- From: Leif Halvard Silli <xn--mlform-iua@xn--mlform-iua.no>
- Date: Wed, 15 Sep 2010 17:17:23 +0200
- To: 'W3C WAI-XTECH' <wai-xtech@w3.org>
Leif Halvard Silli, Wed, 15 Sep 2010 15:06:22 +0200: > Dear WAI-XTECH listees! Is this the right place to propose changes to > ARIA? ... snip ... > Basically, the @longdesc use case can be described as "ability to place > a link inside an image" - which is something that HTML4 allows, despite > that it hasn't worked so well as intended. Thus ARIA has had every > reason to include that opportunity in its role="img" model. Let me > present some examples showing the (lack of) possibilities, in ARIA today ... snip ... > Conclusion: As it stands ARIA has not been designed to cope with the > @longdesc pattern - ARIA does not allow authors to place a link inside > an element with role=img. I would suggest that ARIA is changed so that > example 1 (or 2a/2b) somehow could be made to work. (Example 3 stands > out as mor of a hack.) > > The change that is needed would involve making sure that it is possible > to avoid that a link within a role="img" element is treated as > presentational children. [*] It should also be possible to have the lin > work even if it has display:none. The condition which would make the link visible to AT, could simply be that the link has rel='longdesc' set: > (I have also previously suggested > that rel="longdesc" (<a rel="longdesc" href="*"></a>) could be used to > make clear that the link is a longdesc link, even when it is lacking > link text.) I am ready to file a bug about this: ARIA should operate with a model of role='img' that corresponds to how <img> works in HTML. -- leif halvard silli
Received on Wednesday, 15 September 2010 15:17:59 UTC