Last week, Curtis Chong, the seemingly permanent president of The NFB in Computer Science (NFBCS) published an article in Braille Monitor highly critical of accessibility at Microsoft, especially of the accessibility on its Windows platform. Chong presents a number of indisputable facts with which I agree entirely, there are many things regarding accessibility to people with vision impairment that Microsoft does very poorly. Chong is also correct that accessibility across the Microsoft catalogue is highly inconsistent with some programs providing excellent coverage and others providing none at all. I applaud Curtis for the shedding light on the problems he describes and hope Microsoft will take action to remedy them as quickly as possible.
I also felt that Chong’s article was misleading, that it contained statements that were either inaccurate or unverifiable but, worst of all, it lacks detail in the historical context, an elephant sized hole in the story.
Over the past month or so, I’ve been exploring the concept of leadership in the blindness and technology space. I’ve talked about the changing leadership paradigm in, “Anarchy, Leadership and NVDA,” I’ve discussed leadership in innovation through traditional paths in my CSUN 2015 report and, in my article about Be My Eyes, I discussed another path a team had taken to lead a project of significant value. This article will, in the context of Chong’s piece, explore leadership in technology from NFB and how it, in my opinion, has been a failure for decades.
This article was sourced through public records and through private conversations and communications with 2 former and 2 current Microsoft employees and a number of others who had witnessed some or all of the events described herein. In my role at Freedom Scientific, I was also party to and present at some of the discussions summarized below. Thus, my sources are not “anonymous” but, rather, “unnamed.” As some of these statements are controversial, I will not reveal my sources as they may face retribution. They can, if they choose, self-identify themselves in the comments section. It is very likely that at least one section in this piece will be broken out into a separate and more detailed account of that part of the history as I think you will find it very interesting.
What Curtis Got Right
First, I’d like to recognize that Curtis gets a lot correct in his piece. I’ll confirm that most of the facts I did check are true. The statement he makes in his opening paragraph, “For those of us who are blind, access to Microsoft products is not just something that we would like to have. Rather, full non-visual access to Microsoft products is essential if we are to have any hope of being able to compete in today’s technology-driven labor market, let alone maintain parity with our sighted neighbors at home,” could not possibly be more true.
Chong’s article lists a number of MS products that are mostly inaccessible. Chong’s conclusions, that MS still has a lot of work ahead of them to ensure true and universal accessibility is also true.
The Elephant Sized Hole In The Story
If you haven’t already, please stop reading here and read Chong’s piece immediately.
Now that you’re back, ask yourself, what piece of technology fundamentally important to users with vision impairment does Curtis not mention in his article? If Apple has VoiceOver, a fully featured screen reader, Google has TalkBack, a rough attempt at the same and the Gnome Foundation has Orca, why does Microsoft have no fully featured screen reader of its own? Curtis may have simply been careless in his reporting, he may have been so focussed on application accessibility that he simply forgot to include the missing screen reader in his analysis. Or, as I contend, Curtis left out this detail intentionally.
The elephant sized hole in the story is that NFB has been on the wrong side of the leadership argument in their interactions with Microsoft. NFB’s positions have prevented MS from making its own screen reader and, as we will see later, its continued support for third party commercial AT is part of the reason why Microsoft still has accessibility problems in its technology.
The pressure NFB has put on Microsoft into not building its own screen reader, preferring instead to accept that third party screen readers would provide access to the Windows operating system has been a failure. It is for this reason that, when one launches Narrator, the screen access utility from Microsoft, it tells the user that it is not a fully functional tool and is only useful as a temporary solution until the user installs a real one.
the economic realities of business in the 21st century meant that the “mom and pop” companies like HJ and Blazie Engineering would find their way into a merger/acquisition deal that would put ruthless venture capitalists in charge of JAWS, still the most popular screen reader. The same economic realities have shown the Window-Eyes share, once equal to that of JAWS, drop to single digits.
Perhaps the most notable economic reality that the NFB approach ignored was that, because a screen reader is necessarily a niche product, the only way an independent company can make one and be profitable is by charging a real lot of money for each license. The NFB tact of working against an MS screen reader cost blind people, their employers and their educators millions of dollars that could have been spent otherwise if a no cost one existed. . NVDA, led by volunteers, saw the inequity of blind people needing to spend hundreds to thousands of dollars and its leaders took it on themselves to solve this problem while NFB ignored it entirely.
The Third Party Screen Reader Hypocrisy
As we’ve seen, NFB insisted that MS not do its own screen reader. NFB would later insist that Apple and Google build their own screen readers to provide out-of-the-box accessibility at no extra cost to consumers. I wonder if NFB learned from the chaos on Windows system and realized that it would be better if a screen reader was built into the operating system. Chong seems to praise the Apple experience which, regarding out-of-the-box accessibility is the best available.
In his piece, Curtis suggests that application developers at Microsoft should be tasked with testing against screen readers. This actually makes sense on products from Apple and Google as every developer at those companies have a screen reader they can launch easily and, if tasked with testing for accessibility in their project plans, they need only test against a single user agent. Chong seems to suggest that developers at Microsoft, instead of testing for compliance with the accessibility API, also perform functional tests against third party software. I think it’s absurd to think that developers at MS should try to test against software over which they have no control. If they were tasked with testing with Narrator, it would make sense, MS controls both the application and the AT; suggesting that developers and quality assurance professionals at MS learn JAWS and NVDA (the only Windows screen readers with enough market presence to warrant testing) and test against them is simply absurd.
Is NFBCS An Effective Advocate?
Chong writes, “Year after year, the National Federation of the Blind and the Microsoft Accessibility Team engage in active and ongoing communication, and year after year, we have communicated our frustrations and concerns to this team.” To which I ask, “If Curtis and NFB have been working with the MS ATG for more than two decades and, as Chong expresses in his article, the accessibility job remains mostly incomplete, are Curtis Chong, NFBCS and NFB itself actually effective advocates for this community?”
The Notable NFB Absences
I contend that Curtis and NFB have done a poor job of understanding the technology and, as a result, are ineffective advocates in this space. I didn’t quite know how I could express how an advocacy organization was “ineffective” as proving a negative is a logical impossibility so, instead, I thought I might list a number of very important areas in accessibility for people with vision impairment that, as far as Google can tell us, NFB has not participated. To this end, I used the search engine to look for terms important in technological accessibility with “National Federation Of The Blind” and/or “NFB” in the search terms.
WCAG 2.0 is the single most important set of guidelines for Internet accessibility to all people with disabilities, including we blind people. I googled, “+NFB WCAG 2.0” and found that Google gave us 7 results, zero of which were on an NFB related site. I then googled, using “National Federation Of The Blind” in place of “NFB” and did find one link to an NFB site in the top ten results and it was the consent decree in an NFB lawsuit requiring the defendant to follow WCAG 2.0. I tried a few more search terms and found identical results, it is obvious that no one from NFB, not even Curtis the president of the computer science subgroup, participated in the development of the standard and that there isn’t a single article on the NFB web site explaining this somewhat complex and definitely esoteric set of guidelines. Searching on these terms without including NFB provides one with a panoply of tutorials and other useful information from the entire world of accessibility but, sadly, none of it comes from NFB.
The Web Accessibility Initiative (WAI) has another standard called Aria. Web developers use Aria to include the semantics that a screen reader can use to tell users about complex web applications. In brief, if you use a complicated but also accessible web application like Microsoft Office Online (quite accessible albeit a bit sluggish with NVDA and FireFox), you are enjoying the work the developers did using Aria. So, I googled “+NFB WAI Aria” and, guess what? I found zero entries on any NFB sites. When I spelled out the name of the organization, I find exactly one search result on an NFB page and, once again, it is about legal frameworks and not technology. Searching without the “+NFB” provides one with another large list of tutorials, analysis, explanatory information about Aria from everywhere in the world of accessibility but not NFB.
Microsoft’s accessibility API is called User Interface Automation, if Curtis and NFB are so concerned about the accessibility of Windows applications, surely they must provide readers of the NFB web site with information on how to ensure their applications comply with the API, right? Wrong. If you google “+NFB User Interface Automation” you will probably, as I did, get zero results and two advertisements for contractors who do work using UIA. No matter how we search, we can’t find anything from NFB on this important piece of technology.
The 21st Century Video and Communication Accessibility Act of 2010 (CVAA) is the most important bit of new legislation regarding disability and technology to come along in quite some time. As with all such laws, the agency charged with enforcement must hold a public comment period to determine how it should proceed with the wishes of Congress. During the CVAA public comment period, Pratik Patel (then Curtis’ approximate equivalent at ACB), on behalf of his advocacy organization, filed hundreds of pages in public comments. The combination of NFB, NFBCS and Curtis Chong filed exactly zero. I’m neither a member nor a promotor of ACB but, on this very important task in ensuring that CVAA will be enforced, Pratik and ACB took on a leadership. I’ll assume that the reason NFB, NFBCS and Curtis made no comments was that they were overwhelmingly impressed by Pratik’s genius in this matter and were happy to have ACB speak for all blind people.
The Section 508 Refresh was passed by Congress and opened for public comment. Again, Pratik Patel and the ACB wrote up a ton of documentation and filed it in this important matter. Heck, on behalf of the Free Software Foundation (not a group known for its stellar record on accessibility), I filed a few pages of comment on 508 Refresh. The beauty of public comment is that it’s public so anyone can search the records and discover that NFB filed nothing on this matter. A friend who had attended the public hearings told me that NFB people did attend those sessions but that their only contributions could be summarized as, “blind people need to be involved in the process,” which was already true when they said it and, “NFB speaks for blind people,” which I contend is false as they don’t speak for me. I said the NFB deferred to Pratik on CVAA so I’ll suggest that Curtis and the NFB must have found my comments so brilliant, so illustrative that they chose not to do any of their own and let me speak for the community.
I could go on but, at this stage, I think you get the point. NFB and NFBCS, under Curtis Chong’s leadership, has steadfastly refused to participate in the most important developments in access technology. If, indeed, NFBCS, NFB and Curtis Chong have not contributed to the development, promotion, explanation of these and other extraordinarily important areas in accessibility, they are irrelevant as leaders. It’s easy to write articles like Curtis’, it’s easy to complain, to bitch and moan but it takes actual work to be part of the solution, work that Curtis and NFBCS have thus far refused to join in doing.
Gonz Gets Pedantic
I try to do my best to post articles that are as factual as possible. I may draw a controversial conclusion and use fairly aggressive prose when I express an opinion but, whenever a factual error is presented to me in an article I had published previously, I add a correction to the piece.
This blog is different from Curtis Chong’s articles in Braille Monitor (BM) for a lot of other reasons as well. First, I do not claim to be representing anyone other than myself and those who have given me explicit permission to speak for them. Curtis, in his role, claims to speak for “the blind” which, arguably, would include me. Second, Curtis, as is obvious by the introductory section in his article, claims to speak with authority, apparently derived from talking to other blind people; my own blog profile states that I am a loudmouth, crackpot stoner, I don’t claim any authority or expertise, I let my words speak for themselves and allow the readers to draw their own conclusions. Third, I allow you to post comments on this blog and no NFB publications, including Curtis’ articles in BM, permit any public discourse. If Curtis allowed for comments, we could have had this conversation as dialogue and the rest of you could have contributed as well but NFB sorts speak from “on high” and discourage interaction.
I, therefore, feel it is reasonable to hold Curtis Chong to a higher journalistic standard than I do even myself. He claims to be speaking for all of us and I, therefore, think he should be more careful with the way he states things. To wit:
Curtis writes, “Today only a small percentage of Microsoft products are regarded by the blind as comfortable and intuitive to use…” and as far as I know this may be true. As far as I know, this statement is false. I would like to know what was Curtis source for the things he states as fact, namely, the term “a small percentage.” I would also like to know Curtis’ definition of “the blind” in this sentence as I cannot find a supporting document in my googling.
Curtis, in the same section, continues by stating, “well over 80 percent of Microsoft products remain inaccessible to non-visual users.” I googled using as many terms as I could and could not find this 80% number published anywhere. I’m also curious as to the definition of “accessible” in this context. Did Curtis or others around the NFB actually test every program from MS and, if so, where did they publish their results. I dislike magic numbers when included in prose and, when used to discredit a corporation’s efforts, I believe that such numbers should not be used without a verifiable source as doing so is just ad hominem.
Chong writes, “There does not appear to be any user-experience research being conducted by Microsoft into improving efficiency for keyboard-only users, including the blind.” First, blind people also use crazy wild new fangled things like touch screens and track pads as well as keyboards these days. Second, many years ago, NFB itself published a valuable bit of UX research and is the command set still used on most braille keyboard based devices; as far as I can tell, with Curtis at the helm, NFBCS and NFB have not published any UX research in this area either. Researching user experience for keyboard only users would provide an excellent resource to Microsoft but also to Apple, Google and every other company that hopes to include effective keyboard control of its products. Perhaps, Curtis should be asking, why has it been so many years since NFB published actionable user experience research?
Curtis includes an oddly rambling paragraph on MSAA and UIA, the accessibility API in Windows. He writes, “the screen-access software vendors (very small companies in relation to Microsoft) had to devote considerable resources to make this happen. It would be better if these relatively small companies could spend more time and effort coming up with innovations that improve the efficiency and productivity of blind users of their software.” Or, one might say, if NFB hadn’t pressured MS into not making its own screen reader, we might actually have a company who can afford to keep up with OS releases like Apple can with its VoiceOver software? Why, decades later, Chong still insists that the broken system of high priced third party screen readers should continue is baffling.
In the same section on Windows accessibility API, Chong also neglects to state that FS actively opposed using standard API at all. I am part of the guilty party in this as, when I worked at Fs and for a few years afterward, I argued that an API solution could never provide the kind of accessibility that we could with JAWS using proprietary techniques. We argued vociferously that, rather than a generic API, applications should expose a VB like programming interface so we, the third party screen reader developers could craft custom solutions for each separate program we cared to support. An API solution was fine for simple applications but something fancy, Excel for instance, would always do better if we could write highly customized scripts for the UX. When HJ became FS, we stopped investing as heavily in JAWS development and it was our lack of investment in further support using these non-standard techniques that resulted in deteriorating in application accessibility, not MS. It was FS who rejected MSAA approaches and chose our own non-standard route to accessibility. You can’t blame MS for deteriorating accessibility in the third party screen readers which are entirely beyond its control. If you want to blame anyone, blame me, I fought hard against API back then, so did Glen Gordon and Eric Damery. MS was right, we were wrong.
Curtis writes, “For years Microsoft has left the blind with no access to Windows phones.” This is not true with phones based in Windows 8. I haven’t tried a Windows phone myself and reports from the field say it is a bit sluggish but, if the word “accessible” can be applied to Android as Chong does, it should also be applied to Windows Phone 8 as regards the more than 90% of blind technology users who prefer a synthesized voice interface, Windows Phone does not yet support refreshable braille devices.
About the MS Bitlocker software, Curtis writes, “A blind employee who is required to use a computer with Microsoft BitLocker installed will be unable to turn the computer on and get it running—not to mention use it.” This is strange coming from an advocacy organization that opposes accessible money, beeping traffic lights and other structural bits of accessibility. The fact is, Bitlocker is not “accessible” under any known definition of the word but, as I know a whole lot of blind people whose jobs require using the software daily, suggesting that it is impossible for a blind person to use is misleading. I asked a friend how she used it and she told me, “I turn my laptop on, I wait a little while, I type in my PIN and hit ENTER, my computer starts.” Yes, Apple has made their similar technology accessible and MS should as well but, as many blind people can work around it, it isn’t a functional impossibility as Chong suggests.
I believe that a “leader” in this space, someone who by his statements that he speaks for “the blind” should be much more careful in their publications. I’m a crackpot blogger, Curtis is publishing in an official organ of an advocacy organization that claims to represent our community. I think he should be held to a much higher level of journalistic standards and, as I illustrate above, Chong’s article is filled with problems and outright factual errors.
The community of blind computer users need effective advocacy but NFB, NFBCS and Curtis Chong have demonstrated poor judgement on technological, economic, political and structural issues of the gravest importance to this community. They have not participated in the most important discussions regarding standards, guidelines, API, user experience or anything else in this space. NFB seems to do nothing to promote use of accessibility development tools or standards compliance on any platform including Windows and provides none of the useful explanatory materials a developer hoping to make his work accessible might search on. As far as I can tell, regarding advocacy on technological matters, NFB, NFBCS and Curtis Chong have been present but irrelevant for a very long time now.
To clarify, this article is specifically about leadership and advocacy and discusses the Curtis Chong, NFBCS and NFB as spokespeople for our community. NFB does many other things including having funded the development of KNFB Reader, a really terrific iOS app that I enjoy using frequently. Unfortunately, KNFB Reader is a rare exception in a very large organization.