DRAFT: Response to SC 34's request to use the processing instruction target name token "xml-model"

Below is my draft response.  Once I started writing
it, I thought of some other issues (e.g., general
availability of the xml-model spec), so I threw
them in.  Comments?

(Mohamed, others, any better suggestions for how best 
to address Toshiko Kimura?  He addressed the email to
me "Mr. Grosso".)

(As another aside, I added kimura@itscj.ipsj.or.jp to
public-xml-core-wg@w3.org's accept2 list.)

paul

------------------------------------------------

Eventually-To: Toshiko Kimura [mailto:kimura@itscj.ipsj.or.jp] 
Eventually-Cc: Alex Brown; mzergaoui@innovimax.fr; jirka@kosek.cz;
liam@w3.org

Dear Kimura-san,

Thank you for your request regarding SC 34's desire to 
use the processing instruction target name token "xml-model" 
in Part 11 of ISO/IEC 19757 (DSDL). 

As you are aware, the XML Recommendation reserves names
beginning with 'xml' for standardization in this or future 
versions of this specification.  The XML Core WG feels
any such names would need to be used by other specifications
only in carefully specified, well-defined ways.  We are
therefore pleased you have brought us into the discussion
at the beginning of the process.

In general, we are receptive to your request, but at this
time we do not have a lot of details, so we are eager to
exchange ideas so that we can work together to ensure the
best result.  If we are going to allow the use of the
xml-model name, we would want an xml-model PI to be
generic enough to be used by any application to associate 
any kind of schema with an XML document.  The XML Core WG 
would like to be able to review drafts of the xml-model spec 
throughout its development.

We assume that the general syntax and semantics for most 
of the pseudo-attributes for the xml-model processing 
instruction will parallel that as defined in the Associating 
Stylesheet (AssocSS) spec that defines the xml-stylesheet 
processing instruction.  You should know that we are in the 
process of developing a new edition of the AssocSS spec, so 
you should be sure to coordinate your specification with our 
latest version.  (The additions are mostly clarifications
on error or other edge conditions.)

We assume you will be writing a specification for xml-model
that will be referenceable by everyone on the Web free of charge
so that anyone wishing to support the xml-model PI for any
reason--not just for DSDL--will have an official reference for 
it.  If this is not practical for you, then we would want to be 
able to republish the xml-model PI spec as a W3C publication 
(e.g., a WG Note or a W3C Recommendation).

One detail the xml-model spec would have to address is the
interaction of the xml-model PI with xsi:schemaLocation and
similar attributes.

The XML Core WG looks forward to working together with SC 34
to achieve a satisfactory result for all concerned.

Best regards,

Paul Grosso
for the XML Core WG

Received on Wednesday, 22 April 2009 19:00:33 UTC