Updated ARIA-spiced form example to work in IE 8

I updated my simple form example from the third Easy ARIA Tip to also work in IE 8. I had to explicitly state a doc type to put IE out of quirks mode into proper IE 8 mode, and I also had to change the type attribute’s value of the script tag to “text/javascript” from “application/javascript” for it to recognize the functions declared in the script block.

The example works visually, but has a number of accessibility issues which make testing IE 8 with it a not so pleasant experience:

  • Neither aria-required nor aria-invalid take any effect with either JAWS or NVDA. It’s as if the attributes weren’t set, yet the IE DOM exposes them correctly, as JAWS’s Element Info keystroke, Insert+Shift+F1, clearly indicates.
  • Neither JAWS nor NVDA see the alerts come up, and thus don’t speak them. The alerts appear visually, so the JavaScript is working, but the DOM mutation is not being picked up. Only after a refresh in the respective screen readers is the content of these visible in the virtual buffer.

For testing, I used the latest update to JAWS 10, build number 10.0.1142, and NVDA trunk snapshot 2828. And of course, the release version of IE 8.

For those of you doing web application development and testing, this is an indication that your best bet to get proper results is definitely the combination of a strong implementation of ARIA in Firefox and a supporting screen reader.

Last week in the “Accessible” module, March 30, 2009

The last week was rather short, but no less busy.

First, on the off-code side, I attended the European Accessibility Forum Frankfurt (EAFRA) conference on Friday, March 27. Christian Heilmann from Yahoo! posted a great summary of the event and also caught my guide dog Falko sleeping while I talked. The videos will appear here.

On the code side, the action happened in the mozilla-1.9.1 repository AKA Firefox 3.5b4pre this time. Today, I checked in all the approved mozilla-1.9.1 nominated patches. From Tuesday’s 3.5b4pre nightly build onwards, Firefox will:

  • Expose font sizes in PT units instead of PX, as specified in the IAccessible2 spec.
  • Support the value of “undefined” on aria-checked/aria-expanded etc. attributes, as specified in the ARIA 1.0 spec
  • Drop support for aria-channel, container-channel, and aria-datatype
  • Support aria-expanded on more roles
  • No longer support role=”description”
  • Require aria-grab to be changed to aria-grabbed for drag and drop to work in the future
  • Expose non-editable documents as readonly, regardless of role
  • Expose the ‘checkable=”true”‘ object attribute

This brings Firefox 3.5beta4 very close to the ARIA 1.0 spec. We’ll take another look to make sure we don’t miss any details from the specification for implementation. Thanks to Mike Beltzner for not jumping our throats at these 10 or so approval requests we threw at him at once!

In other news, some progress is being made towards finding the leak that running the accessibility Mochitests on Mac OS exposes. It turns out that these same objects can also be leaked by other tests, which are not accessibility related, but ours are still the best bet at reproducing them. So our master of leak detection, Carsten Book AKA Tomcat has kindly agreed to help debug this beast.

Last week in the “Accessible” module, March 24, 2008

As the accessibility team was at the 2009 CSUN Center on Persons with Disabilities conference last week, not much happened on the forefront that is visible to bug and code watchers. The only changes that landed were:

However, as David mentions in the above mentioned post, we took the chance to put our heads together and develop ideas for future features and also talked about some hands-on issues that need more immediate attention. Also, it was the first time David and I, and David and Alex met in person. I had already met Alex last year in Whistler at Mozilla’s 2008 Firefox+ summit.

Additionally, Ben Hearsum has been active in trying to get our accessibility Mochitests run on Mac unittest tinderboxes, but ran into some leaks that require further digging because we don’t use either class explicitly in the platform-independent or specialized Mac code. We’re probably uncovering a leak that’s hidden somewhere else. If anyone has any good suggestions off the top of their heads, please feel free to step in!

That’s it for this week.