This morning, as on most mornings, I sat down with my first cup of coffee to read my email and found two articles in my Blind News (link above) folder explicitly “blaming” people with vision impairments for the failure of the open source push portion of the Commonwealth of Massachusetts’ open document format (ODF) initiative. The first article, “Blind leading away from open source” ran in yesterday’s ZD Net out of Germany and the second, from the UK based Techworld.com, “Visually impaired prevent Massachusetts move to open source,” both propose the argument that we blinks should take the blame for the lack of accessibility in the open source office applications, “Open Office” and “Star Office” from Sun Microsystems. Both articles admit that the ODF plug-in for Microsoft Office meets the Commonwealth’s requirements of keeping data in an openly documented file format and that Microsoft Office is accessible to all users and doesn’t exclude those with vision impairment.
Blaming the possible victims of the ODF legislation, specifically people with vision impairments who could lose their government jobs in Massachusetts because they hadn’t tools that could, prior to the introduction of the ODF plug-in for MS Office, be used to perform their jobs. Citizens of the Commonwealth and others with vision impairment who chose to access this information would also have discovered barriers to entry without such a plug-in.
Discriminatory statements and blaming the victim, tactics common to most arguments against civil rights movements, simply do not hold water when placed in the light of scholarly scrutiny. If we go back to the middle of the twentieth century, we will find many statements suggesting that company X would hire African Americans if said minorities had the skills to perform the requisite tasks. Of course, as “said minorities” had no access to training for such a job, the validity of blaming the victim of discrimination for not having the skills is demonstrated to be void of reason. In this case, these articles blame blinks because screen magnifiers and screen readers do not work with the open source office applications. Thus, it seems that these people also believe that my transportation problems must result from the lack of a JAWS for Windshields that I could run on our Toyota to get around town.
The ZD Net article quotes Peter Korn, Accessibility Architect at Sun Microsystems (link to his blog above) and one of the smartest, most energetic and outspoken advocates for the digital rights of people with disabilities in the mainstream technology industry as blaming the AT companies for not investing enough in the open source office suites and, therefore, the relatively puny Freedom Scientific, AI^2, GW Micro and Dolphin Systems, according to Korn, should foot the blame and bill to remedy the situation. “If Freedom Scientific and GW Micro and Dolphin Computer Access (makers of JAWS, Window Eyes, and SuperNova respectively) were to make similar investments in scripting and customizing their assistive technologies for OpenOffice.org as they have for Microsoft Office, or if they were to improve their existing scripting and customizations for WordPerfect and Wordperfect were to support ODF, then screen reader users should have no accessibility barriers to equal productivity and efficiency with ODF as they have with Microsoft Office in Windows,” wrote Korn according to the article.
When I worked for Freedom Scientific, a job I left nearly 2 full years ago, Peter and I had this discussion in person, on the telephone and in emails. I have written at length about the general failure of the entire system of providing accessibility in the pages of Blind Confidential many times and, once again, find myself forced to argue against the intellectually vacant myth that one can blame the AT companies or their customers for the accessibility problems in products your much larger and wealthier company prefers.
Freedom Scientific, GW Micro and the others hardly have the resources to keep up with the latest releases of the number one market share products in every market niche. An equivalent in the racial minority example would be saying that if relatively small organizations like NAACP, SNCC and SCLC would only pool their meager resources and start their own General Motors, MIT and professional baseball league, they could integrate them at will. As I wrote in these pages in the past month or so, JAWS 7.1 came out with “support for Windows Media Player 10” listed as a new feature months after Microsoft had released Windows Media 11. Freedom Scientific, the leader in the business of making AT for people with vision impairments, doesn’t intentionally lag behind the mainstream releases for the joy of hearing a cranky asshole like me point out such failures as the collective inability of the AT industry to maintain pace with the mainstream but, rather, its relatively large software engineering team works very long hours sweating blood to put out the best and most current product they can in the current economic climate. I’ve no indication that my friends at GW Micro, AI ^2 and Dolphin Systems behave any differently.
Peter includes the closed source WordPerfect as another example of a product with poor support by the AT companies. WordPerfect, which also does not support ODF, had far better screen reader support in the DOS days than did Microsoft Word for DOS. Why? Because it had a far greater market share, had a little known and hardly documented DOS interrupt that a third party TSR application (like Visual Eyes or JAWS for DOS) or a grammar checker like Right Writer (for which I wrote the TSR interface) could access a lot of information, thus probably making it the first host of an accessibility interface and as, according to John Dvorak, “was a simple application, you went to the DOS prompt, typed ‘wp’ and very little changed.”
Then, the founders of WordPerfect Corporation sold their enormous baby for $1.1 billion to Novell in what was then the single largest acquisition of a private company. Novell played around with consumer software for a few years and watched the Microsoft Office market share surge past the once unbeatable WordPerfect as they mismanaged their entire consumer division. Ultimately, Bob Frankenberg, having killed most of the WordPerfect line of products, sold its remnants, except for the still mostly inaccessible Groupwise, to Corel for $100 million. I then wrote a letter to Bob, now the former CEO of Novell, suggesting that the next time he wanted to spend $1 billion and get virtually nothing for it that he should call me and that I would, for half a billion, guarantee that I would provide nothing of any value whatsoever, thus cutting out the overhead of managing such a loss. Bob didn’t respond.
At one point during my six year tenure holding the reins of JAWS development, we did spend a fair amount of time working with Corel to help them make their office suite work well with JAWS. The strategy we took, obvious to anyone who has a copy of JAWS who elects to open up and read the source code to the scripts for WordPerfect or Quatro Pro, was to advise Corel and coach them on implementing an object model identical to that in Microsoft Office which would let us leverage the existing investment in supporting the Microsoft suite to make theirs work in a similar fashion. Corel paid Freedom Scientific its hourly consulting fee to do this job as, otherwise, FS would not have had the financial wherewithal to work on a product other than the market leader.
When Sun first got into the Open Office/Star Office business, I discussed this strategy with Peter and his then supervisor Marney Beard (also a terrific individual with a ton of energy and great focus in this arena). I said that if they went with an object model identical to or similar enough to the one used by MS and Corel, that either FS or a third party or a group of volunteers or anyone else who cared to could write the scripts using the source to the JAWS scripts for the Microsoft and Corel suites as a template. Sun chose to take a divergent path.
In the period since then, Corel has let its object model support decay and, as a result, the JAWS support for it has also fallen into decline. Is this the fault of the relatively tiny Freedom Scientific or the billion dollar Corel? Also, in the interim, other screen readers, Window-Eyes and Freedom Box System Access for instance, have followed the FS example and have started deriving information from the MS object model to provide their users with information from the market leading office suite. If corel had chosen to keep its object model support up to date and hired their own JAWS scripter, they, for a minimal investment, could remain current with JAWS and, today, probably other screen access products as well. If Sun had elected to add a MS emulation object model, powerful but different from the gnome accessibility API, they could be in the same position and, with some of their current employees, one the former manager of the scripting department at Henter Joyce, could easily have provided an open source office suite with the excellent accessibility available to blind users of the Microsoft products.
Microsoft has never tried to stop any other company from copying the functionality of their object model. They were well aware that FS worked with Corel to provide a replica in Perfect Office and, indeed, employees in the MS Office team helped Corel and FS do this project.
I know Sun Microsystems doesn’t have the same financial power it once had but I’m willing to bet that the Freedom Scientific stockholders would do a one for one trade of companies if offered the possibility. Sun, compared to the entire blindness industry combined, is a giant. The AT companies do what they can to support their customers while supporting their investors who took the big risks to bring talking and magnified computing to market in the years before ADA and 508 forced the hands of the mega-bucks mainstream companies. People like Ted Henter, Doug, Ben, Jim Fruchterman and others shouldn’t be blamed for poor accessibility when an incompatible standard emerges but, rather, be applauded for bringing products as far as they have with so little financial ability.
I don’t want to blame Sun Microsystems or Peter for this mess either. Sun stands as a terrific example of good corporate citizenship in the disability realm and Peter Korn’s leadership has gone a very long way to make the world of computing profoundly more accessible for people with disabilities. The problem, as I’ve described in many ways from many different angles in these pages, has no simple solution.
Blind people, as a group, have never made any objections to open source software as the articles I read this morning would imply. These traditional victims of discrimination merely want to maintain the technological progress they have made so far. Perhaps if one tried to deconstruct the path Massachusetts chose to take to get to its goal of using an open document format the oft repeated problem of including people with disabilities at the end of the process rather than the beginning might pop out as the culprit in the process. I haven’t talked to my old buddy Joe Lazarro, formerly of the Massachusetts Commission for the Blind and now of the Commonwealth’s information technology bureaucracy, in a long time. I wonder, if when he held the position of Technology Director at the commission if anyone in IT called him to ask about compatibility and inclusion issues early in the process?
On the opposite side of the coin, many blinks have joined the open source movement as both advocates and hackers. JAWS, for the most part, exposes it’s most interesting components, nearly its entire user interface, in its scripts for which the source code comes included with every download. One can actually look at most of the code for the JAWS user experience without buying the product as the script source, where the majority of the interface comes from, can be had by anyone who downloads the free demo of the product. There is also a large community of JAWS hackers who, when FS chooses not to support a particular application, does so themselves. Recently, on the blind programming listserv, a group of we blinks have been working together to create a tutorial for using the Microsoft development tools with screen readers and a subset of the tutorial team has been writing an excellent set of scripts and distributing them with source code included to support Microsoft Visual Studio .Net 2005.
The community beyond JAWS also does some really great things in the open source realm. TV Raman’s emacspeak, an outstanding set of extensions to emacs, is a great example of open source development by and for blind users. If I remember correctly, Raman takes a totally free software approach to the world and will not use any product that does not carry the GPL and does so successfully as a blind person in the corporate world.
Others, including those who work on SpeakUp, BrailleTTY, and the many other varieties of kernel modifications for console based flavors of GNU/Linux also work very hard to bring open source solutions to our community.
Finally, major corporations, including Sun Microsystems and IBM, have built their own screen readers for the gnome desktop. Sun has ORCA and I can never remember the name of the one from IBM but I recently received an email that said it had been coming along nicely. Both of these gnome solutions were demonstrated at CSUN as alpha versions and both, when they reach a truly usable level, promise to provide full support for both Open Office and Sun’s Star Office so, if the blind people in Massachusetts can do their jobs as well or better with these solutions, they can move away from Microsoft’s solution to an open source solution when the screen readers for a completely open source graphical desktop reaches maturity.
The gnome accessibility API, one in which I, as Freedom Scientific’s representative, participated in the description of, takes the idea of an accessibility layer further than any I’ve seen from Apple or Microsoft. It includes what might remain the Holy Grail of accessibility solutions, a generic way of exposing contextual information, which, if proven to work, may truly mean the next generation of access for blind users can start to emerge. The operative point being “proven” to work which, in this context, means beyond working in the lab with some highly controlled applications and beyond being an outstanding demo that the market refuses to accept. Sun and the gnome desktop accessibility advocates need to sell this solution both to the community of disability advocates as well as to the market at large before the little guys like FS, GW AI, etc. can be expected to use their scarce resources to support.
I believe in the gnome API on a theoretical basis. I will write anything I can and add my voice to a truly generic standard for accessibility API layers that run on multiple platforms whether the standard turns out to be the one from Sun, Microsoft, Apple or a still unknown source. I will not, however, stand by and watch people who work for multi-billion dollar companies, state governments and major media outlets take cheap shots at the victims of discrimination as they excuse for their own short falls.
As much as the beautiful Massachusetts State Capital Building atop Beacon Hill cannot put out a “Whites only” sign so as to not inconvenience the still racist communities in Chucktown and Southie, they cannot start insisting on using file formats inaccessible by people with disabilities. The blind population should not be blamed but, rather, the disability advocates in the Massachusetts Information Technology bureaucracy should be applauded for stopping what could have forced a lot of people out of jobs before the deadline. The plug-in for MS Office is a reasonable compromise and should be viewed as a victory for people with disabilities and not a defeat for the open source movement.
This is a rare situation in which everyone can win so, instead of pointing fingers, slinging mud and playing the proverbial “blame game” perhaps all parties should take this as a wake up call to ensure that the needs of people with disabilities are considered from the outset rather than when the cow has nearly left the pasture.