- From: Joanmarie Diggs <jdiggs@igalia.com>
- Date: Wed, 18 Mar 2015 20:22:55 -0400
- To: Richard Schwerdtfeger <schwer@us.ibm.com>
- CC: PF <public-pfwg@w3.org>
On 03/15/2015 01:07 PM, Richard Schwerdtfeger wrote: [...] On the rowindex property, you stated: > When applied to an element having a role of "row" the user agent must > expose the proprety to assistive technologies on all descendant cells and > gridcells in addition to the row it is applied to. Does the above belong in the ARIA spec, or in the User Agent Implementation spec? I'm leaning towards the latter. Ditto for this next one on the rowcount property: > When applied to an element having a role of "row" the user agent must > expose the proprety to assistive technologies on all descendant cells and > gridcells unless overridden by applying the property directly to a > descendant cell or gridcell. In fact, at least in the case of ATK/AT-SPI2, the user agent MUST expose the property via the Table interface; the count is NOT a property of the cell on my platform. I think that in both of these instances, user agents MUST expose the properties to platforms in the way those platforms expect them to be exposed. :) As such, I think we should remove both proposed normative statements for the ARIA spec and put the appropriate equivalents in the mapping spec. Thoughts? --joanie
Received on Thursday, 19 March 2015 00:23:29 UTC