HTML validation and extensibility, update [tagSoupIntegration-54 ISSUE-54] [ISSUE-33 mixedUIXMLNamespace-33]

I made progress on this action this week...

ACTION-7 Dan Connolly to Work with Olivier and Tim to draft a position
regarding extensibility of HTML and the role of the validator for
consideration by the TAG

I talked with TimBL a few weeks ago and Olivier, Chris Lilley,
and Doug Schepers this past week. One day the planets will
line up and I'll talk to both TimBL and Olivier at the same
time. Meanwhile, we have email...

Olivier had asked if opensp, the DTD validation software
used by the W3C markup validation service, could be enhanced...
stripping remote namespace elements from parse tree
olivier Thereaux 
http://lists.w3.org/Archives/Public/www-validator/2007Oct/0036.html

Doug was asking, for the purpose of the CDF WG, how we
might validate documents such as this one with inline SVG
http://www.schepers.cc/w3c/role-aria/validity-test-inlineSVG.xhtml

Chris pointed the SVG approach, which is to specify, both in
prose and in NVDL, how to strip away stuff in other namespaces
and then use a RNG schema.
http://www.w3.org/TR/SVGMobile12/conform.html#ConformingSVGDocuments

(by the way, Chris pointed me to oNVDL 
http://www.oxygenxml.com/onvdl.html
which was developed by the oXygen folks, but is open source and free.
I haven't found time to check it out yet.)

Chris also pointed out "Relaxed", which is a service that supports
NVDL and RNG and worked "out of the box" with this case.

I recommended the approach to the Math WG, which is looking
at similar issues in MathML 3 development.
http://lists.w3.org/Archives/Public/www-math/2007Oct/0005.html

I thought that looked useful to a large degree, but it has
built-in knowledge of the namespace combinations, and I wondered how
it fit with bottom-up extensibility such as supported
by XML Schema substitution groups, as discussed in Feb
http://lists.w3.org/Archives/Public/public-xml-versioning/2007Feb/0000.html

It seems feasible that the CDF WG could develop an algorithm
for computing NVDL scripts in "follow-your nose" bottom-up
style.

I think it would be interesting to play around in that design
space as a way of moving the W3C markup validation service forward
from DTDs.

Meanwhile,  XHTML Modularization is in last call.

"This document is version 1.1 of XHTML Modularization, an abstract
modularization of XHTML and implementations of the abstraction using XML
Document Type Definitions (DTDs), and XML Schemas."
http://www.w3.org/TR/2006/WD-xhtml-modularization-20060705/
http://www.w3.org/TR/xhtml-modularization

I wonder if W3C should be investing further in DTDs at this point.

While the HTML 5 design work mostly treats schemas as implementations
of the specification rather than as an integral part of
the specification, I think the SVG/Math-in-weblog-comment
use case that Sam Ruby pointed out in his 13 Oct msg
  http://lists.w3.org/Archives/Public/public-html/2007Oct/0166.html
argues in favor of an integrated approach to HTML 5 style
tag soup and modular language design.


p.s hi tracker. please note this message in connection with
these other two actions too:

ACTION-16 David Orchard Incorporate the NVDL text into the findings

ACTION-48 Henry Thompson XHTML modularization and XML Schema
substitution groups

-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/
gpg D3C2 887B 0F92 6005 C541  0875 0F91 96DE 6E52 C29E

Received on Wednesday, 17 October 2007 22:12:54 UTC