Was the post about urinals serious? Yes, it seriously satirized people who spend an inordinate amount of time and effort searching for things about which to scream discrimination. I also intended to satirize the literary form which authors of such choose to write. Finally, I wanted to say, “Chris, the level of seriousness on your blog has grown beyond proportion and its time to insert something as silly as possible to lower the tone a bit.” Recently, the topics discussed here have moved to the very esoteric so I also wanted to include something as pedestrian as a public restroom to bring us back to issues that more than a few accessibility hackers find interesting. I do find it amusing that at least one person took the post to be serious enough to write a long and critical response and I hope he returns reads, enjoys and perhaps learns something from the other posts that appear in the Blind Confidential blog. People like Will, Peter and me and hopefully others who will join us in the future, are, in fact, among the top experts in access technology and will join the other research and development people to create the next generation and beyond of access technology. So, in the future, if this blog gets far off serious topics, assume it is satire.
Today, I want to explore the definition of accessibility and how the term gets thrown around by different groups. I will, in a serious manner, take a look at standards and the lack thereof and what that means for accessibility and jobs for people with profound vision impairments.
If a web validation and repair tool, like Ramp from Deque Systems, processes a web site and reports that the html passed all of the tests, can it immediately receive the “accessibility stamp of approval.” The answer is no. If a web site includes alternate text for its entire graphics, image map links and graphical links it may remain completely unusable by a person reading the page with a screen reader. Specifically, if the authors wanted to find a way to get past a validation tool without doing any actual work, they could simply put the word “graphic” in every alt tag so, while a sighted user sees “News,” “Weather, “ and “Sports” the screen reader user hears “graphic, graphic, graphic.” Thus the site can claim compliance but not accessibility.
Next, we may encounter a web site that puts useful text in its alt-text tags so the screen reader hears “News, Weather, and Sports” but the site contains so many other elements that finding the points of interest takes a tremendous amount of time, even if the user employs the efficiency features built into screen readers like HPR, JAWS and Window-Eyes. So, can a site that follows some of the spirit of the accessibility standards and guidelines truly receive the label “accessible” if it requires a lot more time to navigate by a screen reader user than a sightie?
Moving on from web accessibility and onto desktop computers, what percentage of features of a program should a screen reader have available before an application gets called “accessible” and does the screen reader need to expose a usable interface before it claims compatibility with an application? Virtually all Windows screen readers claim to be compatible with Microsoft Word, an application which is essential to the jobs of nearly everyone who works with a computer. Virtually none of the screen readers work with more than fifty percent of the features in Microsoft Word so can the AT vendors claims of accessibility be considered true? Or they 50% true if they only work with 50% of the features?
Recently, I did a very unscientific comparison between JAWS, Freedom Box System Access and Window-Eyes in Microsoft Word. I didn’t have a recent version of HAL installed or I would have included it as well. JAWS, according to its help file and my trials did the most but I could find numerous modes and features of Word that it couldn’t handle at all and, as some features can only be used through these other features, I could not get to a number of them to even test their usability (this being true for all three of the products I tried).
Next I went onto FB System Access. The Serotek guys have really come a long way in their latest release. They do quite a lot of the things JAWS can do but they haven’t caught up entirely. FB System Access does deserve applause for exposing some items differently from JAWS and, in some cases, they found a more usable way of presenting information which hasn’t been explored too deeply in the world of screen readers. Of course, the Serotek guys have had all of the terrific features added to JAWS over the past seven years to use as an example and having a blind CEO who needs to use Word provides the Freedom Box guys with extra motivation to push the usability envelope.
Window-Eyes, I sadly say, continues to pull up the rear. JAWS users have been able to read tables, columns and embedded spreadsheets in Word for years. These, along with some other field detection features are recent additions to Window-Eyes. Neither Window-Eyes nor Freedom Box did anything with the collaboration features (JAWS does a passible job) which any writer requires to work with editors to do a successful job, especially if a document has multiple authors – a common occurrence in many workplaces and academic settings. So, in response to what I believe must have been market pressures, GW added some new features to make Word more usable for its users. I recommend they take a copy of JAWS, its help file and the Freedom Scientific MS Word tutorial and use it as a specification for their next set of improvements to the world’s most popular word processor.
Both JAWS and Window-Eyes claim support for Excel. Blind people who want to make spreadsheets beyond the most primitive, though, must use JAWS as very few features work well or at all with Window-Eyes. Freedom Box promises better Excel support in its next release and its current support is similar to that in Window-Eyes.
All three that I researched and tested, though, came up far short of 100% accessibility (one of these days, I’ll set up three or four computers next to each other and create a table with MS Office features on one axis and screen readers on the other and fill in what does and does not work to generate an exact percentage and, perhaps, put a weight on each feature as to my perceived value of each feature to come up with a score that represents actual use rather than features for features sake). All of them claim compatibility and, therefore, the user assumes accessibility.
Back to the question, what percentage of an application or system needs to be usable by a screen reader user so it can be called accessible? Recently, Apple has added a screen reader to its Macintosh line of products. According to Jay Leventhal’s scathing review, it probably works in less than ten percent of the computer that it comes with. Can a Macintosh, with such pitiful usability by a blind person be called accessible?
Durable medical goods, like wheelchairs and such, as well as medical testing devices, blood pressure machines and the like, are all subject to FDA approval. If these access technologies do not meet the government standards, they cannot be sold to the population that requires them. I do not want to suggest that the FDA or any other governmental body get into the regulation of screen readers but I do wish there could be a voluntary program run by an independent organization (ATIA?) that can publish detailed comparisons of screen readers that includes a lot of objective tests (does it work with feature x yes or no) and some subjective tests (can a user complete a task more easily or in less time with product a, b or c?). This can finally give screen reader users a way to see past the marketing hype and half-assed solutions that often carry the label “accessible” in the sales literature from AT companies.
In my mind, a product must be “usable” before it can receive the label “accessible” and that AT companies should do their best to ensure that this is truly the case. Before claiming that something is accessible please make sure that it can be used by your customers. Right now, as I mentioned in my post that talked about how Eric Damery blazes the usability trail, I think that Freedom Scientific and JAWS still lead the pack but my friends at Serotek are making a mad dash to catch up and the market share frontrunners like FS and GW should be looking over their shoulders a bit before Mikey catches them.
