WebAIM - Web Accessibility In Mind

E-mail List Archives

Re: accessible rotating image scripts?

for

From: Al Sparber
Date: Jan 23, 2010 11:33AM

Hi Jared,

Excuse the top-posting but my blind testers have problems following inline
discussions.

Yes, I have read and understand the ARIA specification and we have played
with it and used it in test scenarios. It makes sense for Ajax, which we do
not embrace at all, and it makes sense as a compensation for the lack of a
better solution. So maybe I can distill this debate down to its essentials,
at least as far as my perspective goes.

First, let's forget about bad scripts that make liberal use of display:
none, for example. Let's forget about generalist script libraries versus
dedicated scripts, too. Let's focus on a hypothetical scenario for a
carousel, which is something this discussion has focused on.

Let's say I'm blind. I go to this page:
http://www.projectseven.com/products/tools/horizontal-glider/testing/accessibility/basic.htm

This is a staged example of a carousel that is set to run automatically. It
has no controls. It simply cycles through 3 content panels and then
reverses.

If I'm blind, I really have no interest in what this is. That is, whether
it's a widget. I am only concerned with being able to read the content.

Now, this particular script has the capability of adding all sorts of user
controls and pagination. I have not added them for this test. But if
assistive readers were smart enough (or considerate enough) to support
targeted CSS, then it would be ridiculously simple to add all the controls I
wanted to while ensuring that blind people would not be subjected to them.
All they get is the content.

If that's not more practical and straightforward than Aria to you, then
there is no point in discussing this any further. Seems a no-brainer to me,
as well as to every blind person I have talked to.

The bottom line?

Aria where it is needed, simple logic for everything else :-)

The problem?

Aria exists today (though browser support is an issue). A simple way to feed
selective CSS to Assistive readers does not exist - but boy, if assistive
technology developers and accessibility experts don't get behind such a
simple solution, I question this industry's priorities.

--
Al Sparber



From: "Jared Smith" < <EMAIL REMOVED> >

> And this is *PRECISELY* the reason why ARIA is a great solution, maybe
> even *THE* solution. ARIA provides meaning and semantics and structure
> to screen readers in ways that are not at all possible using standard
> HTML (at least until HTML 5, partially... maybe). There's no way in
> HTML to say, "Hey, this is the navigation", but ARIA provides a
> standard mechanism for content creators to specify the navigation and
> all major screen readers to understand and utilize it. In HTML, you
> can't say, "This is a carousel. Hey screen reader, treat it as such
> and provide standard mechanisms for controlling it." With ARIA, you
> can.
>
> I can assure you that a fully functional, interactive carousel that
> has full assistive technology support and semantics via ARIA is much
> "better" (not to mention much easier to create using libraries) than a
> hacked up list of static panels that have no semantic structure and
> that are marked up to be hidden visually so that screen readers read
> them linearly (which is entirely different than a true carousel
> everyone else gets, no?).
>
> I'm quite surprised at the animosity toward ARIA. Have you read the
> spec and played around with the examples? It's not perfect, but
> support for it is quite good and getting better. It provides standard
> ways of providing accessibility that will never be possible with HTML
> and scripting alone no matter how much you hack at it.
>
>
> Yet another strong argument for ARIA.
>
> I agree that vendors need to be MUCH more involved. However, have you
> seen the level of current support for ARIA? Vendors are involved there
> and implementation is quite good, and ARIA isn't even a finalized spec
> yet. I still curse and wail at JAWS and others for lack of support for
> very basic accessibility in some places, but this doesn't mean we
> should ignore the areas where they are actually making progress - and
> ARIA is definitely an area with great promise.