Over the past week or so, I’ve seen two announcements proclaiming new technologies for community driven projects to label non-compliant web sites that work poorly with screen readers. Adding these to C-Saw from Serotek which has been around for years makes three systems that total to superfluous and counterproductive sectarian standards that should be harmonized as quickly as possible to provide the entire community with a unified database standard so, no matter the AT they choose, all will enjoy similar results.
As I mentioned the other day, Mike Calvo, CEO of Serotek, has offered C-Saw as an open standard to any team that wants to incorporate it gratis. Similar facilities by IBM and WebVisum only came to my attention over the past few days. I understand entirely why user agents to access the database of label information will differ as each needs to fit with the paradigm of the specific screen reader.
WebVisum is a Firefox plug-in and needs to conform to Firefox plug-in standards; System Access is a Windows screen reader that works best with Internet Explorer and, consequently, needs to be compliant with the rules that govern such. All I know about the IBM analogue is that it works with JAWS and, therefore, needs to (at some level) fit into that model. Knowing the IBM accessibility people pretty well, though, I would guess that it is designed for portability and will be offered to other projects as well. I would also assume that Orca and VoiceOver on GNU/Linux and Macintosh could, if they chose, add functionality to access the same database as well.
Putting the obviously platform and technology specific user agents, in this case, any screen reader or web accessibility tool, which necessarily will have variations ranging from OS to browser to user interface paradigm, aside, we can now explore a harmonized database off in the cloud that shares all of the label information. This will give everyone the head start of the more than 4000 web sites already labeled by the users of the screen reader formerly known as Freedom Box coupled with the new systems provided by IBM and WebVisum. If the new systems don’t drop easily into other screen access utilities (Window-Eyes, HAL, NVDA, Thunder, etc.) I doubt it will be too great a technological challenge to add the user agent side.
If the people from WebVisum, IBM and Serotek can get together to discuss and ultimately create a common database standard, the data can sit off in the cloud not caring whether it’s a Mac running VoiceOver, a Windows box running NVDA or a Symbian handheld running Mobile Speak and all of the users will be able to contribute labels to pages that have none and enjoy the labels others have put in before them.
As, at the very least, the current WebVisum and probably the IBM systems are at least somewhat user agent agnostic, building a universal database and populating it with five years of Serotek data as a starting point can really start a fire. Having four thousand pre-labeled sites will provide volunteers with immediate positive feedback that a community based system has been at least partially proven to work. This added layer of usability will, in my opinion, provide added motivation and tear down some artificial boundaries between users of different screen reading technologies.
As, for all intents and purposes, the WebVisum Firefox plug-in works with the other AT out there, if it could harmonize its database with the one Serotek uses, the history of community involvement in labeling pages will be preserved and five (or more) years of Serotek users and volunteers won’t be lost. I will assume the IBM system is similar and encourage them to join trilateral talks on bringing all three projects together.
A long time ago, I published an article to this blog about cooperation being the key to innovation, you can search for it in the box above, which proposed the hypothesis that, as regards underlying technology, AT vendors waste a ton of cycles reinventing the core technologies that occur in all such programs. One reason JAWS has an enormous worldwide market share is not because it has a better off screen model or virtual buffer than the others but because it exposes a user interface to far more mainstream and obscure proprietary applications than do the others.
If all screen readers and web access utilities start using a web labeling technology that shares a common database, those who provide the most comfortable user experience or invent a special way of delivering this content in a manner I cannot imagine at the moment, will win but all screen reader users, from those who cannot afford anything besides the no-cost and free solutions all the way up to the Ferrari programs can, as a single community, share a unified database.
BC, as small a voice as we are, encourages Serotek, WebVisum and IBM to put their collective heads together (they all have some really smart people), perhaps invite people from NVDA, Apple and Sun and bang out a common database format for a set of labeling facilities, work together and take at least one step toward ending some of the counter-productive sectarianism that often causes people in this biz to reinvent the wheel. I also want to emphasize that all parties to this discussion come with an open mind and be willing to make changes to their technology rather than letting their egos get the best of them – all of us are very proud of the work we do but, to build a community, we will occasionally need to put our pride aside and work toward the greater good.