- From: Joseph Scheuhammer <clown@alum.mit.edu>
- Date: Tue, 15 Oct 2013 13:30:11 -0400
- To: Alexander Surkov <surkov.alexander@gmail.com>, Joseph Scheuhammer <clown@alum.mit.edu>
- CC: David Bolter <david.bolter@gmail.com>, W3C WAI Protocols & Formats <public-pfwg@w3.org>
Hi Alex,
> This one goes probably from native markup. So if div was
> used instead input then test should be working.
Confirmed (with a span).
> Or more generic question does ARIA win over native markup
> when they are in direct conflict like <input disabled="true"
> aria-disabled="false">?
The spec states that native states/properties win over ARIA
states/properties. But, it's the opposite for roles: ARIA roles win
over native roles. The details are here:
http://www.w3.org/WAI/PF/aria/host_languages#host_general_conflict
Quoting from that section:
"When a WAI-ARIA role is provided, user agents MUST use the semantic of
the WAI-ARIA role for processing, not the native semantic, unless the
role requires WAI-ARIA states and properties whose attributes are
explicitly forbidden on the native element by the host language."
> So the question is
> should aria-autocomplete:none override value provided by input
> control?
The <input> in the test file has role="combobox". That role wins over
the <input type="text">, and, in fact, the accessible in the a11y tree
has a role of combobox. If the author has further explicitly declared
this combobox as having no autocompletion, why does the <input>'s native
autocompletion have any influence here?
--
;;;;joseph.
'A: After all, it isn't rocket science.'
'K: Right. It's merely computer science.'
- J. D. Klaun -
Received on Tuesday, 15 October 2013 17:30:42 UTC