Re: ARIA roles added to the a element should be conforming in HTML5.

Jonas Sicking On 09-10-22 15.36:

> On Wed, Oct 21, 2009 at 5:39 PM, Leif Halvard Silli
> <xn--mlform-iua@xn--mlform-iua.no> wrote:
>> Lars Gunther On 09-10-22 01.53:
>>
>>> 2009-10-21 15:11, Shelley Powers skrev:
>>>> If you go out and search on Google for "CSS link button" you get many
>>>> tutorials, examples, etc, that focus on styling links as buttons. If
>>>> you search on "JavaScript link button" you get many libraries that
>>>> support this functionality.
>>
>> [...]
>>
>>> So while I do not agree - this time - with Steve and Shelley, I must
>>> emphasize that the wording on the error messages in the validator must
>>> strongly suggest using the button element, and really not be worded in any
>>> way that discourages the use of ARIA.
>> I think the issue must be split it two:
>>
>> (1) If there is conflict between the role attribute the element, then should
>> the validator assume that a misapplied role or a misapplied element?
>>
>> (2) What kind of roles are the <a> element meant for?
> 
> There's also the question, "Should any role be allowed on any
> element". For example I asked earlier in this thread if it should be
> conforming to have an <h1> element with ARIA attributes indicating
> that it is a botton. It seemed like at least some people thought that
> that should be conforming.

In the spirit of "don't break the Web", the most important 
question seems to me be to be "should it work?" Should a <h1> with 
a role="button" be presented as a button in accessibility devices?

It seems to me that it should work - it should be presented as a 
button. But that, at the same time, a validator should consider it 
an error to apply aria="button" to a <h1>.

Comments?
-- 
leif halvard silli

Received on Thursday, 22 October 2009 14:08:50 UTC