## Looking back … and forward

So this whole thing started for me back in 2000. It began as a one way trip from Chicago to Netscape’s Mountain View offices, traveling through small towns and past the Grand Canyon, up through California and finally into the fog of the bay area. This was to be the last major trip for my mini-Cadillac (a 1986 Chrysler New Yorker).

Netscape barely knew what accessibility was. But they gave me an apartment to live in, and an office. I was working for a Braille note-taker manufacturer who needed web browsing capability. My coding skills were all home grown, in straight C and not in C++. I had never touched something as immense as the Mozilla codebase, and still had not grasped the true power of dynamic content. But what I did have was passion and the belief that open source was a beautiful place — and the best place to implement accessibility the right way.

Eventually Netscape hired me to integrate accessibility into the core product. There was a lot to do — not just to make Mozilla communicate XUL and HTML content and events to screen readers via MSAA, but also to add full keyboard support. I also added some special features for all users, like caret browsing and type ahead find (later enhanced by Blake Ross to become the find bar).

Then in July of 2003, Black Tuesday arrived. The Netscape browser team was laid off. I’d like to thank Asa Dotzler and others for their optimism very early on, even before this. Asa knew that Mozilla needed to be free to develop its own vision, Firefox. The truly passionate found a way to continue working on Mozilla. A year after I was laid off, IBM hired me finish the accessibility work. Yay!

All in all, I got married, had two children, went through three jobs, lived in three cities and fixed over 1000 bugs since it all started …. and we have a very accessible Firefox. Mozilla has a major role in leading the charge for accessibility of Web 2.0, not to mention accessibility on the Windows and Linux desktop.

These days, Marco Zehe, Alex Surkov and David Bolter have taken over the core accessibility work. These are the right guys to bring the quality and polish we need. It’s not all boring; there are also new features to implement. In 2009 we hope to become accessible with VoiceOver on OS X, continue to improve our support for WAI-ARIA and Web 2.0 accessibility, and address advanced topics such as accessible math, diagrams and custom widgets. It’s a bit unfortunate that these are still considered advanced topics in accessibility, when they are so basic to the mainstream. However, when Mozilla addresses new accessibility topics, it tends to pave the road with standards and documentation. This makes things much easier for implementers that come after us.

Although I’m no longer coding I work as hard as ever. I still advise in the codebase and act as module owner. I want to help make sure that decisions of the past are understood by anyone working in the code. Documentation is helpful but ultimately not enough.

What am I working on now?

• The WAI-ARIA standard and everything around it. Right now, making Web 2.0 accessible means using a narrow set of technologies which support WAI-ARIA well. The browser must currently be Firefox, and some ARIA features are only supported in certain newer assistive techologies, and definitely not on OS X. And, you’ll either need to develop your own widgets or use Dojo’s Dijit toolkit. Fortunately, these inflexibilities are slowly but surely fading away. IE, Opera and WebKit are all working on WAI-ARIA support. Numerous JavaScript toolkits, such as YUI, GWT and JQuery, are moving forward as well. Assistive technologies are improving support for cutting edge areas like AJAX accessibility via WAI-ARIA live regions. But ultimately, Web 2.0 accessibility will arrive for real when a “killer app” arrives which requires a WAI-ARIA compliant browser. That’s what will finally drive things forward.  In the meantime, I’ll try to make things as clear for other implementors as I can.
• Finding talented individuals and cowriting grant proposals for worthwhile accessibility projects that benefit the web. One example is Silvia Pfeiffer‘s project to add captioning and audio descriptions for HTML 5 video and audio. I’m also working with Eitan Isaacson, a great programmer who helps us in too many ways to count. He developed Speclenium, a tool to allow the accessibility implementations in two browsers to be compared for differences. I plan to use it to help other browsers improve their support for WAI-ARIA.
• Enabling projects to add basic accessibility testing into Firebug and develop a Firebug extension for more advanced accessibility testing. This will bring awareness of accessibility issues to a wider authoring community, finally bring the test tool development community together on a common strategy, and allow us to address Web 2.0 accessibility testing.
• Advancing the accessibility of math, diagrams and custom widgets.
• Participating in efforts such as AIA, to create a cross-platform accessibility API. Today, developing accessibility for native cross-platform applications is a nightmare because each platform has a different solution. It’s best for the web if the entire industry works together on common accessibility solutions — Mozilla, IE, Opera and WebKit.

I’m very lucky to work with the community that has developed around Mozilla accessibility. I think we follow the model of the Mozilla project as a whole, enabling each other to do great things. This is a good opportunity to thank IBM and Mozilla, and in particular Frank Hecker, for the opportunity to work as more of a leader and organizer. It’s a less direct approach than just writing the code yourself, but when the timing is right it’s the right role to take on. You’ll produce far greater results.

### 6 Responses to Looking back … and forward

1. hecker says:

Aaron,Maybe someone more familiar with math on the web can correct me, but it seems like we’ve been waiting a while for math on the web to become mainstream, and it hasn’t happened yet. So I don’t think a11y folks need to feel it’s solely their fault that math on the web is not accessible. Sites like Wikipedia are not using MathML and are resorting to hacks like displaying images.

2. Asa Dotzler says:

Thanks, Aaron, for your years of dedication. You’re an inspiration.

– A

3. Mitchell Baker says:

Aaron

I thought for a moment you were going to say you are changing gears and going to do something altogether different. I read with the proverbial “bated breath” but didn’t find this. I’m very glad it wasn’t there!

I periodically tell the story of you arriving at the early developer day at Netscape with device in hand and a wild determination :-)

4. aaronlev says:

Mitchell, thank you so much. I know you’ve helped out in the background as well, on the Mozilla Foundation board and by getting Marco, Alex and David to work on the core.

Frank, I’m not saying it’s anyone’s fault, but there are steps we can take. Thank you for bringing up Wikipedia — I just checked it out. This is a great opportunity to support the WAI-ARIA role=”math” markup. That allows ATs to know that an image is used for math, and that the alt text is a text representation. For example, in the article on the quadratic equation I see they have:

<img class="tex" alt="\begin{align} x_1 &= \frac{-b + \sqrt {D}}{2a} \\ x_2 &= \frac{-b - \sqrt {D}}{2a} \\ \end{align}" src="http://upload.wikimedia.org/math/4/2/2/422e091d8b012d061430225c8c7e03e9.png" />

We just need to ask Wikipedia to add role=”math” there, and screen readers will already know they can process the ALT as math!

It’s a definite start — not allowing screen reader users to edit math or deal with MathML, but at least read access would work.