Question regarding DR 09-0216

[Doug Mahugh]

Alex,

In the approach we've used for implementing this particular feature, "can" is indeed correct.  The rationale behind that flexibility is that the user may be unaware that the custom XML markup exists, and they could make a change to a document that renders it invalid, by simply deleting a paragraph for example.  In that case, the user may not know what needs to be done to make their change in a schema-valid way, so they have no ability to correct the error if schema validation against the custom schema were required.  So we allow them to save the document anyway, preserving the user's intended content even if it is no longer consistent with the custom schema.

I think it's worth making a distinction between the requirements of dedicated XML editors (whose primary role is help users create instances of specific schemas), and word-processing applications (whose primary role is helping users create documents with specific content, structure and layout).  The customXML element in IS29500 is a lightweight optional mechanism for adding custom semantics to word-processing documents, and the attachedSchema itself is optional.

- Doug


-----Original Message-----
From: Alex Brown [mailto:alexb at griffinbrown.co.uk]
Sent: Tuesday, June 30, 2009 1:06 AM
To: Innovimax SARL; Rex Jaeschke
Cc: SC 34 WG4
Subject: RE: Question regarding DR 09-0216

Dear all,

I too would like to see the questions answered.

On Mohamed's point, ISO wording (not RFC) wording is in force, so "can" is fine. But do we really mean "can". What if the schema makes modifications to the instance; in that case wouldn't different XML be implied depending on whether or not validation took place?

- Alex.

> -----Original Message-----
> From: Innovimax SARL [mailto:innovimax at gmail.com]
> Sent: 30 June 2009 07:42
> To: Rex Jaeschke
> Cc: SC 34 WG4
> Subject: Re: Question regarding DR 09-0216
>
> Rex,
>
> Correct me if I'm wrong but the objection Murata and I did are not
> incorporated.
>
> (I'm just restating this because Shawn answer disturbed me a bit...)
>
> Mohamed
>
> On Mon, Jun 29, 2009 at 9:57 PM, Rex Jaeschke<rex at rexjaeschke.com>
> wrote:
> > I show email traffic (see below) based on my proposed response, but
> no
> > record of agreement on the final words after that. Did I miss
> something, or
> > do we still need final words?
> >
> > Rex
> >
> > 1.      DR 09-0216 — WML: Custom XML and Smart Tags
> >
> > Status: Closed; will be incorporated in COR1
> >
> > Subject: WML: Custom XML and Smart Tags
> >
> > Qualifier: Request for clarification
> >
> > Submitter: Mr. Francis Cave (BSI)
> >
> > Contact Information: francis at franciscave.com
> >
> > Submitter’s Defect Number: 08-00131
> >
> > Supporting Document(s): none
> >
> > Date Circulated by Secretariat: 2009-05-22
> >
> > Deadline for Response from Editor: 2009-07-22
> >
> > IS 29500 Reference(s): Part 1: §17.5.1, “Custom XML and Smart Tags”,
> p. 529
> >
> > Related DR(s): none
> >
> > Nature of the Defect:
> >
> > The second para on p. 529 following the bullets has: "The distinction
> > between custom XML markup and smart tags is that custom XML markup is
> based
> > on a specified schema."
> >
> > It is not clear how "a specified schema" is specified in this
> context. Can
> > there only be one specified schema per document?
> >
> > Solution Proposed by the Submitter:
> >
> > Point to normative text describing how one or more schemas are
> specified, or
> > - if this does not exist - provide new text.
> >
> > Schema Change(s) Needed:
> >
> > Editor’s Response:
> >
> > The exact changes are as follows:
> >
> > Part 1: §17.5, “Custom XML and Smart Tags”, p. 529
> >
> > The distinction between custom XML markup and smart tags is that
> custom XML
> > markup is based on a specified schema, which shall be specified using
> the
> > attachedSchema element (§17.15.1.5). As a result, the custom XML
> elements
> > can be validated against the schema. Also, as shown below, custom XML
> markup
> > can be used at the block-level as well as on the inline (run) level.
> >
> > 2009-06-11 Makoto Murata:
> >
> >>  which shall be specified using the attachedSchema element
> >
> > Is this a recommendation or a requirement?  In other words, is the
> > attachedSchema element authoritative?
> >
> > 2009-06-11 Mohamed Zergaoui:
> >
> > I was also wondering why it is used "CAN" which is not RFC compliant.
> >
> > I would also go for a "MAY" ("XML elements MAY be validated") and
> would also
> > add ("but MUST be valid with respect to the attachedSchema").
>
>
>
> --
> Innovimax SARL
> Consulting, Training & XML Development
> 9, impasse des Orteaux
> 75020 Paris
> Tel : +33 9 52 475787
> Fax : +33 1 4356 1746
> http://www.innovimax.fr
> RCS Paris 488.018.631
> SARL au capital de 10.000 €