- From: James Graham <jgraham@opera.com>
- Date: Mon, 08 Dec 2008 18:35:23 +0100
- To: joshue.oconnor@cfit.ie, "Ben Millard" <cerbera@projectcerbera.com>
- Cc: "Lachlan Hunt" <lachlan.hunt@lachy.id.au>, HTMLWG <public-html@w3.org>, WAI-XTECH <wai-xtech@w3.org>
>> Joshue O Connor wrote: >>> If you can, could you please highlight for me (briefly) what the >>> differences/improvements are? >> >> >> HTML5's biggest improvement is having a clear description of the table >> model with a unified header association algorithm to define how it all >> works together, imho. From this, you can assess what happens in any >> table much more precisely. > > FWIW this is useful work Ben, thank you. However, there is still a need > for consensus on the functional requirements of the language (explicit > semantic associations, aside from the algorithm) that could be ideally > supported by older UAs or, at the very least, be easily implemented by > AT vendors. "The algorithm" in this sense refers to all the processing of table header associations permitted by HTML 5; that is "the algorithm" defines the semantic associations that are made between table cells and headers in HTML5 documents including those marked up explicitly by @headers and those implied by the structure of the markup. Therefore, apart from the final details of the algorithm, I'm not sure what extra you are looking for here? Ideally the algorithm will be associated by browser vendors, not AT vendors, and the information exposed through the accessibility APIs. For use by older AT it should be possible to implement the algorithm in script, either to embed in pages from accessibility-aware authors or to run in AT where that kind of user scripting is supported and the browser has access to the DOM. This kind of script could piggyback on aria or @headers as necessary.
Received on Tuesday, 9 December 2008 08:39:44 UTC