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

Forward to the list - due to address error.

----- Begin forwarded message -----
From: Alexander Surkov <surkov.alexander@gmail.com>
Date: Mon, 9 Apr 2012 16:14:11 +0900
Subject: Re: Alternative to @aria-describedAT: <a role=img>
To: Benjamin Hawkes-Lewis <bhawkeslewis@googlemail.com>
Cc: Leif Halvard Silli <xn--mlform-iua@målform.no>, "wai-xtech@w3.org" 
<faulkner.steve@gmail.com>
In-Reply-To: 
<CAEhSh3fB2hHFz4sRREk0ikjOWtnAFocywNbdemu7vwyZMKfkPA@mail.gmail.com>

I don't feel either that ARIA requirements are clear and I don't feel
that those requirements are always correct.

For cases like <a role="img"> I'd say the UA shouldn't remove native
semantics. The same time UA can use proper mechanism in API to expose
that the link has ARIA role="img" attribute, for example, xml-roles
object attribute in IAccessible2 case. Maybe it's not really hard to
figure out all cases of ARIA usages on native markup that make the web
page less accessible than it was and deny those cases.

Thanks.
Alex.

On Mon, Apr 9, 2012 at 3:49 PM, Benjamin Hawkes-Lewis
<bhawkeslewis@googlemail.com> wrote:
> On Mon, Apr 9, 2012 at 3:51 AM, Alexander Surkov
> <surkov.alexander@gmail.com> wrote:
>> <a role=img href> is a real edge case and it doesn't sound a good ARIA
>> use case.
> 
> I tend to agree that it's not a good use of ARIA, but this doesn't
> help us: if ARIA markup becomes commonplace, bad usages will become
> commonplace, and we need to define how those usages should be
> interoperably interpreted.
> 
>> Also look at:
>> "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"
>> 
>> <a role="img" href> is not described case but the most interesting
>> part that ARIA doesn't require to ignore the native semantics blindly.
> 
> I think the most natural interpretation of the cited text is that it
> does require that user agents to exclude native semantics from the
> accessibility tree in this case, since no attributes required by role
> "img" are forbidden on <a>. The "jump" action is part of the semantics
> of <a>.
> 
>> It sounds reasonable with me if ARIA would add or extend native
>> semantics rather than completely remove it.
> 
> I suspect extension is a lot harder to define, understand, and express
> in platform accessibility APIs than replacement. I think the ARIA spec
> requires replacement, but as so often it's not really clear. Maybe
> PFWG could improve things here.
> 
>> From the user/web author point of view I don't see any benefits that AT 
>> users see an image but
>> sighted users see a normal link.
> 
> Indeed. Hence <a role="img"> is non-conforming in HTML5.
> 
>   http://dev.w3.org/html5/spec/wai-aria.html#wai-aria

> 
> Unfortunately, we still need to define what it does, since
> non-conforming ARIA roles still apply.
> 
> --
> Benjamin Hawkes-Lewis

Received on Monday, 9 April 2012 13:48:17 UTC