit I
> probably just missed it, but since I have followed the chairs
> discussion about document management and the document repository, and
> I have also made an effort to follow TC Process changes when I
> haven't been completely swamped with work, I am surprised I did not
> notice mention of this during the explorations of the document
> repostory question or in the TC Process changes.
>
> Have namespaces been rolled into the new Standards Registry
> Repository? I mention this because it has always seemed like an
> obvious solution, and I assumed such would be the case eventually,
> making the URN for OASIS with a specific namespace location in the
> either the TC's document repository or the standards repository
> (which I would reserve for a family of specs such as legalXML,
> TopicMaps, ebXML, etc.)
>
> The reason I am asking is that a namespace is required to validate a
> schema, which brings me to the last part of this message.
>
> Can we establish a standard validation service for all OASIS
> specification schemata, and require validation using this validation
> service for a Committee Draft (or Committee Specification if we
> change it back to that formulation to avoid confusion) and for
> Submission of a Committee Draft to OASIS for approval of a candidate
> standard?
>
> I have several reasons for asking these questions, like all of the
> TCs in which I actively participate and all of the TCs and SCs in
> which I am an observer.
>
> Ciao,
> Rex
> -- 
> Rex Brooks
> GeoAddress: 1361-A Addison, Berkeley, CA, 94702 USA, Earth
> W3Address: http://www.starbourne.com
> Email: rexb@starbourne.com
> Tel: 510-849-2309
> Fax: By Request
>



[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]