Re: Conflicting inclusion/exclusion criteria for elements in the accessibility tree (Was: Re: [ARIA] Agenda: March 3, 2016 WAI-ARIA Working Group)

In that case, I think the author has written a very broken website.

aria-hidden should only be used for content that is either decorative or
redundant (or is currently visually hidden or offscreen).  If an author is
adding mouse-only interaction to a decorative or redundant part of the
interface, without also implementing keyboard equivalents elsewhere, they
are a horrible person and I don't mind if there website behaves in a very
broken way.

In the converse, what if an accessibility-minded author has two different
controls for the same function, one is mouse/touch oriented and the other
is keyboard/accessibility oriented.  Why should we second-guess their
decision to put explicit attributes such as aria-hidden and negative
tabindex?  For example, maybe you can pan & zoom a map or diagram by
dragging the mouse and scrolling the mouse wheel, but you can also pan &
zoom by using slider controls or buttons.


On 15 March 2016 at 11:58, Joseph Scheuhammer <> wrote:

> On 2016-03-15 1:52 PM, Amelia Bellamy-Royds wrote:
>> If something has negative tabindex *and* aria-hidden="true", it should
>> definitely not be in the tab order!
> How does that work for a keyboard only user?  On the assumption that the
> user can see it, wants to put focus on it, and the only way they can is via
> TAB navigation ...?

Received on Tuesday, 15 March 2016 18:10:56 UTC