[dsdl-discuss] Fw: CDF WG feedback on NVDL FCD

From: MURATA Makoto <murata@hokkaido.email.ne.jp>
Date: Fri Nov 04 2005 - 01:27:09 UTC

Folks,

Here is a feedback on the NVDL FCD from W3C.

I am afraid that my spam filter mistakenly blocked this message.

Martin, is this good enough to put this issue on the
agenda?

Cheers,

Makoto

Forwarded by MURATA Makoto <murata@hokkaido.email.ne.jp>
----------------------- Original Message -----------------------
 From: Mark Baker <distobj@acm.org>
 To: murata@hokkaido.email.ne.jp
 Cc: member-cdf@w3.org
 Date: Mon, 24 Oct 2005 11:20:31 -0400
 Subject: CDF WG feedback on NVDL FCD

----
Makoto - I didn't see an official email address for feedback, so if you
you could please forward the following along as our response, that'd be
great, thanks!
See you in Atlanta!
CDF WG - this closes my ACTION-258.
=======
The following are the comments of the W3C Compound Document Formats WG
on the final committee draft of the NVDL specification[1].
Issue: Assumed dispatch mechanism
NVDL appears to make some assumptions that may or may not be compatible
with the base CDF framework.  In particular, it seems to assume that
dispatching is based only on the use of XML namespaces, and not, say,
media types, or some combination thereof.  The CDF WG has, as of today,
not decided upon a dispatching mechanism.
I'll use an RDF example (D.2) in the spec to describe the issue;
<?xml version="1.0"?>
<html xmlns="http://www.w3.org/1999/xhtml">
 <head>
  <title>Some Page</title>
  <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
     xmlns:dc="http://purl.org/dc/elements/1.1/">
   <rdf:Description rdf:about="http://www.w3.org/" dc:title="W3C Homepage"/>
  </rdf:RDF>
 </head>
 <body>
 </body>
</html>
As-is, that example and the NVDL script that decomposes it into XHTML &
RDF sections, should work fine AFAICT.  But consider that not all RDF
documents have the RDF namespace as the namespace of the root element.
For example, this is a perfectly valid RDF document;
<Person xmlns="http://example.org/people">
  <Name>Mark</Name>
  <Height>180</Height>
</Person>
With that document embedded in the XHTML, there's no general NVDL script
that could be authored which could extract all RDF embedded within XHTML
because the namespaces could be anything.
Depending upon what the WG decides, this may or may not be a big issue
for NVDL.  The group thought though, that if NVDL were to provide the
ability to dispatch based on the result of an XPath expression (or
similar), that this might provide a certain degree of insulation from
what the WG eventually decides.  But of course, we can't guarantee that
this mechanism will provide a completely general solution.
Issue: Triggers
The group doesn't fully understand the trigger mechanism and how it is
expected to work.  We feel it could be better specified.
Question: Can NVDL support this example?
We wondered one would author an NVDL script to validate the following
document;
<html xmlns="http://www.w3.org/1999/xhtml" 
xmlns:xf="http://www.w3.org/2002/XForms">
  <head>
    <title>Title</title>
  </head>
  <body>
    <table>
      <!-- Invalid based on XHTML2 spec -->
      <xf:model id="model1">
        <xf:instance id="inst1">
          <stuff xmlns="">
            <morestuff>Stuff</morestuff>
          </stuff>
        </xf:instance>
      <tr>
        <td>
          <!-- Valid based on XHTML2 spec -->
          <xf:model id="model2">
            <xf:instance id="inst2">
              <stuff xmlns="">
                <morestuff>Some More Stuff</morestuff>
              </stuff>
            </xf:instance>
        </td>
      </tr>
    </table>
  </body>
</html>
Specifically, other than dispatching on a given namespace to a validator, 
do we need to use additional constraints in DSDL?
General Comments:
In section 5, data model has local names and ns URIs separated out,
which is good, but ns prefixes in scope are lost. What if anything
happens with qnames in attribute values?
Needs more examples, especially of creating element section on the
basis of something other than namespace URI, also of sections which are
not validated - how to say - "well formed non validated stuff goes
here".
Need to further study how inherited attributes like xml:lang and
xml:base are handled. 6.4.2 may be sufficient, not clear.
The use of 'mode' would be clearer with an example, too. This seems to
relate to an XML processing pipeline, but it's not totally clear.
In 8.7.4 "NOTE This part of the standard does not introduce any standard
options. Schema languages or their implementations are recommended to
introduce options so that they can interwork with NVDL." What
specifically should W3C XML Schema, and RelaxNG, add by way of options
to interwork better? I assume that something specific is meant and that
the spec is careful not to say what it is; it would still be useful to
know what is meant even if the spec remains silent here.
Thank you.
  [1] http://www.jtc1sc34.org/repository/0586.pdf
--------------------- Original Message Ends --------------------
-- 
MURATA Makoto <murata@hokkaido.email.ne.jp>
--
DSDL members discussion list
To unsubscribe, please send a message with the
command  "unsubscribe" to dsdl-discuss-request@dsdl.org
(mailto:dsdl-discuss-request@dsdl.org?Subject=unsubscribe)
Received on Fri Nov 4 02:27:28 2005

This archive was generated by hypermail 2.1.8 : Fri Nov 04 2005 - 20:03:02 UTC