OpenDocument Accessibility

David A. Wheeler

November 20, 2005

In recent discussions about OpenDocument there have been a number of discussions about accessibility. It's a good thing that accessibility is of such deep concern, but I fear that for many it's a political excuse instead of a real concern for those with disabilities.

There have been some good articles about OpenDocument and accessibility. Peter Korn has an excellent analysis, under the assumption where you want to use OpenDocument and refuse to use Microsoft Office. In short, even if you choose to not use Microsoft Office, in many cases there's no problem (in fact in some cases it's better than the alternative), and people are working the resolve the rest of the issues very rapidly. David Berlind has a good article about OpenDocument accessibility too.

Too many seem to be ignoring one obvious solution: use Microsoft Office and a plug-in or service for OpenDocument, if Microsoft Office has an accessibility option for that individual. Complaining that this would require a third party plug-in is a double standard; for many disabilities, supporting accessibility already requires a third party plug-in. Don't expect Microsoft's new Office suite to be fully accessible for some time; in every previous release, Office has not been accessible to many, and it takes a long time for third parties to re-engineer what needs doing. In fact, as Curtis Chong (president of the National Federation of the Blind in Computer Science) clearly notes, "accessibilty of Office-based solutions is largely due to the 'heroic' efforts of third party software developers whose software routinely breaks everytime Microsoft upgrades its software" -- Microsoft has never documented the interfaces necessary to support full accessibility. Massachusetts' ITD has clearly stated in their policies that using Microsoft Office and plug-ins would be fine.

I recently got a very interesting letter from Wesley Parish, who is himself disabled and who argues that fully open formats like OpenDocument should be required to allow full accessibility. Here's his letter. It's long, but worth it, and in particular look at the last paragraph:

I am disabled, having suffered a Traumatic Brain Injury, an extradural haematoma, in 1988 in my first year of University Study.

I prefer to focus on what I [can] do rather than what I can't, and so in order to understand what had happened to me, I bought and read some books like A.R. Luria's "The Working Brain", Muriel Lezak's "Neuropsychological Assessment" and Guyton's "Basic Neuroscience: Anatomy and Physiology" - light reading! ;) , thus making me a somewhat atypical TBI survivor.

Part of the consequence of that TBI was short-term memory loss, and a relatively long-term clinical depression that lasted five years. I survived by the skin of my teeth. A consequence of that also is long-term unemployment, which strikes me as strange - I mean, I suffered the accident and have preferred to define myself in terms of what I can rather than what I can't do. Whereas the employers seem to concentrate on what I can't do - in reality making themselves the more disabled - "We are surrounded by insurmountable opportunities!"

As part of my redefining myself following the accident, I took up computers and proceeded to learn Linux intermittently in 1994, 1996, 1997 and haven't looked back since then - though along the way I have picked up various other OSes such as MS-DOS, IBM's OS/2, MS' Win9x and the MS NT-based branch.

I do part-time volunteer work in a community-based computer training and learning centre, and manage a network of Microsoft Windows 98 machines. My favourite Office Suite on Microsoft Windows and Linux is OpenOffice.org. That is because it is there, unlike MS Office, which doesn't exist on the Linux platform; and because it doesn't have a problem with viruses.

My current default file format is RTF; that is because it works in approximately the same way on Microsoft Office and most other word processors; it is limited to word processors, however, so it won't be of much use to me if I ever get into business for myself.

Now what does a non-discriminatory, open standard file format mean to me? For a start, one of the few books on nursing TBI patients that I found in my parents' home town public library, stated that one of the major causes of post-accident depression was uncertainty about surviving financially after the accident. The book was published in Britain, where the state takes up what is called "unemployment insurance" in the United States.

I found that that was not an imaginary worry. The (NZ) Department of Social Welfare took a perverse delight in acting as if I was a malingerer until the University I had been studying at, informed them they were mistaken; then I got a version of a hurried apology.

I hate to think what it might have been like if the relevant laws had been in a file format that the relevant computer company had abandoned a few months before, and the relevant department had undertaken a hurried upgrade to the updated software which rendered most of the laws too much trouble to consult.

This experience colours my attitude to file formats. It is necessary for the disabled to have access to all government information relevant to them, in a file format that is readily available for as many different applications from as wish it, one that does not insist that one jump through licensing hoops in order to implement it, one that can be readily extended in the future according to need - and one that can not be used as an excuse by lazy bureaucrats to deny me my rights!

The question currently buzzing in Massachussetts is , "Does Open Document Format limit accessibility?" For myself, I find it does not. In 1994-97 when I gave myself a rather hurried form of computer education, reading books like Tanenbaum's "Operating Systems: Design and Implementation" and "Computer Networks", CJ Date's "An Introduction to Database Systems" and Frederick P. Brooks, Jr's "the mythical man-month" - light reading again! ;) - I found one of the most persistent concepts was a strict separation between data and executable code.

ODF provides that strict separation, defining data separately from the code. This means that software libraries that record and play the spoken voice, etc., can be used to manipulate documents stored in ODF. An open specification that allows ANYONE to implement accessibility solutions is the way to solve the problems of access by the the blind and other disabled. Otherwise, government data will be tied to specific programs and NOT accessible to all, and in time, NOT accessible at all.

Wesley Parish, wes.parish (at) paradise.net.nz

Obviously, every disability is different. But I think his closing point is important -- if we really want to make sure that information is available to all, then we need to make sure that information is stored in a format that allows "ANYONE to implement accessibility solutions." The current Microsoft Office binary formats are a poor solution, from that viewpoint; the secrecy of the format makes it more difficult to create effective solutions.

When I starting writing this, Microsoft was using an incredibly restrictive license on their proposed XML format (MOOX) that was clearly unacceptable for general use (it excluded legal use by competitors). As I write, Microsoft has announced that they are changing this license. Unfortunately, the true terms and implications are not yet clear. It may still be totally unacceptable, if it turns out to continue to exclude competitors. It'd be great if it did allow completely open competition (e.g., allowed GPL implementations, subsets, taking format concepts and adding them to other formats, etc.) At the very least MOOX will be suboptimal, due to the lack of lengthy outside review and correction. The release of Office 12 will not fix this; most accessibility solutions have to be extensively modified after release to work with a new product, so for quite some time it's likely that many of the disabled will be unable to use Office 12. MOOX is about to enter the ECMA process, and is a long way from release as a standard -- it's not clear how different the final result will be, if there ever is one. (Java entered ECMA, but its backer decided to halt standardization and it never became a real standard; billions of dollars were invested on that failed promise!) OpenDocument is an OASIS standard now, has undergone more review than Microsoft is even contemplating for MOOX, and there are products avaiable today. It's clear that using OpenDocument is a reasonable choice, even for the disabled.

The OpenDocument article on Wikipedia includes a section on accessibility, if you're looking for more.

Feel free to see my home page at https://dwheeler.com. You may also want to look at my paper Why OSS/FS? Look at the Numbers!, and my papers and book on how to develop secure programs.

(C) Copyright 2005 David A. Wheeler. All rights reserved. (Parish's letter is obviously his own; he's given me permission to republish it.)