IE and document.write [was: Exploring new vocabularies for HTML]

Chris,

Another question from a while back from Thomas Broyer, this one
regarding IE8 handling of document.write

Thomas Broyer <t.broyer@gmail.com>, 2008-04-10 17:55 +0200:

> On Thu, Apr 10, 2008 at 4:48 PM, Philip Taylor wrote:
> > Thomas Broyer wrote:
> >
> > > To add to the discussion, here are small tests of IE's "namespace
> > support":
> > > http://www.ltgt.net/ie-namespace-tests.html
> > > I've run the tests on IE7 on Vista, IE6 on WinXP SP2 (Virtual PC image
> > > from Microsoft) and IE8b1 on WinXP SP2 (Virtual PC image from
> > > Microsoft), the results are exactly the same.
> > >
> > > What's noticeable:
> > > ...
> > >
> >
> >  The behaviour is different if you document.write the namespaced HTML code -
> > namespace declarations are inferred automatically when you write an element
> > using a new prefix. Compare
> > http://philip.html5.org/misc/ie-namespace-tests.html to the version above:
> > the <baz:quux:foo>, <baz:foo> and <quux:foo> elements get put in 'baz' and
> > 'quux' namespaces (with no urn) in IE6/8b1.
> 
> Yup, that's why I got confused comparing the behavior in the Live DOM
> Viewer and the one described in Microsoft's documentations; and later
> comparing the "live" and "dead" renderings in the "Zombie DOM Viewer".
> 
> Hence a test-document hosted on my domain ;-)
> 
> Now, I don't think document.write()ing namespaced elements is much
> used "in the wild" (I suspect namespaced elements in IE to be mainly
> in use with "element behaviors", deployed in intranets –I did code
> such things–; outside document.write()s), so not having these inferred
> declarations with document.write() shouldn't be a big deal.
> 
> Again, maybe someone at Microsoft have figures?
> 
> -- 
> Thomas Broyer

-- 
Michael(tm) Smith
http://people.w3.org/mike/
http://sideshowbarker.net/