- From: Joanmarie Diggs <jdiggs@igalia.com>
- Date: Thu, 02 Apr 2015 10:45:16 -0400
- To: Alexander Surkov <surkov.alexander@gmail.com>
- CC: W3C WAI Protocols & Formats <public-pfwg@w3.org>
Hi Alex. Comments inline. On 04/02/2015 09:56 AM, Alexander Surkov wrote: > On Wed, Apr 1, 2015 at 11:59 PM, Joanmarie Diggs <jdiggs@igalia.com [...] > 1. https://rawgit.com/w3c/aria/tables/aria/aria.html#aria-colcount > 2. https://rawgit.com/w3c/aria/tables/aria/aria.html#aria-rowcount > > > I don't thumb up on these until I understand why they are allowed on > elements other than table-like containers. I guess I missed some > discussion where it was explained so can I have some summary on it please? I *believe* the use case had something to do with backwards compatibility or ATs that access things via the DOM or ... something to that effect. I'll leave it to someone who knows the answer to answer it. :) (Merely responding now to be clear that I'm not ignoring your question or concern.) > 3. https://rawgit.com/w3c/aria/tables/aria/aria.html#caption > > > I don't see a need for it as long as the author can label the table. Do > I miss something? If the caption is caret-navigable, it should have an accessible object. For my particular platform, if the author used some other role (label?) and that other role was a valid owned element of a table, and user agents mapped it to ATK_ROLE_CAPTION, then I wouldn't have any objection to not having the role in ARIA. That said, I'll leave it to others to chime in. I was tasked with proposing all the roles. Caption seemed like it might be something authors would find helpful, so I proposed it. But as we saw with tbody, thead, and tfoot, I don't always get it right. ;) [...] > 7. https://rawgit.com/w3c/aria/tables/aria/aria.html#gridcell > > Do we need role="gridcell" having role="cell"? One is a widget/control; one is not. I think we need both. --joanie
Received on Thursday, 2 April 2015 14:45:56 UTC