RE: RDF in XHTML [was: Re: Authors describing what their URIs mean]

> > that text becomes well-formed.  In other words, if I decided to use
a
> > new scripting language that was based on XML and namespaces, as soon
as
> > I use it in my <script /> block, the page won't validate -- XHTML:
use
> > anything but XML..
> 
> Again, I'm not hearing a distinction between markup and document
content.
> The criticism is thus unfounded.

I can think of at least two scripting facilities that use XML markup as
the document content.  I am probably confusing myself, but it seems that
the problem comes in when the markup *is* the content, right?  Suppose
we have a scripting language of the sort:

<script language="SomeSillyLanguage">
 <event sink="onload" target="document">
  <replace newtext="the page loaded" target="#span1" />
 </event>
</script>

To me it seems that in XHTML we have only two choices:
1. make sure that the data model expressed by SomeSillyLanguage can be
validated by DTD instead of some more appropriate schema validation
tool.
2. or obfuscate/encode the script to ensure opacity

> If we agree we're not talking about content but markup, validation is
> validation, regardless of RDF or XHTML. You use the same validation

Except that XHTML requires that *all* markup in the document be
validated by DTD, which is not always the best tool.


> tool even. Putting a Dublin Core module into XHTML makes good sense,
> but as we seem to agree there's no possibility of an "RDF module,"
> nor does one even make any sense. That'd be like having an "XML
module."

If we define "makes sense" as "can be validated with a DTD", then I
agree.  But keep in mind that RDF is simply an XML format for encoding
assertions.  RDF is an application of XML, and expresses edge-labeled
graphs of logical assertions, rather than node-labeled graphs of
"anything under the sun".  I understand your concerns about enforcing
agreement here, but again that is not the domain of XHTML, and not the
domain of a DTD.