WebAIM - Web Accessibility In Mind

E-mail List Archives

Re: Ajax

for

From: Alastair Campbell
Date: Oct 10, 2006 10:10AM

John Foliot wrote:
> I suspect however that the bigger issue/requirement is that
> the AJAX widget simply degrades gracefully without the
> requirement of the end user to indicate "turn it off".
> Often, it may not even be an issue of turning
> something On or Off, as Robert Nyman discovered one day
> [http://soap.stanford.edu/show.php?contentid=65#ajax2]

Presumably if the app works without JavaScript (i.e. unobtrusive/hijax)
then that situation (where the JavaScript was removed) should be fine?

> And testing with various Adaptive Technologies is always
> important. But as us "old fellers" are always reminding -
> this web accessibility thing is for
> more than just the blind/visually impaired.

No argument from me there, it's just that the issues for just about
everyone except screen reader users tend to be of the usability kind,
affecting *everyone* except maybe the technically gifted. With screen
readers, it can go from difficult to impossible, and a pretty clear
accessibility issue.

I assume there are also issues when things update away from the current
focus for people using screen magnifiers, are there any other common
issues?

> it also impacts users of alternative
> user-agents (our oft quoted "cell phones/PDA's").

Definitely, my old phone's browser choked so badly on JavaScript it
became unusable, and I couldn't turn it off.

However, I've been meaning to post about the webkit based browser (that
safari is built on), as my current phone has excellent JavaScript
support, I even managed drag and drop on my Google home page!

> You mentioned that there are a few caveats based on your
> current testing - care to share?

Sure, sorry for being cryptic. I need to gather together the JAWs
testing (i.e. talk to my colleague who completed it), I'll publish that
when I get the chance.

Using Voiceover (OSX), AJAX is not intrinsically difficult at all, the
changes within the page are updated and available to the user.

The first caveat is that updates via JavaScript aren't announced, and if
you click a link, you are often redirected to the top of the page as
though it had reloaded (although it hadn't). From that point, you can
then navigate and get to the updated content as though the page had
simply refreshed.

The second caveats are that within page links simply don't work in
Voiceover. The screen scrolls, but the cursor doesn't move (although one
person on the MacVisionaries list came up with a very impressive but
difficult workaround). One of Brothercake's tests worked as intended,
and a couple almost worked as intended (i.e. the text updated, was read
out and left the cursor there).

I'll write up the results when I get a chance.

Kind regards,

-Alastair

--
Alastair Campbell | Director of User Experience

Nomensa Email Disclaimer:
http://www.nomensa.com/email-disclaimer.html