At 10:58 AM -0400 8/30/06, Edward Lass wrote:
>  >>> "Peter Krantz" <peter.krantz@gmail.com> 8/30/2006 8:53:31 AM >>>
>>  Am I interpreting it correctly if I use it like this:
>>
>>  <section role="legal:paragraph">...
>>
>>  ...provided that the "legal" namespace has already been established
>in
>>  the document.
>
>I think this specific example would be a misuse of the XHTML2
>Structural Module, since that same module's <p/> element already means
>paragraph (and "XHTML2's paragraphs represent the conceptual idea of a
>paragraph, and so may contain lists, blockquotes, pre's and tables as
>well as inline text").[1]
>
>In this case, the @role attribute could be applied to <p/> to,
>literally, specify the role of this type of paragraph within the
>document.

Not so sure. If legal:paragraph has a content model that requires a
header, and xhtml2:header's expect sections as their @about
container, it could be more appropriate to derive legal:paragraph
from xhtml2:section.

And marking the role on a <section> or <p> would convey the
same information to legal:namespace-aware processors; what
would vary would be the handling in XHTML2 processors that
were legal-unaware.  In which case to process as a <section>
might be the highest and best fit in pure XHTML2.

[I blather.  The problem with ontologies is that there are *so many*
ways to skin a cat.]

But if what the legal community calls a paragraph walks like an
xhtml2:section and quacks like an xhtml2:section, it is best to use
the most restrictive element type (from the more commodity
vocabulary) that fits, before extending with @role qualification.

Al

>
>- Ed.
>
>[1] http://www.w3.org/TR/xhtml2/mod-structural.html#s_structuralmodule