RE: Agenda: April 28, 2016 WAI-ARIA Working Group

Sounds good to me. My main concern was that this would be overlooked. The text looks good to me as is.


-----Original Message-----
From: Joseph Scheuhammer [mailto:clown@alum.mit.edu] 
Sent: Thursday, April 28, 2016 9:21 AM
To: Bryan Garaventa <bryan.garaventa@ssbbartgroup.com>; Rich Schwerdtfeger <richschwer@gmail.com>; ARIA Working Group <public-aria@w3.org>
Subject: Re: Agenda: April 28, 2016 WAI-ARIA Working Group

On 2016-04-28 11:57 AM, Bryan Garaventa wrote:
> Currently the only option listed in this section is the use of aria-owns, which is problematic in this case.

The text states for this special case that "... the element referenced by aria-activedescendant is not in fact a descendant ...".  This is the most general way of stating the problem.  Using aria-controls does *not* solve *this* problem.

The only thing relevant here is that there is a case where aria-activedescendant does not actually reference a descendant, but it nonetheless works.  That is all the needs to be documented in the aria-activedescendant section.

> I believe one of the original proposals too was to allow the use of aria-controls as well in special cases when an explicit association is needed, such as when a text field is focused and includes role=combobox, but references another node such as role=option via aria-activedescendant.

You are correct about the use of aria-controls with comboboxes.  But, that is fully documented in the combobox role section with Matt's new text.  I don't see the point of reiterating that in the aria-activedescendant specification.

--
;;;;joseph.

'Die Wahrheit ist Irgendwo da Draußen. Wieder.'
                 - C. Carter -

Received on Thursday, 28 April 2016 16:47:26 UTC