Re: misc editorial questions/issues on spec (finding XHTML profiles)

On Thu, 2007-06-21 at 10:53 -0400, Booth, David (HP Software - Boston)
wrote:
[...]
> Q11: Sec 4, normative definition of metadata profile name is unclear
> a. The definition does not say that the "XHTML document" must conform to
> the XHTML 1.0 specification.  It requires only that the root element is
> "html" and has namespace http://www.w3.org/1999/xhtml .  Is this
> intentional? 

Yes.

RESOLVED: that the premise of the rel="transform" rule depend only on
(a) XML-wf-ness (b) root element name "html", (c) root element namespace
http://www.w3.org/1999/xhtml

 -- 24 Apr minutes

http://lists.w3.org/Archives/Public/public-grddl-wg/2007Apr/att-0212/2007_04_25-grddl-minutes.html

(cited from changelog entry Revision 1.255  2007/04/26 15:47:24 )

> b. Must the "head" element be in the http://www.w3.org/1999/xhtml
> namespace?

Oops.

A literal reading of the spec says no, and likewise the WG decision.
But I think the implementations all say yes. And as Jeremy points
out, so does the informative rule.

Maybe it's clear enough as is? Or is it worth copying the
XPath expression from the informative rule into the normative rule?
Before I make that change (which obviously
involves re-considering the decision above) I'd like somebody
else to concur. Chime? Harry?


> c. Must the "profile" attribute be in the http://www.w3.org/1999/xhtml
> namespace?

No.

> d. Must the "head" element be a child of the root element?

Yes. (again, this is somewhat implicit in the current text).

> e. What if there are multiple "head" elements?

Ouch. That makes my head hurt. I guess I agree
with Jeremy: all profiles count.

-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/

Received on Monday, 25 June 2007 22:47:15 UTC