COMPUTER SCIENCE UPDATE - SUMMER, 1994 Published By National Federation of the Blind in Computer Science 3530 Dupont Avenue North Minneapolis, Minnesota 55412 Phone: (612) 521-3202 Table of Contents Message from the President by Curtis Chong Minutes of the 1993 Meeting of the NFB in Computer Science by Mike Freeman Message from Microsoft Correspondence on the GUI Problem Letters of Support from NFBCS Message from the President by Curtis Chong Greetings to the members and friends of the NFB in Computer Science. As you doubtless know, we have not produced a newsletter in almost two years. Much of the reason for this rather significant lack lies in the demands and other pressures placed on the NFBCS president. I am hoping to remedy this problem by finding someone who would be willing to serve as the newsletter's editor. The articles in this issue of Computer Science Update deal primarily with an issue that is of concern to the NFB in Computer Science--namely, the graphical user interface. Last year, at our meeting in Dallas, we heard from four companies doing work in this area. By the time of our meeting in Detroit (less than a month away), I expect that a number of companies will have announced that they are either working on or distributing programs to access Windows 3.1. So, at least with respect to Windows, we are seeing some progress. Of course, everybody is concerned that the progress we have made will be nullified when the next release of Windows (code named Chicago) is released to the general market. In this regard, we will be able to hear from someone representing Microsoft at our Detroit meeting. All is not sweetness and light with respect to X Windows. Despite the growing acceptance of Unix in the workplace, we have yet to see any commercial access product for the blind computer user who needs to run an X Windows application. The Disability Action Committee for X (DACX) is now working on the promulgation of standards designed to support hooks in X that would allow screen reading programs to obtain the information they require. I am happy to report that a good bit of progress has been made here. Beth Mynatt, from Georgia Tech University, is trying to get a commercialization grant to push forward the work she has done with something called Mercator, the first ever access technology for X. Although I was pleased to write a letter of support in my capacity as president of the NFB in Computer Science, the grant has not yet been approved as of this writing. As many of you know, IBM has Screen Reader/2, the only screen reading technology for OS/2 Presentation Manager, its graphical operating system. Last fall, when it appeared that priorities within IBM might be shifting and the Screen Reader/2 project might be severely diminished, I wrote a letter of support on behalf of of the NFB in Computer Science. So, as you can see, although I have not done as well as I would have liked in terms of getting a newsletter out, I have been able to establish the NFB in Computer Science as a positive voice of support for the projects that are of importance to blind people needing and wanting independent access to various computer platforms. Minutes of the 1993 Meeting of the NFB in Computer Science by Mike Freeman The 1993 meeting of the National Federation of the Blind in Computer Science was held on July 5, 1993 during the annual convention of the National Federation of the Blind (NFB) at the Hyatt-regency DFW hotel in Dallas, Texas. PRESIDENT Chong opened the meeting at 1:11 p.m. President Chong announced that IBM had donated a copy of either Screen Reader for MS-DOS or OS/2 to be given away in a drawing to be held at the end of the session. Another copy of either Screen Reader for MS-DOS or OS/2 has been donated to NFBCS; this will be given away during the NFB convention banquet. These programs will be supplied either with or without an adapter card. David Andrews announced that TeleSensory will be giving away twelve copies of ScreenPower as door prizes during the convention; the program will come out sometime during the fall. President Chong said he would discuss developments at Microsoft. He announced that a representative from Microsoft (one Greg Lowney) would be speaking at the meeting of the NFB Research and Development Committee, to be held later in the week. Mr. Lowney currently is speaking to VIDPI. The first program segment dealt with Optical Character Recognition (OCR) technology as it affects the blind. In his introduction, President Chong stated that he gets calls from many blind persons who want to buy reading machines. They don't have the best training in blindness skills or attitudes about blindness. They think that a reading machine will obviate the necessity to use readers. This is, of course, not true. However, reading machines can be useful. "Optical Character Recognition Technology and its Practical Application for the Blind" was presented by Bill Schwegler of Arkenstone Corporation. He said that new generations of both standalone and PC-based reading machines are being introduced. Consumers will see more choices of products, increased machine performance and reduced costs. Arkenstone's new family of products (introduced during the last year) typically are 40% less expensive and have improved performance, he said. Improvements in OCR technology during the last year and the advent of more powerful PC's at a reasonable cost were responsible for these product improvements. As more powerful PC'S capable of doing sophisticated analysis of scanner output become more affordable, commercial vendors are increasingly adopting the PC-based approach. However, card-based products have not disappeared. TRUESCAN-600 half-size cards are still available. But in general, the focus has shifted from hardware-based to software-based products--both for standalone reading machines and PC-based systems. New reading machines have automatic contrast selection and more automated page-orientation. There will be more "one-touch" reading machines. The most important thing a person must do when considering which reading machine to buy, said Mr. Schwegler, was to test the machines on typical documents that the person will be reading. No reading machine is perfect. Next, Keith Loris, Vice-president for Technology of Xerox Imaging Systems, presented a talk entitled "Document Recognition, the Intelligent Reading Methodology." Mr. Loris said that his firm's mission was document recognition -- the translation of data on paper into forms useable by a computer. Documents are infinitely variable; this makes it tough for a computer. A reading system must decide what should and should not be read and in what order items should be read. Xerox applies many analysis "engines" to help in this process. Screen pixels are separated into homogeneous regions of space which are turned into text. Xerox has developed a font recognizer/lexifier (on which a patent is pending) which permits recognition to not be hampered by non-textual material. The lexical classifier extends the notion of a dictionary-based recognition system (which is text-based) to include non-textual objects. This same technology has yielded a quantum leap in language knowledge. This yields better performance out-of-the-box. In the question-and-answer session on this item, David Andrews said that none of the reading machines in the International Braille and Technology Center would read his bank statement! "IBM Continues to Attack the GUI Problem" was the next presentation item. Presented by James Thatcher, Staff Researcher, Thomas J. Watson Research Center, IBM Corporation, it concerned how one uses Screen Reader and what access to the GUI is like. Under OS/2, one can access MS-DOS applications, WINDOWS applications and applications written specifically for OS/2--all at the same time. One task is in the "foreground"--this is the task one is currently interacting with--and the rest are in the "background." In addition to knowing such text-based character attributes as color, in the GUI environment, one must also know such things as the font of characters displayed on the screen. To select a different application to be placed in the foreground, one chooses from a "window list" of background applications. Such GUI artifacts as dialogue boxes and buttons are spoken in a useable form (such as questions that can be answered). There is, however, a fly in the ointment. Mr. Thatcher said that a critical problem for access software developers is emerging: how does one find the focus of interaction for an application (such as where one types on an insertion bar or where the software cursor is)? Different applications use different graphical symbols to represent this. In answer to a question, Mr. Thatcher said that windows access via AIX was "going slowly." The next agenda item was the NFBCS business session, which began at 2:53 p.m. The minutes of the 1992 NFBCS meeting were approved as published in the NFBCS newsletter. President Chong said he had spoken twice before "the rehab folks" at conferences sponsored by Mississippi State University. "We all know how the braille agendas came out," he observed wryly. The NFBCS has been asked to help with research of the quality of adaptive technology. President Chong then announced that Microsoft had designated a person, Greg Lowney, to work on access for people with disabilities to Microsoft products. Mr. Lowney will come on Wednesday evening and will speak at the R&D Committee meeting. President Chong then read a letter from Mr. Lowney. In the letter, Mr. Lowney stated that Microsoft has started distributing specs and documentation on how Windows works to companies developing access software. The NFBCS Treasurer reported last year's balance as $1,568.09. The ending balance this year was $1,712.89. The report was approved. The next program segment concerned access software for Microsoft Windows. The first presentation was entitled: "Slimware Window Bridge, the First Solution to the Windows Access Problem," and was made by David Kostyshyn, President of Syntha-Voice Computers, Inc. Mr. Kostyshyn is not convinced that any access solution will be the total answer to the problem of access by the blind to the Windows environment. Syntha-Voice took a theoretical, conceptual approach to the development of access software for Windows. They developed the Slimware Window Bridge software without help from Microsoft. The aim was to provide Windows access with the feel of familiarity and comfort level of the MS-DOS environment. Slimware Window Bridge doesn't worry about the size of windows or box coordinates; it behaves as if each window uses the whole screen. (One must be prepared for any screen layout, Mr. Kostyshyn averred). One uses the same keystrokes to access applications as does a sighted person. There are the following limitations: some bit-mapped graphics cannot be accessed, some concepts (most notably in desktop publishing) which cannot be conveyed, proportional placement of data is tough to describe and video support is limited to 640 x 480 pixels. Kostyshyn would like to see the product developed to the point that a blind person could install Windows by him/herself. "OutSPOKEN, from the Macintosh to the PC," was presented by Peter Cantisani of Berkeley Systems, Inc. The access software uses the same interface to access Windows as is used by outSPOKEN on the Mac. Berkeley Systems is developing an access "toolkit" which will aid access software developers in creating their own access products for Windows. The Berkeley Systems product supports most major synthesizers (some allow different voices for cursor, icons, text, etc). One can bring up outSPOKEN for Windows in three ways: from the Windows init file, by running the program from Windows or from the command line. The program is in "Beta 1" testing; that is, the system still crashes and the problems must be fixed before it can be marketed. Berkeley Systems hopes to have the product out in a couple of months. Berkeley Systems plans to port outSPOKEN to the UNIX environment. "Artic Technologies' Answer to the Windows Access Question" was presented by Dale McDaniel, Vice President of Marketing, Artic Technologies International. After discussing the "advantages" of Windows (e.g., multitasking and standardization of applications), Mr. McDaniel stated that a prototype of WinVision had been made available in March, 1993, and Artic now had a production version. Mr. McDaniel said Artic shared some of the experiences Mr. Kostyshyn had described. Artic had little contact with Microsoft. However, Microsoft has discussed Windows/NT with Artic. The problem is that the system was designed for the government with security considerations in mind. Microsoft has said they could put hooks in for access software. Mr. McDaniel said that WinVision loads easily, does not use a specific video driver (so any video configuration will work), and if one is running Business Vision under MS-DOS, one can install Windows independently. WinVision tracks the mouse, and there are keyboard mouse commands (Windows already has keyboard commands). Windows boxes, icons, buttons, checkboxes, etc. are automatically spoken. One can tell which font is being used. The last program item was "JAWS for Windows, Yet Another Answer to the Windows Access Question." It was presented by Charles Oppermann of Henter-Joyce. Mr. Oppermann said that the program was still in the early stages of development--the "pre-pre-alpha" stage. There are still many arguments about exactly how things should be designed and some strange areas to work through. The program will implement a "soft keyboard" via macros. The macro language will be redesigned into a procedural language. There will be implemented 26 integer variables, 26 string variables and 200 user functions. There will be 3 cursors: the "PC cursor", the "JAWS cursor" (the review cursor) and the "mouse cursor"; all cursors can be moved with the same commands. An applications language will obviate the necessity for users to employ macros. Henter-Joyce found Microsoft to be extremely helpful. The program is almost entirely written in C with about 5% written in Assembler. The program will read lines as lines (it uses the "base-line."). There will be field functions to inform the user of font type, etc. The program should be released by the end of the year (1993). The meeting was adjourned at 4:46 p.m. Respectfully submitted Mike Freeman Secretary National Federation of the Blind in Computer Science Message from Microsoft From the Editor: At our 1993 meeting in Dallas, we were not able to hear from Greg Lowney, Senior Program Manager, Accessibility and Disabilities Group, Microsoft Corporation, in person. However, Mr. Lowney was kind enough to send us the following letter. _________________________________________________________________ Microsoft Corporation One Microsoft Way Redmond, WA 98052-6399 Tel 206 882 8510 Fax 206 936 7329 July 2, 1993 Curtis Chong National Federation of the Blind Dear Curtis, Thanks very much for inviting me to speak at the NFB in Computer Science meeting at the annual convention. I'm sorry that I could not be there in person on Monday afternoon, but I appreciate the opportunity to say a few words in this remote fashion. I know that for many of you, Microsoft is a word not to be said in polite company. In fact, I'll even agree that much of the criticism has been earned. In the past, Microsoft has turned out many software products without really considering their accessibility to people with disabilities. My main message to you today is that we ARE trying to improve. This is my job: for the last year or so I have been running Microsoft's Accessibility and Disabilities Group, working full-time on making Microsoft's products and services more accessible. I'd like to give you some examples of the work we are doing, but first I think I'd better address what is probably the most pressing concern, and that is Microsoft Windows. Windows has probably done more than anything else to earn Microsoft the enmity of the blind community. Microsoft has been both hated and feared by many people because we were promoting a graphical operating system without making sure that it could be used by people who are blind, and the results have been disastrous for many people. I am not going to claim that Microsoft is innocent. As an insider I can partially understand how these errors came about, but I cannot justify them. Let me explain our perspective on Microsoft Windows. At the most basic level, Windows is a graphical operating system that runs on top of MS-DOS. It allows you to continue running MS-DOS applications, but it also enables a new class of Windows- based applications. Windows provides a much richer set of services than does MS- DOS, with the result that applications can avoid reinventing the wheel for each application. Windows offers much better memory management than MS-DOS, allowing you to run networks, full-featured applications and accessibility aids without running into strict memory limitations of MS-DOS. Windows is also a graphical environment. Windows was, to many people, also inaccessible. Why? Most people would probably give the following answers: because it is graphical; because there weren't any screen reader utilities; because Windows-based applications use a complicated interface involving menus, dialog boxes, buttons, scroll-bars, and so forth; and probably because Microsoft didn't care. Well, it's true that Windows is a graphical operating system, but don't let that mislead you. "Graphical" has been a bad word in the past because it was usually impossible to make graphical applications accessible through speech or Braille output. That's because MS-DOS doesn't provide any tools to help, so each application had to do its own graphical output in its own way. No standardization meant that there was no way accessibility aids could work with these products. Windows is a graphical environment because the underlying output technology is based on graphics images, rather than on fixed rows and columns of discrete characters. Just like graphical MS-DOS applications, once a letter is drawn on the screen it is reduced to a graphic image and no one but the application that drew it remembers what it represents. But that's where the similarity ends. Windows offers a richer set of services to programmers, so they don't have to do things themselves. They don't draw text and graphics themselves, they ask the operating system to do it for them. That is a big improvement over MS-DOS, and that's what makes it possible to write utilities that work with almost all applications. It is true that writing a screen reader for Windows is a much harder proposition than in a text-based environment, because the operating system doesn't remember the screen information for you. A screen reader is burdened with the extra responsibility of keeping track of this information, in addition to presenting it to the user in a useful way. But it is possible. If that's so, then why did it take years for screen readers to become available for Windows? Did we finally do something to make it possible? No, the accessibility aids that are now on the market could have been written years ago, and if they had been then all our lives would have been a bit less stressful! But every time a new computing environment comes out, it takes a while for third-party products to adapt. It took time when MS-DOS came out, and some platforms are still inaccessible. And face it, for many years Windows was a flop, and no one took it seriously. Also, for a long time everyone assumed that it was not possible because of the word "graphics", so no one tried. To be fair, it was not an easy task--Windows-based applications are complex and hard to write, so it is natural that Windows-based programs will be complicated as well. Now, Microsoft has taken steps to make this task easier. Last year we started distributing documentation to companies interested in developing Windows-based screen readers, explaining the issues and recommended techniques. This document has helped "jump start" many developers and resulted in more products available in a shorter period of time. Right now I work closely with five companies who have publicly announced such products, and more who are still working quietly. We are also listening to these companies, and modifying our future products to help them build better and more reliable screen readers. And we're helping them migrate their utilities to our new operating system, Windows NT, so that this time we won't have a long lag before the accessibility field catches up. I'm often asked whether Microsoft will be developing a screen reader of our own, or in some cases told vehemently that we should. We currently have no plans to do so; we feel the community is best served by having a diverse set of products from which to choose, and the best way to attain this is for Microsoft to put our energy into empowering and supporting all independent software developers. The last thing I want to do is compete with them or put them out of business. Finally, let me address the question of whether Windows-based applications are inherently less accessible because of their more complex user interface. Certainly they are more complex than the old command-line oriented programs. But the world does not stand still. Even without Windows, we're seeing virtually all applications move to complex displays designed to convey more information to the sighted user as quickly and intuitively as possible. We're finding more applications using graphics, and a graphical application for MS-DOS is truly inaccessible. Only by moving these graphics applications to a common platform, where the use of graphics is standardized and which itself can be made accessible, can we allow blind users to continue to access the majority of applications. The big advantage of Windows is that it offers a greater degree of standardization. Almost all Windows-based applications make use of the standard user-interface elements, and that is a big improvement over MS-DOS. When a screen reader utility understands the standard Windows interface it automatically understands much of all applications. And this standardization benefits ALL users because once you learn to use a single Windows-based application you have the basic skills to find your way around ANY of them...or almost any. Of course, not all Windows applications are as "well behaved" as we'd like, and some of the rules that help them work with screen readers are, unfortunately, unwritten rules. That's why one can't get too upset at developers who ignore them. And that's why Microsoft is currently developing programming guidelines which tell software developers how to make their applications more accessible. When these are released, application vendors will not be able to claim ignorance as a defense! But as strongly as we word these guidelines, they are recommendations only. Microsoft cannot force compliance. If we are going to get all software developers to follow them, it will still be in large measure up to organizations such as the NFB to help us drive home the urgency to each and every company. And that includes Microsoft, because this is, after all, a very big company that's made up of lots of little companies and working on well over a hundred separate software products. As with any bureaucracy, it's going to take years before all the problems are corrected. In closing, let me just throw in a few other notes about what we are doing at Microsoft. In the present, we are already making all of our product documentation available to Recording for the Blind, so they can be distributed in accessible format. We're also working with them to automate the production system, so that we can make ALL of our titles available at low cost and in a timely fashion. We're also members of the International Committee on Accessible Document Design, working on standards to make more documents available in accessible format. In order to make sure we're doing things right, we're including accessibility companies, organizations, and users with disabilities in beta test programs for our future products so we can get their input before the products are finalized. And Microsoft's working towards "Information At Your Fingertips", which means improving the user's ability to sort through and utilize the vast amounts of information that are available in a digital environment. Today, I hear from many people that the best feature in MS-DOS is the /B switch on the DIR command, which lets you get sorted directory listings. Extend that to being able to locate any document on the network by name, by the author, or even by a combination of words inside it. That's an example of the technology of the future. And Windows is going to become increasingly object-oriented, both to the user and on the inside. We hope in this way to come closer to one of our long-term goals: screen readers should not need to be trained or given profiles to adapt to each new application. Rather, the application should provide all that information to the screen reader as part of its standard design. And Microsoft is also working on voice-input and voice-output technology which should help make a more flexible operating environment adaptable to a variety of input and output styles. In short, the computer should adapt to the individual, rather than the other way around. The future is going to be based on information, and Microsoft is, today, at least, trying to make sure that information is accessible to everyone. Don't judge us by the past, but if you see a problem be sure to let us know. Greg Lowney Senior Program Manager Accessibility and Disabilities Group Microsoft Corporation Voice 206 936 8510 Fax 206 936 7329 TT/TDD 206 936 2627 Internet greglo@Microsoft.com CompuServe 70714,1542 Microsoft Corporation is an equal opportunity employer. Correspondence on the GUI Problem From the Editor: As Federationists know, I spoke at the second US/Canada Conference on Technology for the Blind on the topic "Problems and Chalenges of the Graphical User Interface." This talk was reprinted in the January, 1994 edition of the Braille Monitor. Some time after the January Monitor was published, I received a note via electronic mail from Christopher Chaltain, a National Federation of the Blind scholarship winner and an employee of IBM. Mr. Chaltain has had extensive experience with applications using the graphical user interface. He has used and developed such applications. Mr. Chaltain felt that in my speech, I had not given enough credit to IBM for its work on Screen Reader/2 and, further, that the tone of my speech was unduly pessimistic. I responded by saying that IBM had indeed done a tremendous job in the development of Screen Reader/2 but that, nevertheless, blind people still had many challenges to overcome before we could achieve full access to applications using the GUI. Here is the exchange of correspondence. _________________________________________________________________ March 20, 1994 From: Christopher J Chaltain To: Curtis Chong Subject: Comments to your remarks at the technology conference Dear Curtis, In the Braille Monitor, I read your comments you made at the meeting of the Joint Committee on Technology held at the National Center in 1993. I have some critical comments on your presentation and other remarks you made at the meeting. First, you stated that IBM first demonstrated Screen Reader/2 at the 1992 national convention held in Charlotte, North Carolina. Actually, Jim Thatcher demonstrated a prototype of Screen Reader/2 in 1990 at the national convention held in Dallas. Furthermore, in 1991 in New Orleans, IBM was demonstrating a version of Screen Reader/2 for version 1.3 of OS/2. This version was also available on a limited basis to IBM customers. What was unique about the demonstration in Charlotte was that it marked the general availability of Screen Reader/2 which ran on OS/2 2.0 and provided support for MS Windows applications. Second, you implied that Microsoft's interest in making MS Windows accessible began with the passage of the Americans with Disabilities ACT (ADA). Unfortunately, the structure of your presentation also had you referring to IBM in the surrounding comments, which could lead the listeners to infer that IBM's efforts were also a result of ADA. As you know, this is not the case. Besides the dates mentioned in my previous paragraph, I have been aware of Jim's efforts to make OS/2 accessible to the blind since 1988, and I believe he started earlier than that. This shows that IBM's interest in making OS/2 accessible predates the passage of ADA. The accessibility of OS/2 by the blind is more a result of Jim Thatcher's drive and efforts than any other single cause. Third, when the subject of the American Printing House for the Blind (APH) producing a tape to instruct the blind on using the graphical user interface (GUI) came up, you recommended that the target system should be a Macintosh with outSPOKEN, since that is the only graphical user interface accessible to the blind. I have never used a Macintosh running outSPOKEN, but I cannot imagine that it is more accessible than an OS/2 machine running Screen Reader/2. I have been using OS/2 and Screen Reader/2 exclusively since the first month of 1991. Not only has this allowed me to advance in my career, but it has made me a more efficient and productive employee. Finally, the overall tone of your remarks was negative and pessimistic regarding the blind's use of the graphical user interface. I, on the other hand, am optimistic and encouraged by the work done by IBM, Berkeley Systems, and others. There is a reason that the graphical user interface is becoming so popular among our sighted colleagues, and those very same reasons make it an exciting opportunity for the blind computer user. As I stated above, I am a more efficient and productive computer user because of my access to a GUI. Furthermore, I have access to applications I never could have accessed under DOS like the desktop publishing software, FrameMaker for Windows. Under DOS this WYSIWYG application would have been totally graphical and unaccessible to the blind user. Obviously, the blind user faces some challenges with the graphical user interface. It is not as intuitive for the blind user as it is supposed to be for the sighted user. However, once the blind user has mastered the additional complexity of a GUI and the associated access application, the benefits of the Common User Access (CUA) standards and multi-tasking make it well worth the effort. I guess I was particularly distressed by your comments. I, like many blind computer users, hold you in high regard and value your opinion. I was under the impression from your comments at previous conventions that you were impressed with the work of Jim Thatcher and IBM to make OS/2 and MS Windows accessible to the blind. I was also under the impression that you had an open mind to the benefits the GUI could have for a blind computer user. None of this came out in your comments, at least not in my reading. Would it be possible for me to get a copy of your remarks? I am sure they are available somewhere on the Internet or on some bulletin board. I would like to pass them around to a few people to see if my comments are shared by any of the other blind GUI users. Christopher _________________________________________________________________ March 21, 1994 From: Curtis Chong To: Christopher J. Chaltain Subject: My Remarks on the GUI Dear Chris: Thank you for taking the time to write to me with your thoughts concerning my speech on the Problems and Challenges of the Graphical User Interface (GUI). I am always glad to receive constructive and thoughtful criticism about the articles and speeches I write. First, I would like to set the record straight concerning my view of Screen Reader/2 in general and my high regard for Jim Thatcher in particular. In many informal conversations, in person, on the phone, communicating over the Internet, on NFB-NET, or in the CompuServe Disabilities Forum, I have expressed the belief that IBM deserves a lot of credit for developing Screen Reader/2. As you so rightly point out, IBM's work on this program pre-dates the Americans with Disabilities Act (ADA). Screen Reader/2 is, today, the only screen access solution for blind people who want or need to use the OS/2 operating system. If a corporation selects OS/2 as its "platform of choice," blind people who are affected by this decision will be able to keep their jobs because of Screen Reader/2. And yes, I know that Screen Reader/2, together with OS/2, can provide access to applications designed to run under Microsoft Windows. As I learn more about Screen Reader/2 (having recently converted my office computer to OS/2), I am impressed by the amount of planning, forethought, and downright genius that has gone into the development of this software. I thank God, quite literally, for Jim Thatcher. Although he would probably not admit it, I believe that he has been the inspiration, the driving force, and the architect for the entire Screen Reader project. Jim Thatcher possesses great personal warmth, public charm, tremendous enthusiasm, and intuitive genius. I have made no secret of my high personal regard for him. This is why, year after year, I have invited him to speak at annual meetings of the National Federation of the Blind in Computer Science. Each and every year, Jim has never disappointed me. He always has something interesting and thought provoking to say, and he is always upbeat about the potential for blind people to use the GUI on the same terms as their sighted peers. Why then, in my speech, did I not devote more space to IBM and Screen Reader/2? Why did I fail to mention the fact that in 1990, and again in 1991, prototypes of Screen Reader/2 were demonstrated at Federation conventions? Why did I not express more optimism about the potential for blind people to use GUI applications? I think to understand why my speech turned out as it did, you have to be aware of the context in which it was presented and the audience I was attempting to address. As you know, the Second US/Canada Conference on Technology for the Blind brought together people from four groups: leaders from the field of work with the blind, leaders from organizations of the blind, leaders from companies manufacturing or marketing specialized technology for the blind, and representatives from the principal computer companies in private industry having a major effect upon the ability of blind people to use commercial software (i.e., IBM and Microsoft). Because of the diverse nature of the group, it was difficult for me to come up with points that would mean the same thing to everybody. I wanted to shake up the rehabilitation professionals--to stop them from wallowing in DOS-based solutions for their blind clients. I wanted to send a message to private industry to the effect that today, the GUI is still a problem for the blind, despite Screen Reader/2. I wanted to make the point that access to the GUI is not a matter to be considered once and then forgotten; it is something that must be considered each and every time a new operating system or application is developed. I readily admit that at the time my speech was written, I knew next to nothing about how to run a GUI platform. When I wrote my speech, I was getting a constant stream of queries from blind people around the country who were concerned that their jobs were on the line because of conversions to a graphical platform or application. The platform most frequently mentioned was Windows. I had no current and specific information from knowledgeable, articulate blind people (such as yourself) about the ability of OS/2 and Screen Reader/2 to provide real access to such software as Word for Windows or (in your case) Framemaker for Windows. I was thinking in terms of the average blind computer user--the computer user who could not get early releases of software from IBM, the computer user who knew nothing about the differences between DOS and an operating system using the GUI, and the computer user who was told to convert to Windows but not to OS/2. I do recall that in 1990 and 1991, prototypes of Screen Reader/2 were demonstrated at NFB and other conventions. I remember going to a Wednesday afternoon demonstration of Screen Reader/2 at the Charlotte convention in 1992. Because the 1990 and 1991 demonstrations were of prototypes, available only to a limited set of individuals, I did not regard them as having much significance in the over-all scheme of things. It was perhaps because of this perception that I did not mention them in my speech. The 1992 demonstration was quite another matter, as you know. By then, Screen Reader/2 was a viable product, soon to be generally available. That, I felt, was worth mentioning in my speech. And while we are on the subject, I did not actually say that the Charlotte demonstration was the first time IBM had ever demonstrated Screen Reader/2. My exact words were, "At the 1992 convention of the National Federation of the Blind, IBM demonstrated its screen reading system for the graphical OS/2 Presentation Manager." Although this statement fails to make note of the 1990-91 prototype demonstrations, it was never meant to imply that IBM had done nothing in this area until 1992. You say that the over-all tone of my remarks was negative and pessimistic. I would prefer to think of the tone as realistic. As you say, you have been using Screen Reader/2 and OS/2 Presentation Manager since the beginning of 1991. You have doubtless had access to Screen Reader developers, OS/2 support personnel, and perhaps even some intensive training. With all of these resources to help you, how could you not feel positive about the GUI and your ability to develop and use applications built around it? On the other hand, I and a growing number of blind people are only now beginning to use GUI operating systems and applications. In my case, although I am fortunate to have contact with some key IBM people such as Jim Thatcher, I found that I was not getting enough help on a day-to-day basis to understand the intricacies of this new graphical operating system, OS/2. No one where I worked could tell me how to manipulate objects on the OS/2 desktop without a mouse, not to mention learning about Screen Reader/2. I was continually frustrated by the fact that the documentation, even though it was online, provided very little in-depth information about how everything worked together. Installing a simple DOS application (WordPerfect 5.1) would have been far more difficult if I had followed the instructions in the OS/2 User's Guide. In the end, it became necessary for me to arrange to receive a week's worth of training from Frank DiPalermo, a Screen Reader/2 consultant. Fortunately for me, my employer was more than willing to pay for the training. How many other blind people do you suppose will find themselves in exactly this situation? Quite a few, I would bet. How many of them will be as fortunate as I was? I simply don't know. I think that it is also important to point out here that OS/2 is not the only graphically-based system that has created concern among blind computer users and professionals. More and more blind people want to know when a commercial access product will be available for X Windows applications. I have received complaints from frustrated blind Macintosh users who tell me that Berkeley Systems is diminishing its support for the outSPOKEN program. Do these problems cause me to be negative and pessimistic? I prefer to think of them as helping me to be "concerned." As you say, blind people still face challenges accessing the graphical user interface. Screen Reader/2 is one solution to the problem, but it is by no means the only solution--nor should it be. Screen Reader/2 and OS/2 are fine systems in their own right. Together they provide access to a wide variety of GUI applications. This is the message I have been communicating to blind people in a variety of forums. I chose not to promote Screen Reader/2 quite so heavily in my speech because I was attempting to communicate a different message. If you feel that I was not as positive about Screen Reader/2, OS/2, and IBM in my speech as you might have liked, I can only say that I have had other opportunities to demonstrate my support in ways that you may not know. For example, just a few months ago, on behalf of the NFB in Computer Science, I wrote a letter of support for the Screen Reader project when I learned that it was being re-examined by IBM top management. I circulated the letter quite widely and caused other blind people to write letters of their own. Under separate cover, I will be shipping you an electronic copy of my speech as an ASCII text file. Feel free to circulate it around to other blind GUI users or to anyone else you think would be interested in reading it. When you solicit reactions to the speech, I hope you will keep in mind what I have said here. Cordially, Curtis Chong, President National Federation of the Blind in Computer Science Letters of Support from NFBCS From the Editor: During the past year, the National Federation of the Blind in Computer Science has written two letters of support: one in support of the Screen Reader/2 project and another to support a commercialization grant for the Mercator project. They are reprinted here for your information. _________________________________________________________________ NATIONAL FEDERATION OF THE BLIND IN COMPUTER SCIENCE October 13, 1993 Shmuel Winograd, Director Mathematical Sciences IBM Corporation Thomas J. Watson Research Center P.O. Box 218 Yorktown Heights, New York 10598 Greetings: As president of the National Federation of the Blind in Computer Science, I would like to express my personal support and that of the organization for the work IBM has done and continues to do regarding Screen Reader. Particularly noteworthy is IBM's work on Screen Reader/2, the ONLY screen reading technology that enables blind people to use OS/2 applications without sighted assistance. I recognize that in the DOS arena, other competing screen reading systems are available to the blind and that, arguably, IBM should never have developed Screen Reader for DOS. However, it should be pointed out that without the experiences gained in developing and then marketing the DOS version of Screen Reader, IBM would have had a much more difficult task developing and promoting its Screen Reader for OS/2. As it is today, a lot of blind people, including myself, appreciate the fact that Screen Reader developers at IBM have done a better job with the OS/2 Screen Reader as a direct result of the experiences gained in the development and marketing of Screen Reader for DOS. In today's corporate data processing environment, an increasing number of systems are converting to the graphical user interface (GUI). Some of these conversions are to the Windows platform, and others, to OS/2. Inasmuch as OS/2 provides access to Windows applications, there is no reason why a blind employee (who may be adversely affected by any conversion to Windows) cannot use OS/2 (and by extension, Screen Reader for OS/2) to access the same Windows-based applications as his or her sighted colleagues. Although there are other companies marketing Windows access products for the blind today, none of them has the depth and experience with the GUI that has already been evidenced in IBM's Screen Reader for OS/2. As an ever-increasing number of corporations convert to the use of the GUI, IBM's role in making the GUI accessible to the blind through Screen Reader/2 and OS/2 will necessarily grow in importance. If I may, I would like to cite my own personal situation as an example. Last August, my employer, IDS Financial Services, made a sweeping platform decision. It was decided that the platform of choice for IDS would be IBM's OS/2 as opposed to Windows, Apple Macintoshes, etc. This decision has a significant effect upon everyone at IDS whose job requires the use of a computer. Even now, large scale conversion plans are being drawn up by a multitude of departments as we work to implement this major decision. If not for IBM's Screen Reader/2 and the significant development effort that went into this product, blind computer users at IDS, including me, would have been effectively shut out from the computer. At least now, we have the very real possibility that we will continue to be able to enjoy the independent access to corporate applications (such as the E/MAIL system with which I am sending this letter) that we have today. What I am trying to say in a round-about way is that IBM Screen Reader, particularly Screen Reader for OS/2 (and perhaps for other GUI platforms) is a vital component in the total mix of products enabling blind people to use critical applications in the graphical environment. For OS/2, Screen Reader is the ONLY choice we have, and it is a good choice. I encourage you and the other folks who are working at IBM to continue your essential and highly desirable efforts in this area. Yours sincerely, Curtis Chong President National Federation of the Blind in Computer Science _________________________________________________________________ NATIONAL FEDERATION OF THE BLIND IN COMPUTER SCIENCE March 15, 1994 Sheila Stanley Advanced Technology Development Center Faculty Research Commercialization Program 430 Tenth Street, NW, N-116 Atlanta, Georgia 30318 Dear Ms. Stanley: On behalf of the members of the National Federation of the Blind in Computer Science (NFBCS), I would like to take this opportunity to express my personal support and that of the organization for the work that is being done at Georgia Tech University on the Mercator Project, directed by Elizabeth Mynatt. I understand that there is a proposal being developed to produce a commercial system which would provide access to Motif applications on Unix workstations for users who are blind. I am writing this letter to express support for that proposal. Blind computer users today, whether on the college campus, in the workplace, or at home, have achieved a high degree of access to text-based applications running on IBM PC's and compatible computers using the Disk Operating System (DOS). Today, it is not surprising to find a blind person using the same word processor, communications program, spreadsheet, or data base system as his/her sighted peers. For example, I am writing this letter to you using WordPerfect 5.1 for DOS. However, blind computer users are being threatened by the growing acceptance and use of graphically-based applications--applications that run under Windows, OS/2, or Unix. I think it is safe to say that blind people have actually lost jobs because of this trend. Why? Because screen access technology has not been able to keep pace with the accelerating conversion of applications from a text-based interface to one using the GUI. This problem is particularly acute for blind people using Unix applications. Although standard communication programs such as Procomm Plus, CrossTalk, Telix, and the like permit a blind person to "Telnet" into a text-based Unix application, they do not provide any kind of access to graphically based programs. With respect to Windows and OS/2 Presentation Manager, there are screen reading systems that can be purchased by a blind person who needs or wants to use applications running under these software platforms. Not so with Unix and applications written for X. Today, there is not one single commercial screen access system that a blind person can buy to use a graphical application written under Unix. I can tell you that a lot of blind people around this country are very interested in the work of the Mercator Project. We would welcome any effort to make a commercial offering available in this area, and we hope that we would have something to say about how the system functions from the viewpoint of the blind computer user. I appreciate this opportunity to express my support and that of NFBCS for the effort to make a commercial X access system available. Rest assured that it is badly needed by a growing number of blind computer users and professionals around the country. Yours sincerely, Curtis Chong President National Federation of the Blind in Computer Science