WebAIM - Web Accessibility In Mind

E-mail List Archives

Re: When will developers be able to rely on ARIA?

for

From: Al Sparber
Date: Mar 4, 2009 11:25AM

Jared Smith wrote:
> On Wed, Mar 4, 2009 at 8:19 AM, Despain, Dallas wrote:
>> But when will we be able to release such technology with confidence
>> that users will reap the benefits?
>
> Now! You lose nothing by implementing it now. If site visitors are
> using an older screen reader or browser that does not utilize ARIA,
> they don't get the enhanced accessibility. While you should do what
> you can to make it accessible without ARIA, if you implement ARIA you
> can only make things better for those that have compatible
> technologies.

Hi Jared,

I have been testing various aria techniques, as well as online test cases,
in Jaws 10 with very mixed results. Granted, I am not blind, but trying to
leave Jaws to its own devices (and default settings) the results do not seem
very good. Consider this example page in Firefox 3 using Jaws 10:

http://codetalks.org/source/widgets/tabpanel/tabpanel1.html

I find myself "trapped" in the tab widget and unable to intuitively get to
the content that comes after it. Moreover, if the tab panel contents were
simply text content, instead of form controls, Jaws will not read it on its
own.

And of course, it's pretty much unusable in IE7.

I know far less about Jaws than you do, so perhaps I'm missing something? Is
there another way to approach the testing?


> Based on our screen reader survey results, it appears that most screen
> reader users have relatively up-to-date browsers and screen readers.
> I'd say the vast majority of them have ARIA compatible user agents
> now. An IE8 release with proper ARIA support will only increase
> adoption.

Do you think that in the real world, the majority of Jaws (or Window Eyes)
users are running Firefox?

It would seem to me that ARIA is being used by Ajax developers to mitigate
some accessibility questions, while leaving many more wide open - such as
sighted users with script disabled, which might number higher than blind
users, when the real need is for a solution to semantically correct and
valid client-side widgets that utilize display: none. For that, there seems
no clearer solution than Eric Meyer's "plea" from a few years ago:

http://meyerweb.com/eric/thoughts/2005/06/27/dont-read-speak/

That approach would, in my opinion, solve everything and leave Ajax
developers to come up with a logical solution for their widgets, which have
far more serious accessibility limitations - going far beyond the realm of
blindness.


--
Al Sparber - PVII
http://www.projectseven.com