- From: Leif Halvard Silli <xn--mlform-iua@xn--mlform-iua.no>
- Date: Mon, 09 Nov 2009 18:01:11 +0100
- To: stephane.deschamps@orange-ftgroup.com
- CC: 'Sam Ruby' <rubys@intertwingly.net>, 'HTMLWG WG' <public-html@w3.org>, 'W3C WAI-XTECH' <wai-xtech@w3.org>
stephane.deschamps@orange-ftgroup.com On 09-11-09 14.15: > Hello Sam, > > Here is one real use case: > > I designed DHTML menus, like everyone using unordered lists and links, and > then added roles for browsers that do understand them. > > Here is the thought behind that: if your browser or AT don't understand > ARIA, you will understand that the menu is made up of links, and then > clicking these links will take you to whatever page/action this link points > to. > > If you have an ARIA-enabled browser, you'll likely prefer to hear that this > link is menu item 3 of 5 in such-and-such submenu. > > Thus I *need* to add a @role to an A element, for accessibility purposes as > well as backward-compatibility (my company is stuck with IE6 and will > probably be stuck with IE7 or IE8 for a few years in the future). May be I did not get what you meant. But ARIA is an advanced accessibility feature. Not all user agents supports it. If we permit @role to diverge in a confusing way from the default role of the element, then we risk that old and new user agents see the same element differently. Therefore I would argue that it is more backward compatible to use the HTML 4 included <button> element, which in ARIA aware user agents defaults to role="button", rather than to use the <a> elements with role="button", and risk that the old user agents presents it as a link. If IE6 doesn't support the button element (does it?), then we should look at that specific problem. May be that's and issue that means we require looser rules for ARIA @role. However, authors should still, in some way, be recommended by the validator to use the <button> element instead of <a>. -- leif halvard silli
Received on Monday, 9 November 2009 17:01:49 UTC