- From: Henri Sivonen <hsivonen@iki.fi>
- Date: Fri, 28 Aug 2009 15:30:45 +0300
- To: Lachlan Hunt <lachlan.hunt@lachy.id.au>
- Cc: HTML WG <public-html@w3.org>, W3C WAI-XTECH <wai-xtech@w3.org>
On Aug 28, 2009, at 15:00, Lachlan Hunt wrote: > Henri Sivonen wrote: >> On Aug 28, 2009, at 13:32, Steven Faulkner wrote: >>> the example for footer in the html 5 spec does not fit the >>> definition of contentinfo in the ARIA spec. >> It seems silly to have almost one-to-one mapping between the new >> HTML5 sectioning elements and ARIA landmarks but not quite one-to- >> one. Surely the reasonable thing to do here is to adjust either >> HTML 5 or ARIA or both so that the following mapping holds: >> role=main == <main> > > I'm not so keen on using the name <main>, and would prefer we used > <content>. <main> is less ambiguous than <content>. Also, see http://lists.w3.org/Archives/Public/public-html/2009Aug/1360.html > <main> seems like it could only be used once per document, but if we > introduce an element for this purpose, it should be able to be used > within multiple sections of a page, just like header and footer can > be. Doesn't that defeat the point of the element? (Of course, error recovery needs to be specced for the multi-<main> case. What does JAWS do with multiple role=main?) >> role=contentinfo == <footer> >> role=banner == <header> > > Do you think it would be acceptable to make these mappings subject > to where those elements are used, like I suggested earlier? No, I think the interface between the browser and AT should be exactly the same for role=contentinfo and <footer>, etc. That is, it should be possible to add <footer> support to browsers without needing changes to AT. I think the AT shouldn't be able to tell if the original markup was <div role=contentinfo> or <footer>. >> role=complementary == <aside> > > The spec currently sets the note role as the default, but allows > complementary or search to be used. Should the default be changed? That doesn't make sense to me. role=search should be applicable only to <form> as I understand it. -- Henri Sivonen hsivonen@iki.fi http://hsivonen.iki.fi/
Received on Friday, 28 August 2009 12:31:30 UTC