Re: What do you think about the use of aria-label on elements with no role?

While I see a need for authors to give pronunciation hints to speech 
engines I don't think the engine should ever be required to use them. If 
pronunciation hints are given for web content they should never replace 
the accessible name in the accessibility API. So this is a vote against 
allowing aria-label on a div or a span.

Ideally I would like the ability to specify them, the AT will read that 
there is a hint that this word is pronounced in a certain manner, and 
give the user a choice as to whether to add this word to the global 
speech hints dictionary or to ignore this hint and "never ask me again"

This is important due to the use case where a user has content which 
uses the same hard to pronounce words in both web content and non-web 
content such as Word documents. It may be very important that the user 
knows that they are both using the same word (or company name) even if 
they are both pronounced incorrectly. It would be great to be able to 
fix the pronunciation system-wide rather than only in the content being 
accessed.

Regards,
James



On 4/7/2015 6:31 AM, White, Jason J wrote:
>> On Apr 6, 2015, at 19:17, James Craig <jcraig@apple.com> wrote:
>> Still open for discussion. I'd probably err on the side of ~"not apply to those with a specifically generic role (e.g. div/span) but still open on those with a role that is not mapped to ARIA (e.g. <video>)." The goal is to get us closer to a 1:1 mapping for ARIA to each host language so these ambiguities will fade over time.
>>
> I would also prefer not to honor aria-label on elements with a generic role. Some authors would very much appreciate the opportunity to write <span aria-label=“pronunciation”>original text</span>. However, this has the undesirable effect of also presenting the value of aria-label in braille output, where it can be particularly verbose and inappropriate, making the text harder rather than easier to read. The author’s implicit assumption that text to speech is being used simply does not hold in this case.
>
> Nevertheless, the above-mentioned scenario does point to a genuine need to be able to customize pronunciation, or indeed braille output, to ensure high-quality rendering. There is a need for implementations of SSML or CSS 3 speech properties, or, failing this, the creation of a suitable mechanism that authors can easily use to specify pronunciation.
>
> Tempting as it may be, aria-label is not the right mechanism for this, and the observation earlier in this thread that it is inconsistently implemented reinforces this conclusion.
>
>
> ________________________________
>
> This e-mail and any files transmitted with it may contain privileged or confidential information. It is solely for use by the individual for whom it is intended, even if addressed incorrectly. If you received this e-mail in error, please notify the sender; do not disclose, copy, distribute, or take any action in reliance on the contents of this information; and delete it from your system. Any other use of this e-mail is prohibited.
>
>
> Thank you for your compliance.
>
> ________________________________

-- 
Regards, James

Oracle <http://www.oracle.com>
James Nurthen | Principal Engineer, Accessibility
Phone: +1 650 506 6781 <tel:+1%20650%20506%206781> | Mobile: +1 415 987 
1918 <tel:+1%20415%20987%201918> | Video: james.nurthen@oracle.com 
<sip:james.nurthen@oracle.com>
Oracle Corporate Architecture
500 Oracle Parkway | Redwood Cty, CA 94065
Green Oracle <http://www.oracle.com/commitment> Oracle is committed to 
developing practices and products that help protect the environment