Published Every Other Month by Raised Dot Computing, Inc., 408 South Baldwin Street, Madison, Wisconsin USA 53703. Telephone: 608-257-9595. Fax: (608) 241-2498.
Subscriptions: $18/year Print, $20/year Audio Tape, $30/year Apple II BEX data disk or MS-DOS data disk. (Kindly add $20/year for postage outside N. America.)
Single issues: $4 each (specify medium).
Submissions are always welcome, especially on diskette. All are subject to editing for style and clarity. All opinions expressed are those of the author. Editors: Caryn Navy and David Holladay.
Entire contents copyright 1991 by Raised Dot Computing, Inc., All Rights Reserved. Nothing may be reprinted in any medium--print, braille, audio, or electronic--without prior written permission from RDC Inc.
As we said in the last issue, our Newsletter survey revealed an almost universal desire for more information about new products. In this issue we are starting a new feature called "New Product Alert" which lists new products in the sensory aids field. We would greatly appreciate receiving reviews from users about these products.
We remind you about our new policy regarding technical phone calls. As of January 1, 1992, we will no longer deal with technical phone calls from people who are not subscribers to the Newsletter. Unless the Newsletter you are reading now is a free sample, you pass this simple test. In organizations where many people use sensory aids equipment, we only ask that each person making technical phone calls has access to recent issues of the Newsletter.
We thank all of you who sent us information in our Newsletter survey. We summon a drum roll as we announce that Sally Langston of Soldotna, Alaska has won the prize of $100 in credit toward RDC purchases in the Newsletter survey drawing. Congratulations from the lower 48 for being chosen from among 220 entries.
In February 1991 Raised Dot Computing began selling Hot Dots 3.0. Since then we have added a number of improvements. At this point we would like to announce Hot Dots 3.01. By now all Hot Dots 3.0 purchasers should have received a braille and print letter explaining how to get a free copy of Hot Dots 3.01 to replace Hot Dots 3.0. Simply mail us the letter plus your original Hot Dots 3.0 program disk (both program disks, main and supplemental, for the 5.25 inch package). We will mail you a copy of the Hot Dots 3.01 program. Don't return the manual disk. The manual is the same; there is a supplement to the manual on the new (main) program disk in the file {EXTRA.TXT}.
Below is a list of new features in the program. Please notice that most of these features require using Hot Dots from the command line. If you have been slow in graduating from using the {DOTS} menu to using Hot Dots from the command line, now you have more incentive.
Many people have commented on the lack of a facility in Hot Dots for printing multiple copies of a document. Comment no more. There is a new batch file in Hot Dots 3.01 called {MULTI} which allows you to print multiple copies. For example, to make 17 copies of a file called {FROG.BFM}, type in the following: {MULTI FROG.BFM 17 COM1 <enter>}. This directs the output to the embosser connected to COM1 (serial port 1). (Of course you can select other files and other output ports.)
You may want to print only a few pages of a document. For example, if your dog stepped on pages 51 through 78 of a document, you might want to rebraille only those pages. To do this, type the following: {VANILLA FLY.BFM LPT1 /B:51 /E:78 <enter>}. This would print out only pages 51 through 78 of file {FLY.BFM} through the embosser hooked up to the parallel port. {/B} stands for begin, and {/E} stands for end. If you typed {VANILLA FLY.BFM LPT1 /B:51 <enter>}, printing would begin on braille page 51 and continue to the end of the document. If you typed {VANILLA FLY.BFM LPT1 /E:78 <enter>}, printing would begin at the start of the document and end with page 78.
You can substitute a file name for the output device. This creates a new file containing only the specified pages. Let's say you want 17 copies of pages 10-19 of the file {FROG.BFM}. To do this, type the following lines: {VANILLA FROG.BFM TRASH.BFM /B:10 /E:19 <enter>
The desired output is sent to the parallel port. After you have finished brailling, you can delete the temporary file {TRASH.BFM}.
You can also view only a few pages of a document on the screen. To view pages 26 through 34 of the file {FROG.BFM}, type the following: {VIEW FROG.BFM /B:26 /E:34 <enter>}.
To find out how many braille pages a document has, type the following: {VIEW FROG.BFM /P <enter>}. The slash P feature (P for pages) suppresses all screen output except for the report on the total number of braille pages. The final report gives the correct number. (Hot Dots 3.0 gave a number that was too high by one.)
Some people want a way to make inkprint output that shows the inkprint together with the braille dot pattern on a line-for-line basis. You can now do this with the Hot Dots Dipner driver. You get a line of inkprint and then a line of braille dots. This new feature is available when you add slash D to the {DIPNER} option. The D stands for dual output.
To get output on a Diablo 630 (or compatible letter quality printer), use {DIPNER /D}. To get output on an Epson printer (or compatible dot matrix printer), use {DIPNER /E /D}. To get output on an Apple ImageWriter, use {DIPNER /I /D}. There are two ways to use this driver program. You can use it at the command line like this: {DIPNER FROG.BFM LPT1 /E /D <enter>}. Or you can use the statement {SET HDDRIVER=DIPNER /E /D <enter>} to modify the behavior of the {DOTS1234} and {DOTS234} batch files.
Raised Dot Computing has developed an optional driver for PostScript printers to make inkprint dots with the same spacing as tactile dots. This driver program was written for the sign making industry, which is now required to put grade two braille on door signs. There is an additional charge for this software module. For information and sample output sheets, contact Raised Dot Computing.
If you have a textfile which is a poem or a verse play (e.g., Shakespeare), importing it into Hot Dots with the {ASCD} file type does not do a good job of formatting the final output. There is a new file type {VERSE} to deal with these formats. For example, you may have a file called {HAMLET.TXT} that contains all of the Shakespeare play Hamlet. To turn this into braille, enter the following: {DOTS1234 HAMLET.TXT VERSE LPT1 <enter>}. You will get appropriate indent and runover by using the file type {VERSE}. This feature was inspired by the work that Raised Dot Computing has done with CD-ROM material. Since it is possible to collect an entire Shakespeare play in one textfile (from the Library of the Future CD-ROM disk), we felt it was important for Hot Dots to do a good job of producing this material in braille.
4DOS and NDOS are non-standard variants of MS-DOS. These operating systems have more commands than standard MS-DOS (and you can combine the commands in more interesting ways). 4DOS and NDOS have their own command called GLOBAL which gets in the way of the Hot Dots GLOBAL command. If you use 4DOS or NDOS, to straighten things out add the following lines to your {AUTOEXEC.BAT} file: {ALIAS DOALL *GLOBAL
This renames the 4DOS or NDOS GLOBAL as DOALL. Then any use of GLOBAL invokes the Hot Dots GLOBAL.
Hot Dots now offers a debugging mode in which you can see some of the internal commands executed by Hot Dots. To turn on this debugging feature, type {SET HDSHOW=1 <enter>}. To shut this feature off, type {SET HDSHOW= <enter>}. Remember, when you use a {SET} command, don't put any spaces next to the equal sign.
When you import a file with the {ASCD} file type, Hot Dots must decide for each carriage return whether it is hard (retained) or soft (deleted). When a line is short, Hot Dots treats the carriage return at the end of the line as a hard return. But how short must a line be to qualify as a short line? Hot Dots uses an internal parameter called {PARACONTROL} as a guide. If a line in the input file has length less than the value of {PARACONTROL}, Hot Dots treats the return at the end of the line as a hard return. Otherwise, Hot Dots assumes the line ends with a soft return. The default for this number is 30. You can change it by typing {SET PARACONTROL=25 <enter>} (or any other number). The lower the number, the fewer hard returns you get. The higher the number, the more hard returns you get. We picked the number 30 since it seems to give a good balance. Feel free to adjust {PARACONTROL} to fine-tune the software for the results you want.
Since February 1991 we have fixed a number of problems in the Hot Dots translators. For example, Hot Dots used to use a letter sign in ordinal numbers like 2nd and 3rd. This problem has been eliminated in Hot Dots 3.01. Your feedback on the translators is still most welcome, as explained in the article "RDC Increases Translation Bounty in 1992" in this issue.
If you run into any problems with Hot Dots, do not hesitate to call our technical line at (608) 257-8833.
In the November/December 1990 issue of the Newsletter, we announced a bounty on words which are mistranslated in Hot Dots. In that article, we announced a bounty of $1 in credit for each error. Unofficially, we have been crediting people $5 per error.
Betsy Doanne reported that deregulate was translated with the (er) contraction. We are giving Betsy $5 in credit. Richard Hartness reported that kilonewton was translated with the (one) contraction, and he also pointed out a number of other translator problems. We are giving Richard $25 in credit. We are also giving $25 in credit to Rick Roderick for pointing out several problems in the translator.
For 1992, we would like to announce a further increase in the bounty to $10 per word (as long as the word is not a proper name). We are also offering a $5 bounty on translation errors involving punctuation, emphasis, or formatting commands. We are offering a $1 bounty on mistranslated proper names. We are also offering a $1 bounty on errors in the back translator. All of these bounties are in credit for future purchases from RDC. Please read the following list of restrictions carefully.
With deep regret, Raised Dot Computing has had to lay off two hard-working full-time employees, Phyllis Herrington and Carolyn Briggs. Phyllis did technical support and recorded this Newsletter, manuals, and other documents. She also worked on other projects. Carolyn did shipping and promotional projects. With her flare for promotion, she became our first "shipping goddess."
We are doing our best to maintain a high level of support for our customers with our reduced staff. We thank you for your understanding as we adjust to working as part of a smaller team.
I hoped that no one would have to be laid off. I overestimated the interest in producing braille that the ADA (Americans with Disabilities Act) would generate. I thought that more groups would be purchasing braille translation software and braille embossers now that they are legally required to prepare documents in braille. Instead, it seems that these groups are waiting to see if anybody gets sued for non-compliance with the law before they invest in braille production systems.
We deeply appreciate the level of service that Phyllis gave to Raised Dot Computing for almost six years. Since she has left, many customers have told us how much they appreciated her help. If you are interested in getting Phyllis's resume, please call Raised Dot Computing at (800) 347-9594. We also deeply appreciate the energy level and creativity that Carolyn brought to RDC, and we regret that her time here was so brief. We wish both Phyllis and Carolyn all the success in the world!
Every year we take time to reflect upon the past year. We think about those things that have been accomplished and grimace at those things left unaccomplished. Regardless of what the year brought, there is a new year ahead ready to unfold itself in a joyful and exciting way. I would like to use the pages of the RDC Newsletter to do a little reflecting.
The past six years of giving technical support at RDC have been some of the happier years of my life. I learned rather quickly how to dress for the Wisconsin winters and even how to deal with blizzards. But the most important things I've learned have come from my coworkers at RDC and from all of you. We all have seen so many new and exciting changes in adaptive hardware and software. We all have figured out the various interfaces and cursed those which have proved problematic. But when the solution was found, we sighed a great sigh of relief and jumped up and down with joy.
I want to publicly thank Caryn and David for being two of the best bosses one could have. They believed in me many times when I didn't have much confidence. They allowed me to grow as an individual and learn more about computers than I ever dreamed of knowing, and they stretched me when I needed it. I can see a few blushes coming from David, so I won't belabor the point. Caryn and David, it's been fun. Thanks for everything.
I've had an opportunity to meet many of you at conferences and via the tech line. I've enjoyed the privilege of trying to solve your technical difficulties and to encourage whenever possible. You are the most gracious people I know. You've been patient while we've figured out the solution to a rather difficult problem. I've also enjoyed getting to see some glimpses of your lives beyond the computer.
What the future holds is not clear at this point. One thing is certain. I may not be a permanent fixture at RDC, but you never know when I might appear in the pages of this Newsletter. I look forward to the coming year and to the opportunities it brings.
Again, thank you RDC staff present and past. I can look upon the past six years with much fondness and gratitude. Just keep plenty of hot water for tea.
I wish all of you a wonderful holiday season and happy new year. SHALOM!
Nemeth Code is the code for mathematics and scientific notation in North America. During my years of studying and teaching mathematics, I developed a great respect for Dr. Abraham Nemeth, who had developed this braille code. He has had a very active academic career in mathematics and computer science, and he remains very active in his retirement. I was delighted that he agreed to talk with us about the background of the Nemeth Code and about himself.
Q. How did you get involved with developing the Nemeth Code?
A. I began working on my braille math code in 1946 or 1947. At that time I had a day job at the American Foundation for the Blind (AFB), and I was taking night classes in math at Brooklyn College. Boys were returning home from World War II and going back to school. Many of them had passed Calculus I before the war and were now taking Calculus II. After that long interruption they needed some extra help. There was a room with a large blackboard where volunteers helped them with their math problems. Each student stationed himself at one panel of the blackboard and wrote out the problem he wanted help with. I was one of the volunteers. I asked the student to read me the problem, and then I worked out the solution on the blackboard. I didn't find it difficult to write on the blackboard. Apparently the Chairman of the Math Department happened to observe me and was impressed. One day I received a telegram from him asking if I could replace a member of the math faculty who was ill. The telegram asked if I could start next Monday. I said yes.
When I wanted to take notes, I needed a way to write things down. At the time people used the Taylor Code from England for writing mathematics in braille. I thought that the Taylor Code used too many grouping symbols. I had already come up with rules to tell my readers how to read mathematics aloud to me. I began working on a braille code which simulated my rules for speech. For example, when you say "x to the n power," the phrase "to the" means "begin a superscript," and the word "power" means "return to the baseline." So in my braille code I created symbols that mean "begin superscript" and "return to the baseline." My personal code for braille mathematics began to evolve. I used it for my work in calculus and statistics.
Q. Before you tell us how you came to share your private braille math code with the rest of us, I'd like to know how you started taking evening math classes at Brooklyn College.
A. I was always interested in math. I went to the New York City public schools, and I spent a lot of after-school time at the Jewish Guild for the Blind. I had a good buddy there who was younger, and I showed him a lot of math. When he got to high school, he took a math placement test. The result was no surprise to me. It showed he knew a lot of algebra and could skip the first algebra class. That was the first case of advanced placement I ever heard of.
Anyway, I always liked math. But various counselors told me that I couldn't have a career in math because I was blind. I heard this from so many counselors that I believed it. After all, there's a saying, "If three people tell you that you're drunk, you'd better lay down." So I majored in psychology. I got a B.A. in psychology from Brooklyn College and an M.A. in psychology from Columbia University. But it wasn't so easy to get a job as a psychologist either. I got a job at the AFB but not as a psychologist. My first wife, Florence, who died in 1970, knew how much I loved math. She asked, "Wouldn't you rather be an unemployed mathematician than an unemployed psychologist?" So I started taking math classes at night at Brooklyn College and then got the teaching position there. I worked toward a Ph.D. in mathematics at Columbia University. I got a mathematics teaching job at the University of Detroit and finished my Ph.D. at Wayne State University in Detroit.
Q. How did your braille math code become an official code?
A. Another blind employee at AFB was Dr. Clifford Witcher, a physicist from Columbia University. One day he asked me if I had a table of integrals in braille. I said that I had one, but it was in my own private braille code. So he asked me to teach him my code. When I showed him the code, he really liked it. Dr. Witcher happened to be a member of the Mathematics Subcommittee of the Joint Uniform Type Committee. This committee, an ancestor of BANA, was responsible for braille codes in the U.S. and England (the word "Joint" for U.S. and England).
[Editor's note: A table of integrals is a long list of formulas for performing a calculus operation called integration. A table of integrals is part of the holy liturgy for calculus students, engineers, physicists, and many others.]
Q. This really brings back memories for me. When I was a freshman in college, my new friend, David Holladay, asked me if I had a table of integrals in braille. I told him that my braille calculus textbook had a table of integrals at the end. He said that he had a much better one in print and wanted to braille it for me over Christmas vacation. He asked me how the braille math code worked and spent half an hour taking one page of notes on Nemeth Code. He took my Perkins Braillewriter home over the vacation and did a really good job of brailling his favorite table of integrals.
Anyway, what happened after Dr. Witcher became a fan of your private braille math code?
A. Dr. Witcher asked me for a document proposing my braille math code to the Joint Uniform Type Committee. Various members of the Mathematics Subcommittee were supposed to write different parts of the official code. But they ended up using my proposal, with minor editorial changes, as the official code book. That was the Nemeth Code, 1952 edition. It was published by the American Printing House for the Blind (APH). In 1956 they reorganized the code book. They made a separate section at the back of the book with all the rules not needed until after tenth grade. They soon realized what a mistake that was. The whole math curriculum changed when the U.S. wanted to catch up with the Russians after the Sputnik launch in 1957. Set theory was moved into the elementary school curriculum, but the set theory symbols were at the back of the Nemeth Code rule book.
Q. If Nemeth Code was proposed to the Joint Uniform Type Committee for the U.S. and England, why don't they use Nemeth Code in England?
A. I don't know. At some point the Braille Authority, with only three members, became responsible for braille codes in just the U.S. Later they added Canada, and it became BANA (the Braille Authority of North America). I think they should add New Zealand and Australia to make it BANANA.
Q. What were the other revisions to the Nemeth Code?
A. APH published two newer editions of the official code book in 1965 and 1972.
Q. Were there any major changes in these revisions of the code book?
A. In 1965 we got rid of a rule about using two spaces to switch between text and mathematics. We also made parentheses more consistent. In 1972 the changes were noticed by transcribers but probably not by anybody else. Since 1972 there have been two addendums, for key caps (symbols for keys on a keyboard) and for ancient numeration systems.
Q. When will the Nemeth Code be finished?
A. When will mathematical notation be finished? The purpose of changes is improvement, not changes. Mathematicians are intrinsically lazy creatures. They spend years trying to find an easier way to do things.
Q. How do braille teachers, blind students, and transcribers learn Nemeth Code?
A. The official rule book is a reference book, not an instructional manual. There have been books written to teach the Nemeth Code. One was Introduction to Braille Mathematics by Bernard Krebs and Helen Roberts. The Library of Congress, National Library Service for the Blind and Physically Handicapped (NLS) adopted it as their instructional manual without sending me a copy. They had to add many correction pages. I was not happy that the name "Nemeth Code" is nowhere to be found in that book. There is another instructional book about Nemeth Code for braille teachers written by Ruth Craig. When I learned that APH was going to publish it without the name "Nemeth Code" in the title, I called my lawyer.
Anyway, Nemeth Code is part of the college curriculum for braille teachers. Blind students learn it from their braille teachers and by reading and using it. NLS certifies Nemeth Code transcribers and offers a correspondence course for those preparing for certification. Braille transcribers also have an active network through the National Braille Association (NBA) and the California Transcribers and Educators of the Visually Handicapped (CTEVH). Both of these groups conduct Nemeth Code workshops, distribute the workshop materials, and publish articles about Nemeth Code. You do not need to understand the mathematics that you are transcribing in Nemeth Code. My personal transcriber, Pearl Hartman, had not gone to college. When she talked about transcribing a topology book for me, her friend asked what topology is. She said, "I don't know. It must be very advanced because the book doesn't have any numbers in it."
Q. What have you been doing more recently?
A. I have developed a new uniform braille code. You can use one code to transcribe literary material, mathematics, and computer notation. With a uniform code, you don't have to use different braille signs in different codes for the same inkprint symbol; the braille sign for a dollar sign is always the same. You don't have to learn more than you need. If you read only literary material, you don't need to learn how mathematics is done. The main difference between literary material in my uniform code and in the current literary braille code is that numbers are done with dropped digits. I serve on the Research and Development Committee of the National Federation of the Blind (NFB), and I wrote a 20-page document in this uniform code to share with the committee. They didn't even realize until page five that they were reading a different braille code. I would like to see a uniform braille code adopted by BANA.
I have written a scientific calculator program for MS-DOS computers which is sold by the NFB. One of my current projects is writing a program in the C language for a new formatter for the Braille 'n Speak. It makes better use of space at the end of the line. If they like it at Blazie Engineering, they may use it in the Braille 'n Speak.
Q. How long have you been involved with computers?
A. For a very long time. Back in 1959 and 1960 I wrote a grade two braille translator in machine language on the IBM 650. The IBM 650 required a special room that was always cool. It took so long to warm up the machine in the morning that we said it had morning sickness. In 1976 my wife, Edna, and I went to the Soviet Union for a computer conference at the invitation of the All Russia Society for the Blind. The topic of the conference was methods of getting braille computer output.
Q. So you have witnessed many changes in the sensory aids field?
A. Yes. The Braille 'n Speak is a much more powerful computer than the IBM 650, it does not require a special room, and you don't have to wait for it to warm up.
[Editor's note: Dr. Nemeth heard this poem on a show called Chip Talk on the CBS Radio Network. We don't know who wrote it. He liked it so much that he called his local radio station, and they gave him a copy on tape. Here it is, with thanks to the unknown author:]
Have you ever wondered how much it costs to make a braille book from start to finish? The answer may surprise you. The first step in modern braille transcription is data entry. If you use a commercial service to do data entry, it costs about one dollar per thousand keystrokes. To accomplish data entry, these commercial services use optical scanning or the labor of overseas typists (usually in the Far East). Since there are about a thousand inkprint keystrokes per braille page, that works out to a dollar per braille page.
That is only the cost of raw data entry. To turn inkprint into rough braille requires inserting markup (formatting commands), running the translator, outputting the braille on an embosser, and bursting and binding the braille. To make sure the braille is correct, you need to have each page proofread. This usually involves having a certified braille proofreader work with a sighted copy holder. Any errors need to be fixed on the computer. Then the file needs to be retranslated and a new copy embossed. I estimate that a reasonable cost for making quality braille is $6 per braille page. This assumes that you are paying a salary to all those involved. This also assumes that the book does not contain material requiring specialized braille codes (i.e., Nemeth Code or braille music).
The economics of braille are distorted by the volunteer labor which is often used to produce it. Friction occurs when those who volunteer have to deal with those who do not (i.e., those on salary). When a volunteer group needs a book already transcribed on disk by a non-volunteer group, they resent being charged for the time and effort of embossing a copy. Both kinds of groups are wary of sharing master disks with each other.
I have no simple solutions. I suspect that if an effort was made by the major braille groups, both volunteer and salaried, to chart the real costs of their braille output, there would be a better appreciation of all the effort that goes into producing quality braille. Such an accounting would be useful to software vendors, since it might show where to concentrate development time to further lower the real costs of producing braille.
I know that these numbers will arouse some blind persons who are advocates of braille. Many people would prefer to lower the standards for braille if it would lower costs and lead to a wider and more timely distribution of braille materials. The problem is in the production of textbooks for school children. No one wants to put second-rate braille books in the hands of a student. For these materials there is no choice but to follow all the rules. If these high costs are frustrating to you, use your energy to find ways of simplifying the braille rules, provided the simplifications do not interfere with readability and ease of use.
For the last year, I have been working with CD-ROM technology and braille. (For those who need background information on CD-ROM systems, see the Jan/Feb 1991 issue of the Newsletter.)
Each CD-ROM has the capacity to hold about 6,000 braille volumes. The manufacturing costs for CD-ROM's are very low, about $1.50 each (once the master disk is made). From these bare facts I make the fantastic leap that CD-ROMs are the ideal carriers for all the braille formatted files that transcribers produce. If braille masters were distributed on CD-ROM, it would mean that anyone with a PC, an embosser, a CD-ROM drive, and lots of paper could emboss any volumes they wanted.
I know how each organization treasures its master disks as if they were crown jewels (as well they should). If we accept a low figure that each page of braille "costs" $4, then a collection of 5,000 braille volumes represents a collective human endeavor worth two million dollars.
My concept is based on the premise that any activity that involves so much patient work is worth preserving in a way that makes it available to many people.
I have no plan or agenda to carry this out. I suspect that some private foundation might be induced to fund a pilot project to distribute those files that people are willing to share.
I also know that there are many different disk formats out there: IBM with its different sizes and densities, Apple with BEX and ED-IT, Commodore with Tabby Cat. I know that transcriber groups are getting smarter about transferring files from one format to another. I recommend low density 3.5 inch PC disks as the ideal way to save braille files. They hold a lot (720k) and rarely fail.
I know that there have been experiments to optically scan paper braille. If this technology could be perfected, then there would be no need to save all the paper masters. Some organizations could save a lot of money in rent if they didn't need all the storage space for their irreplaceable paper masters.
I know the idea of putting thousands of volumes worth of braille in one disk sounds fantastic, but think how far the braille field has come in the last generation. We have gone from slate and stylus to extensive use of quality transcription software in a very short period of time. Who knows where the next generation will take us?
In the July/August 1991 issue of the Newsletter, we described a new state law in Texas regarding braille production. The law requires book publishers to submit disk copies of their textbooks to authorized braille producers as a condition for doing business in the state of Texas.
I have been appointed to a state commission to assist in the implementation of this new law. I just attended the first meeting in Austin, Texas. It was a chance for braille producers, braille and electronic text software developers, typesetters, book publishers, teachers of the blind, and blind consumers to compare notes and learn from each other.
Several important things came out of the meeting. One was the sobering realization that getting the text of books on disk would not appreciably speed up the production of books in braille. Getting the material on disk is a big help. But coping with material that was prepared on disk to meet other needs has its own set of problems. The result of a big step forward and a small step backward is probably a small step forward.
I was shocked to learn that the first batch of books delivered on disk had no markup at all. The files contained just the raw text of the book (with carriage returns showing the start of paragraphs). There were no marks for start or end of emphasis (italics, underlining, or boldface). There were no marks showing the beginning and end of headings. When this information is lost, the person producing braille must type the formatting commands to tell the braille production system about these and other text attributes. It seems like such a waste. First the publishers spend hours stripping all these commands away. Then the braille transcribers have to spend countless hours typing them back in again.
To prevent this from happening, a subcommitee is trying to come up with a recommended list of markup commands for the publishers to use in these files. The members of this subcommittee are Joseph Sullivan of Duxbury Systems, George Kerscher of Recording for the Blind, and myself. We are working on a list of commands which is easy for the publishers to use and which the different braille production software packages can use. We are basing our work on the American Association of Publishers subset of SGML. If you are interested in this aspect of the project, call me at Raised Dot Computing.
Another surprise for me was learning from Dr. Phillip Hatlen (the superintendent of the Texas School for the Blind, who chaired the meeting) about the need for interlined braille. Interlined braille means a line of print followed by the corresponding line of braille. So far, no brailler sold in the US properly interlines braille. The one that comes closest is the Ohtsuki, but it puts the print under the braille instead of above it. When the inkprint is below the braille, the student's hand covers up the print for the braille that he or she is pointing to. Besides, we all agreed that the inkprint that is displayed should be provided by the applications software, not internally generated by the brailler.
I suspect that there would be a strong market for a properly designed device that did both print and braille. I joked that if people felt strongly about this issue, they should announce that as of Jan. 1, 1994, the state of Texas will only buy devices that do proper interlining. Nobody laughed. Instead, I was put on the Interlining Committee. So TeleSensory, Enabling, Blazie, HumanWare, American Thermoform, and Index, pay attention. If your equipment does not do this little trick in two years, you may not have a market.
Recently, we have received a few phone calls from people needing to move data between the Macintosh and the IBM-PC. The Macintosh system software has a feature called Apple File Exchange that can read or write PC 3.5 inch disks. In the July/August 1990 Newsletter, we described how to use Apple File Exchange to read and write Apple II ProDOS disks on a Macintosh. This article explains a similar procedure for reading and writing PC formatted 3.5 inch disks on the Macintosh.
When you use this procedure, you lose virtually all formatting information in the conversion process. If you have a large amount of data conversion to do (such as an entire book), consider purchasing a specialized data conversion program that can deal with the formatting information. The advantage of the Apple File Exchange method is that you do not have to buy anything extra (except a few more blank disks).
This technique works only with 3.5 inch disks. Your PC must have a 3.5 inch disk drive. When you are writing PC disks on the Macintosh, you need several disks already formatted on a PC.
You need a copy of the Apple File Exchange program on your Macintosh hard disk. It is found on one of the Utilities disks that come with your Macintosh System Software. If you have not already done so, copy the Apple File Exchange program onto your Macintosh hard disk.
The PC uses a carriage return-linefeed pair to force a new line. The Macintosh uses just a carriage return to force a new line. When a file goes from the PC to the Mac, you need to eliminate the linefeeds. When a file goes from the Mac to the PC, you have to add linefeeds.
On the Mac, a linefeed shows up as a square at the beginning of the line. A linefeed is really a control-J. Depending on your inclinations and the tools you have available, you can eliminate linefeeds on the PC or on the Mac. On the PC, I use the Hot Dots GLOBAL program to wipe out all the {~0A} characters in a file (a linefeed is the 10th ASCII character, a 10 in hexadecimal is 0A). On the Mac, I use the Change All utility in Word to change linefeeds to nothing at all. Both approaches work.
After converting a file from Mac to PC format, I use the Hot Dots GLOBAL utility to add a linefeed after each carriage return. I change {~0D} into {~0D~0A} (carriage return into carriage return followed by linefeed).
Another carriage return issue is how often you want them. Most textfiles on the Mac have just one carriage return at the beginning of each paragraph. Even if the paragraph has 600 characters, you have to wait until the end before you find another carriage return.
On the PC, most textfiles are designed as print image files. You have a carriage return-linefeed pair at the end of each inkprint line.
Most Macintosh programs allow you to choose to save "text-only" or "text-only with line breaks." Text only means returns are placed only at the beginning of paragraphs. Text only with line breaks means returns are placed at the end of each inkprint line.
On the PC, WordPerfect allows you to save both kinds of files. If you press control-F5 3 in WordPerfect, you get to save a "Generic file" (returns only at paragraphs, the kind of file that the Mac likes). If you press control-F5 1 1, you get to save a "DOS textfile" (returns at the end of each inkprint line, the kind of file that the PC likes).
If someone hands you a Mac disk and you want to use the material on the PC, then you have very little control over the format. Depending on the disk, it may take some work to make the data usable in your PC application. If you have a global replace utility on the PC, you can probably fix some of the data problems quickly. On the other hand, if you are doing data entry yourself on a Macintosh for use on a PC, then you have control over the data and its format.
Here are the steps in the conversion process:
You have two ways of eliminating linefeeds: you can do it on the PC (with Hot Dots) or you can do it in your favorite Macintosh application. Of course, you need to wipe out linefeeds only once.
Here are the steps in the conversion process:
Working out a good system for your application may take some experimentation. The pay-off is in the many worry-free conversions you can perform after establishing your system.
Every few years I get this wild urge to change things around a bit. I had been using two PC's, a Toshiba portable and an ancient creaking XT machine. I was running both with Artic Business Vision, version 2.10 as my screen review program.
I wanted to upgrade my desktop system, and the choice of a new computer system seemed easy. Just go out and get the fastest 386 machine you can afford with an adequately sized hard disk. However, selecting a screen review package for this new system from the many choices available was something else again.
How to decide? For starters, I knew I didn't want to purchase an upgrade of my present software. I wanted something new and different. I had taught students to use Flipper and Vert, and I knew I didn't want either of these packages, although they are fine programs. I had tried a demo copy of Jaws, which works well with macros. I had also used a demo copy of Vocal-Eyes. Although it is very flexible, giving extreme ease of customization, I decided not to purchase this package either.
I longed for the days of my Apple 2E when I could boot up with BEX controlling the whole system and just go. Most of the programs I had tried did not seem to have that lightning fast response I so fondly remembered. I wanted a screen review package that would be as transparent as possible to my applications, yet supremely flexible. I finally figured out what I was looking for. I wanted a program that had a good deal of automation in it that would not require setting up dozens of configurations for my applications programs. ASAP, which stands for Automatic Screen Access Program, had recently introduced a bit of artificial intelligence into the process of telling the blind user what is on the screen. It seemed to be just the thing to fill my needs.
Last fall I contacted a friend in Philadelphia who until recently had been a confirmed Apple user. She had vowed to die rather than switch to the PC environment. I was really surprised when she said, "You have to try Larry Skutchan's ASAP package. It makes a PC as simple to operate and as user-friendly as an Apple system."
To make a long story short, since this friend and her husband sell ASAP and computer systems, they made a sale that evening. I purchased a 33 megahertz 386 system with the ASAP screen review software and the DoubleTalk speech synthesizer by RC Systems. The DoubleTalk and ASAP come bundled as a package for $795.00.
The press release for ASAP states that while there are other programs that provide speech access to applications software, ASAP is the only package available that lets the computer do the work for you. The manual states further that you don't have to be a genius or even a very patient person to start using thousands of programs with speech. Unlike other speech access software, ASAP lets you use the computer like your sighted peers, concentrating on the application. You just start the program of interest without having to configure your speech system. These seemed like incredible claims, too good to be true!
I decided to put this to a test with WordPerfect 5.1. With my previous package, I had set up a window to read the status line, and all of my function keys in WordPerfect were voice labeled; for example, if I pressed F3 for help, the computer would announce "help" and then go to the help facility. Upon trying ASAP with WordPerfect, I found that it reads most of the prompts automatically. During the operation of WordPerfect, it loads various {.set} files which change the speech environment. Therefore, I do not have to set up a separate window to review the choices presented by the spell checker. Also, ASAP lets you voice label keys in an applications program.
This package supports some interesting features. For instance, there are three ways to silence the speech output. The method you use depends upon your needs at a particular time. If there is a lot of DOS output in progress, you can completely turn off the speech by using the Alt key. If you want to silence the speech and make it catch up with real time operation, use the control key. Finally, and unique to ASAP, you can temporarily silence speech, to the end of the line, with a press of either shift key. This last command is very useful when using the DOS directory command. You can invoke the {DIR} command and then keep hitting the shift key to skip down through the lines until you find the file name of interest.
While ASAP operates completely automatically, it also provides you with a way to repeat the relevant or emphasized text on the screen. You refresh your view of the material on the screen by pressing both shift keys simultaneously.
In addition, you can read parts of the screen in real time without having to go into review mode, or the control panel as ASAP calls it. You do this with ASAP by using the numeric keypad on an enhanced keyboard. The numbers 7, 8, and 9 read the previous, current, and next line respectively. The numbers in the next row, 4, 5, and 6, are for reading the previous, current, and next word. The numbers in the third row, 1, 2, and 3, read the previous, current and next letter on the current line. By pressing two twice, you get a phonetic pronunciation of the current character followed by additional information--the attributes of the current character, the cursor position, the status of the locking keys, and a color code used for setting up some ASAP functions remotely.
The commands mentioned above are associated with a reading cursor, independent of the application's cursor. It is therefore possible to examine the screen while your program is doing something else, without having to freeze the program and go into a review mode.
ASAP also lets you make broad movements around the screen by pressing the 0 key on the numeric keypad and then one of the other number keys. You can move the reading cursor to the bottom of the current window by pressing 0 followed by 2 on the keypad; ASAP moves the reading cursor and reads the text on that line. You can also press 0 followed by 8 to type the word located at the position of the reading cursor. For example, you can get a directory in DOS and then use your keypad to move the reading cursor to the filename of a program. Press 0 8, and ASAP will type out the filename for you. Then you just have to press Enter to start the program.
You enter the control panel by pressing Ctrl-\. The original command was Alt Spacebar, but this was changed recently because of a conflict. While you use the control panel, ASAP takes over all keys and uses them for its own purposes. In the control panel, you use the letters A through Y to move to a specific line on the screen and read it. To move up one line and read it, you use the semicolon. To move down one line and read it, you use the slash. If you press the shift key along with semicolon or slash, ASAP reads the previous or next line only from the reading cursor's position to the right.
In addition, you can read text in review by word or by character. The comma moves the reading cursor to the previous word and announces it. The period moves the reading cursor to the next word and announces it. By pressing the shift key along with the comma or period, you move by character. Tab reads the rest of the screen from the line following the reading cursor.
The main use for the control panel is fine-tuning the system to optimize the speech environment for the particular program in use. From the control panel, you can set such parameters as attribute monitoring (Alt-A), reading by screens (Alt-B), announcing keystrokes (alt-K), and most punctuation notification (Alt-M). Perhaps one of the most interesting options you can set with ASAP is "noise with DOS output" (Alt-O). This feature is useful when you use terminal communications software. It makes a click each time normal DOS output prints a character. You can use this to monitor a remote computer's activity. As long as clicking continues, characters are coming in through the modem. It's a really handy feature!
After you have finished setting options, you press escape to exit from the control panel. You can also go into the control panel for just one command by pressing 0 on the numeric keypad. An example of this would be the time announcement. You press 0 and then Alt-t. This one shot function lets you make necessary adjustments to the ASAP environment without completely entering the control panel.
You can also customize ASAP from the command line in your {AUTOEXEC.BAT} file. All of the settings available for adjustment from the control panel can also be set from the command line. You could perhaps start the program with volume 3, speed 7, and no keyboard announcement with the following command: {ASAP 3V9S@KN}
Simply and directly, I really like this screen review program. I have tried it with the terminal programs Procomm version 2.42 and Telix version 3.15, and it has worked flawlessly. It also works well with my favorite word processing programs. In addition, it works very well with my banking program, MoneyMate, which puts a lot of extra junk on the screen. And now, ASAP works well with Hot Dots, once you make a {DX.SET} file. ASAP is transparent to your applications and very responsive and fast.
I appreciate that ASAP comes with a lot of pre-made {.SET} files for setting up the speech environment. I found the manual's discussion on {.SET} files and windows to be fairly complex. I would have preferred a menu system for setting the many options, like the one which Artic Vision uses to prompt you through setting up windows, rows, and columns.
If I have any quibbles at all, they are with the RC Systems DoubleTalk speech synthesizer. Although the speech has a very clear-sounding male baritone voice, it cannot pronounce some consonants correctly. If you come across the filename command dot com, the synthesizer says something that sounds like "command vop com." The other major problem that I have with the DoubleTalk is that even at rate nine it is not nearly fast enough for my taste. I guess that instead of complaining about the DoubleTalk board, I could have bought the Audapter. But for the low price of $270.00, you can't beat the really clear speech.
Even with these minor quibbles I recommend this package most highly. Larry Skutchan, the program's author, has written excellent software, as he did for the Apple environment. He gives excellent technical support for ASAP via modem. Extremely responsive to user input, he welcomes suggestions on how to improve the product. The philosophy of automatic screen reading is really something. Yes, Virginia, it does really work. It is not just marketing hype.
ASAP can be bought bundled with the RC Systems DoubleTalk speech synthesizer for $795.00, or with the RC Systems external speech synthesizer called Lite Talk for $895.00. These products can be purchased from MicroTalk (see Facts on File).
[At this time, RDC is arranging to be a dealer for these items. Please contact Raised Dot Computing for more details. We will also continue to sell Flipper.]
Eric Clegg (amateur radio Call sign AA6XG) lives and works in Sacramento, California as a free-lance talking computer instructor. He can be contacted in any medium at: 686 "N" Street, Sacramento, CA 95814; (916) 446-0301 evenings or weekends.
We intend to make this column a regular feature of the Newsletter. In this column, we will announce new products that we hear about and supply a capsule description, based on our understanding of the manufacturer's representations. If you have one of these items, we invite you to submit a review.
Some of these products are not new to the marketplace. They just need a proper introduction in this Newsletter.
Recording for the Blind has merged with Computerized Books for the Blind. RFB is now distributing books in electronic form. We would appreciate a review from someone who has received several electronic books from RFB.
This is an interpoint brailler that does a double-sided page in 27 seconds and costs $10,500. It boasts a high-resolution graphics mode (132 dots per line) and an expandable 30k buffer. Available from Enabling Technology.
This is a souped up version of the Bookmaker. It does a double-sided page in under 15 seconds and boasts a 512k buffer. Available from Enabling Technology.
This is a souped up version of the Braille 'n Speak with 640k of RAM memory. It costs $1300 new, or $500 if you trade in your old Braille 'n Speak. Available from Blazie Engineering.
This is a small portable notetaker with a refreshable braille display (like a miniature VersaBraille). It comes in two versions: one has 20 braille cells, weighs 2 pounds, and costs $5,000; the other has 40 braille cells, weighs 4 pounds, and costs $8,000. Available from Blazie Engineering.
This is a portable notetaker with voice output and a single-cell braille display. It costs $1600. Available from TeleSensory.
This is a low cost ($270) speech synthesizer for the PC. It is a circuit card that plugs into most PC's (though not the high-end PS/2 machines or some of the Tandy 1000's). The DoubleTalk has very clear speech, and it is supported by a broad range of screen access programs. Available from RC Systems.
Omega Engineering sells the Omegasays, a high-performance multimeter (measures voltage, current, and electrical resistance) and thermometer for $349. Since we know very little about this device, we do not know how suitable it is for use by blind persons.
This is a personal computer system designed expressly to work with the Arkenstone Reading System. The product has a single set of earphones to combine all of the audio input for the user. Available from Arkenstone. Prices start at $2,500.
I am interested in purchasing a used VB-II Plus. Price negotiable. My address is: Debbie Stein, 4666 N. Leclaire St., Chicago, IL 60630; (312) 286-8560.
I have a VersaPoint model C purchased in October, 1988. I paid $3,600 for it. It is in perfect condition (recently checked over by certified technicians). I am open to offers. My address is: S.K. Shin, 305 Memorial Drive, #209A, Cambridge, MA 02139; (617) 253-7729 days; 225-9571 nights.
David Holladay, President; Aaron Leventhal, Software Development; Linda Millard, Bookkeeper; Susan Murray, Office Manager; Caryn Navy, Vice-President.
Written & edited with RDCUs BEX on an Apple IIgs. BEX commands changed to RTF/Interchange format control words with BEX's Contextual Replace. File transfer with BEX and Hayes Smartcom II to an Apple Macintosh Plus. RTF commands interpreted and then spell checked by Microsoft Word 4.0. Pages composed with Aldus PageMaker 3.02, output on an Apple LaserWriter, and printed at the Print Shop. Two track audio edition mastered on APH Recorder and copied on high speed Recordex 3-to-1 duplicators.