- From: James Craig <jcraig@apple.com>
- Date: Tue, 01 Apr 2014 10:20:51 -0700
- To: Matthew King <mattking@us.ibm.com>
- Cc: W3C WAI Protocols & Formats <public-pfwg@w3.org>, Shane McCarron <shane@aptest.com>
On Apr 1, 2014, at 7:23 AM, Matthew King <mattking@us.ibm.com> wrote: > My understanding is that <div role="region" roledescription="chapter" .... would still expose the role of the div to the AT as region. None of the role information is lost. That's correct. One minor nit though, we're still proposing an aria-prefixed attr, not an unprefixed attr like role or tabindex. > Presumably in this case, an e-pub reader would reveal navigation commands to users based on the roledescription, e.g., a "Next Chapter" button with a keyboard shortcut and/or its own gesture. That'd be controlled by the EPUB spec and each implementation. That type of behavior could not come from ARIA until if and when we decided "chapter" was worthy of its own ARIA role. > If an element is not interactive, then changing the name of the role revealed to an AT user could make content more understandable without risking any deterioration in operability. The possible improvements in understandability with this feature could be substantial. Making up for the ambiguity of generic roles like region or group with labels creates messy, hard to understand experiences. In fact, as an end user, I almost detest roles like region and group just for this reason. I'm glad you agree. … > Since the semantic role would still be revealed to the AT for an interactive element, the AT could base any user hints or tutor information about operation on the role but speak or reveal the role of the element using the roledescription.
Received on Tuesday, 1 April 2014 17:21:33 UTC