- From: Bryan Garaventa <bryan.garaventa@ssbbartgroup.com>
- Date: Thu, 7 Apr 2016 18:54:04 +0000
- To: Joseph Scheuhammer <clown@alum.mit.edu>, Rich Schwerdtfeger <schwer@yahoo.com>, ARIA <public-aria@w3.org>
Thanks, that change looks good to me, and also works when aria-activedescendant on the focused DOM node points to either role=option, role=treeitem, or role=gridcell. Matt's point about using aria-controls to set an explicit relationship is valid too. -----Original Message----- From: Joseph Scheuhammer [mailto:clown@alum.mit.edu] Sent: Thursday, April 07, 2016 11:24 AM To: Bryan Garaventa <bryan.garaventa@ssbbartgroup.com>; Rich Schwerdtfeger <schwer@yahoo.com>; ARIA <public-aria@w3.org> Subject: Re: Agenda: April 7, 2016 WAI-ARIA Working Group On 2016-04-07 12:25 PM, Bryan Garaventa wrote: > That looks good to me. All right. However, it occurs to me that both situations are possible. That is, the text input field uses aria-activedescendant to reference the faux focus. In the case you are concerned with, that input has role="combobox". In the other case, it has a textbox role. However, in neither case is the element referenced by aria-activedescendant actually a descendant of the input. Both situations have to be addressed. With that in mind, here is another attempt that covers both: "A special use case is a combobox where its text input field has DOM focus, and that input element uses aria-activedescendant to reference the currently active item within the combobox's popup. In one case, the text input field has role="combobox" whereas, in another case, it is a textbox. Regardless, the element referenced by aria-activedescendant is not in fact a descendant of the input. Even so, the aria-activedescendant property is treated by user agents as the active descendant of the combobox." -- ;;;;joseph. 'Die Wahrheit ist Irgendwo da Draußen. Wieder.' - C. Carter -
Received on Thursday, 7 April 2016 18:54:37 UTC