W3C home > Mailing lists > Public > wai-xtech@w3.org > April 2012

Re: Alternative to @aria-describedAT: <a role=img>

From: Benjamin Hawkes-Lewis <bhawkeslewis@googlemail.com>
Date: Tue, 10 Apr 2012 22:03:53 +0100
Message-ID: <CAEhSh3cPcxq-Cr=igAAa=J6jCL+5O4p5Ucd5a=wcUOwSOw7JSg@mail.gmail.com>
To: Leif Halvard Silli <xn--mlform-iua@xn--mlform-iua.no>
Cc: Alexander Surkov <surkov.alexander@gmail.com>, wai-xtech <wai-xtech@w3.org>, Steve Faulkner <faulkner.steve@gmail.com>
On Tue, Apr 10, 2012 at 9:18 PM, Leif Halvard Silli
<xn--mlform-iua@xn--mlform-iua.no> wrote:
> Benjamin Hawkes-Lewis, Tue, 10 Apr 2012 18:22:14 +0100:
>>    http://www.w3.org/WAI/PF/aria/complete#intro_ria_accessibility
>  When I take what your presentation of what WAI-ARIA says about <a
> role=img href=* > ...
>> User agents, user agent addons, or external AT can perform ARIA
>> "processing" directly on the DOM rather than through the mediation
>> platform accessibility APIs. For such processing to be compliant with
>> the spec, they must interpret the element in terms of its ARIA role
>> ("img") not its native semantics ("a href"). Similarly, when exposing
>> the element to platform accessibility APIs, user agents must interpret
>> the element in terms of its ARIA role not its native semantics.
>  together with what WAI-ARIA says about  <a role=presentation href=* >
> ...
>> The spec means Safari's behavior is incorrect. Since the element *is*
>> focusable, the presentation role should be dropped in favour of the
>> element's native semantics, and it should be exposed to the
>> accessibility tree (used by VoiceOver) as a link.
>  ... my conclusion, is that, per the ARIA spec, then a <a role=img
> href=* > is distinctly different from <a role=presentation href=* >.

Per the HTML5 and ARIA specs, yes.

> Which strengthens me in the view that HTML5 ought to permit <a role=img
> href=* >.

I don't see how that follows from the difference.

> Agreed?

If anything, I think it would be more consistent to expand the
processing of "presentation" so that other role annotations that do
not imply interactivity (like "img") never supplant native semantics
that do imply interactivity (like "a href").

> Btw, it seems seems to me that not only VoiceOver+Safari but also Jaws
> 13 have incorrect behavior as it does not expose the link-i-ness of <a
> role=img>.

Is that because of JAWS or because of what browsers are exposing to
the accessibility tree?

Benjamin Hawkes-Lewis
Received on Tuesday, 10 April 2012 21:04:43 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:51:46 UTC