This page will offer a brief introduction to two activities related with the WAI-PF mission. One is the Global Grid Forum, in particularly how it has been warming up to the idea of Web Services. The other is the pursuit of "Access Interfaces" within the U.S. National Committee on Information Technology Standards Technical Committee V2 (NCITS V2). For information, NCITS is usually pronounced 'encites.' It takes framing some context to explain why these are of particular interest.
At the Global Grid Forum meeting in Vienna, VA this July, there was an accelerating realization that heavy duty distributed computing is looking at the same consumer base as are the emergent standards for Web Services. Web services looks like the next default or commodity client/server arrangement for services in their league. This is of interest to WAI PF because the Grid Forum customers from Computational Science have some of the same needs for very-well-described resources that ataptive strategies serving people with disabilities do. The Global Grid Forum is based in a community doing big computations and serving into big delivery contexts. NCITS V2 holds up the small end of the scale, with concern to be able to access pea-brained devices such as the elevator and toaster, from reasonably small devices that are miniaturized to a mobile phone form factor at least, in some cases. But Braille Note Takers won't necessarily compress much in their form factor but they will be able to grow radically in compute capability, so the NCITS V2 work can't isolate a community of either servers or clients or networks that are uniformly small. It just has to work down to the small end. General Web Services fall somewhere in between. They assume more server and network than does the entry-level NCITS V2 scenario of two nodes on a BlueTooth piconet. But don't necessarily worry about the upper end that the Grid Forum worries about. I have been looking at the permutations and combinations and don't see a place to draw a line in Device Independence. The principles we can suggest to the DI group scale across all these cases and there are people with disabilities who need almost all the cases. So it is not clear that less is good enough.
In the Global Grid Forum, one finds Advanced Collaborative Environments which has an interest in how to cook session resources so as to be presented in situation-adapted and user-directed views. This is where the issue of "what is the accessibility API for Flash" fits. They have people working on Scene Graph libraries, and they will be pushed by their users to do more to make it easy to incorporate commercial software in their collaboration resource base. The Grid Protocol Architecture working group is trying to make a simple story of what makes it all go together. As this group makes progress, it is likely to contain patterns that we can steal from in our attempts to frame a coherent roadmap for adaptation for accessibility and device independence.
We have a problem still with the model of how resources blend in the user interface. SMIL and SVG and the Common Formatting Properties work all have useful technology, but in the time to come with more data, more natural modes of expression, and so on we will need a more powerful and effective model of how a user interface is composed from the client-side representation of resources and services. The Advanced Collaborative Environments Working Group has the same problem and they have more resources to prototype and experiment than we have. So I am working to get our issues on their radar screen, so some of those resources get applied in ways that we learn from and the W3C document set can be smarter as a result.
A few links to get started, here:
We have recently released a draft of XML Accessibility Guidelines. This is a design guide for Web Applications, more than Web Documents. Of course it inherits a lot of ideas from the guidelines for Web Content, User Agents and Authoring Tools. But as XML is a language-building language, this document carries the accessibility into this deeper step of how on puts an XML dialect together. The exploading use of XML messages to carry dialogs [XML Protocol, SOAP 1.1] suggests a real need for this. In particular, running a Remote Console dialog over SOAP for Alternate Interface Access as the NCITS V2 is prototyping, is exactly one of these XML applications and its purpose is accessibility of a broad class of target devices and services.
The best thing for accessibility on the Web would be if the Device Independence Activity were to develop a very powerful application generator capability so that voice browsing, XForms, and the like would drop out as just two of parametrically-tuned morphs, that would scale up to the Grid Forum's powerwalls and down to someone with a single switch and vibrator. The XML Accessibility Guidelines, or XMLGL for short, are an attempt to lay out the accessibility requirements for such a technology. The more of these principles are accepted into the mainstream in the Device Independence and Web Services technology, the more accessible the Web world is going to be, let's face it. We would prefer not to have accessibility documents at all. But this won't happen automatically or overnight. PF and V2 will have to continue to plow ahead and break new ground before that generic version of a solution to our problem matures. People with disabilities have historically been technology pioneers.
Links related to this topic:
Hopefully the combined efforts of the Advanced Collaborative Environments group from the large-node end and the Device Independence group with their small-device focus will convince the mainstream Web Services evolution to actually do something about independence of the service delivery context, which if it is there for devices will have some capability we can turn to advantage for people. If this happens, it will be very good for us, because the dissemination of knowledge about how to mount device-independent services on the web will create a lot of technology and skills in the industry that will then be available to the developers of our targets. The XML Accessibility Guidelines are at the moment just a rough sketch of how to go about it; but they have already received compliments on providing a compact but systematic guide to developing Web Aps in XML that doesn't seem to be available elsewhere. So the document may have real influence. The NCITS V2 scenario is a very important special case squarely in the center of the problem space that this document addresses.