NEWSLETTER #19 08/01/84

CHANGING THE PROGRAM IN YOUR VERSABRAILLE UNIT -- Noel Runyan

Many VersaBraille tech heads have wanted to "get into" the control programs inside their VersaBraille units. They would like to change the VersaBraille programs to do something special for them. For example, David Holladay would like to have a special VersaBraille overlay that would permit his BRAILLE-EDIT program to read the VersaBraille table of contents. This would give him much better control of the system and would let BRAILLE-EDIT be even more useful for VersaBraille users.

So why doesn't TSI make these overlays for its customers, or at least let people write their own overlays? Is TSI opposed to letting users write their own VersaBraille control programs? Since I am not working for TSI now, I cannot tell you what their current official policy is. However, here's what we were telling people who wanted to change the VersaBraille programs.

Contrary to popular belief, TSI has not been opposed to letting people develop their own VersaBraille programs. In fact, if you develop a good program they would likely encourage your effort. Furthermore, they might even distribute it and pay you a royalty for it. If your custom VersaBraille program was well written and well documented, it would clearly be in TSI's interest to help support it.

So why haven't David and others written any VersaBraille overlay programs? The truth is that it couldn't be done without some help from TSI. At the very least you would have to convince TSI that they should give you a copy of the VersaBraille program listing. It's possible that they would give you a listing, but it's very unlikely.

Here are the hard facts of life for the VersaBraille.

To say the least, modifying the main VersaBraille program or writing new overlay programs is a tremendously frustrating and time consuming task for TSI or anyone else. Because it would take up a lot of TSI's time, I doubt seriously that anyone could ever convince TSI to help them develop new VersaBraille programs.

Will TSI develop any new overlays or new upgrade versions of the VersaBraille system? It's very unlikely. As a productive product, the VersaBraille system is getting old and obsolete. It is hard to imagine that TSI would be interested in spending any more time and money on the development of more programs for a limited number of future VersaBraille sales.

Maybe the only way to get a new VersaBraille program is for somebody to volunteer to do the work for free on TSI's own TI development system at TSI headquarters, in Mountain View. You might have a chance if you can figure out some way that takes none of TSI's time or money.

Given these gloomy prospects for changing the VersaBraille system's programs, what can be done to expand the capabilities of your VersaBraille unit? I think that the best and most practical way is to hook it up to a personal computer with a good custom program such as BRAILLE-EDIT. The VersaText program on a CP/M computer is also a good example of expanding the VersaBraille system with an external computer. The VersaText program makes heavy use of the VersaBraille remote control commands. There is nothing secret or special about the way that the BRAILLE-EDIT or VersaText programs use remote control commands to make the VersaBraille unit dance for them. Remote control of a VersaBraille is described in its user manual. For additional information you could contact the TSI applications engineer.

If you develop your own custom program for a personal computer, you might want to share it with others. If your system is good enough, TSI might be interested in supporting it as they have done with the BRAILLE-EDIT and VersaText systems.

The International VersaBraille Model P2(D) -- Noel Runyan

What is the VersaBraille Model P2(D), and how is it different from the other models? Is it better than the older models? Would it be appropriate to upgrade an older model to the P2(D) version?

The VersaBraille model P2(D) is just a model P2(D) that has been modified to make it more useful in foreign countries. This international model gives a foreign language user a choice of 5 different computer braille codes. It also includes improved methods for display and entry of control characters and characters with either upper or lower case.

The hardware design of this new model is identical to the design of the standard P2(C). The only changes made are in the VersaBraille firmware or software programs. In fact, upgrading a (C) to a (D) only involves changing the VersaBraille program memory chips and master overlay tape.

The 5 choices for computer braille codes includes the standard American computer braille or MIT braille code. However, it has no Grade I braille code. We had to remove the Grade I braille translators to make room for the added computer braille translators and the other changes.

The method for handling control characters is different. Single-cell symbols are used instead of the old two-cell representation scheme. A keyboard chord command lets the user switch control character display modes. Control characters can be displayed as a braille 'dash' or as the standard symbolic character ("m" for carriage return, for example).

To make it easier to distinguish between upper and lower case characters we included a user-selectable option to cause uppercase characters to vibrate. When this mode is switched on with a keyboard chord command, the uppercase characters pulsate at a rate of about two times per second.

The p2(d) unit now makes it easier to switch to uppercase when entering text. In addition to the commands for selecting uppercase or lower case, there is now a chord-"s" or shift-next-character command. This command causes the character following it to be entered as uppercase (regardless of the current case lock state).

There are also some subtle differences in the international unit. For one thing, its interface data transfer rate is slightly slower than the previous models. In some applications, this can cause problems. Duplicating tapes between two VersaBraille units will be more difficult because of the slower speed of the P2(D). Duplication and some other interface applications that did not require handshake protocols on previous VersaBraille units will need it for reliable data transfers with the (D) model.

Another important consideration is reliability. The model P2(C) has been field tested for quite a while, and it has proved to be reliable. Its program is now well debugged. On the other hand, the international unit has not been as thoroughly debugged and tested. Given time, TSI might be able to get it working as well as the P2(c). Initial tests indicate that the international unit will work well with BRAILLE-EDIT.

For users in the U.S., the lack of the Grade I braille code is probably the biggest drawback of the international unit. Many users have found the Grade I braille back translator very helpful in the earlier VersaBraille models. For example, in writing the VersaBraille interface manual I did my whole manuscript in Grade I. One of the main arguments for not having a Grade I braille code in the new machine is that the new methods for handling the display and entry of upper and lower case characters reduce the advantage that Grade I had over the old computer braille system. Hopefully this is true. The international unit will certainly be nicer for computer programmers who have to work with the full ASCII character set and who are finding more situations today where handling of upper and lower case is important.

If you own an earlier model VersaBraille unit, having TSI upgrade it to a P2(D) will not be cheap. As of May of this year, TSI is quoting a cost of about $500 for upgrading a model P2(C) to a P2(D) and quoting $1,300 to upgrade older versions. These high prices may encourage users to do their own upgrades, especially when they consider the ease of copying PROM chips and master overlay tapes.

So which model is better? David Holladay is noticeably more enthusiastic than I am about the international VersaBraille unit. It seems slightly ironic that I should be less enthusiastic, since I spent so much time and energy on the design, simulation, and testing of the P2(D). David has a good point concerning users with easy access to an Apple and BRAILLE-EDIT. Using BRAILLE-EDIT and its Grade II translators removes most users needs for the VersaBraille Grade I translators.

With or without BRAILLE-EDIT the P2(D) might become the most popular version, even for U.S. users. For the time being, TSI is going to continue to offer both the P2(C) and P2(D) for sale. I have not heard anything about stopping the P2(C). They may be waiting to see how the market reacts to the international unit.

If you are considering upgrading or buying an international unit, I would recommend that you also wait to see what problems and popularity is in store for this new international device. If you want to drive a printer directly from your VersaBraille unit without using BRAILLE-EDIT, and you do not like computer braille, the standard P2(C) would probably be best for you. If you would use BRAILLE-EDIT and/or you plan to use computers heavily, you should probably wait until we find out whether the letter "D" in P2(D) stands for "Dog" or for "deluxe".

Toward Guidelines for Braille Convention Programs

Last spring I foolishly volunteered to try to find a way to transcribe the program into braille by technological means for the Annual Convention of the Council for Exceptional Children, which was to be held here in Washington, DC in April. What I hoped to do was get the program from a word processing system at CEC headquarters in the Washington suburbs, edit and translate it using BRAILLE-EDIT, and send it to David for hard-copy brailling. I really wanted to prove it could be done, and perhaps develop strategies for assisting other organizations in providing braille programs.

The first news was good. The program was on a Wang word processor. Also there was a very cooperative word processing supervisor. But that was about the end of the good news. The first bad news was the length of the program. It eventually turned out to take 790 braille pages. Then, the modem connection did not work for some reason so I had to spend about ten hours at CEC with my VersaBraille downloading the program.

During the downloading process, I discovered that there was no way I could really do a braille program like the print program, because extensive hand lay-out had been done after the word processing was finished. I also learned that the text was full of type setting commands, of which there were so many that I could not easily write a transformation chapter to get rid of them. (As it turned out, it took so long to do the hand-editing I had to do, that I could well have spent more time trying to think up transformations. Well, we live and learn!) There were also format problems which one has in braille translation, such as centered titles being too long to center in braille.

Well, suffice it to say that I got it done, finally! Not perfectly, but done! The next problem was how to distribute such a program. Should we provide a pushcart with each copy? Also, we had no idea how many blind people would want braille programs, given that they are not usually available. I decided to have David print fifteen or so copies of the overview section of the program, a good-sized volume in itself. Then we printed two copies of the rest, about eight more volumes. We decided to give anyone who wanted one a copy of the overview, and then let them borrow other sections from the disabled service desk in the registration area as needed.

I would really like to stimulate a dialogue among readers regarding what I should have done, how much braille material blind conference attendees should expect, ("the way, I have a very strong prejudice against being told that I don't need the whole program), and generally, how such situations should be handled. I feel that in the past braille programs have been an afterthought and a step-child. I want organizations, especially those in the "handicapped business", to understand that brailling conference programs is just as much a staff routine responsibility as the production of ink-print programs.

RECREATIONAL COMPUTING

In Search of Integer -- Paul Edwards

I had intended to spend this column talking about some of the neat and wonderful games that are available for virtually no money from sources of public domain software. However, in my travels lately, I have discovered that there seems to be a new disease abroad in the land which is called INTEGER PHOBIA. While this malady is not life-threatening, it causes many people to be unable to access many of the early programs written for the Apple and, hence, must be seen as intellectually debilitating. For those who have always wanted to get Integer programs to talk and have never managed to figure out just how they can do that and for those who want some ideas about writing HELLO programs, this month's article is for you.

First, what is Integer anyway? Well, to put it simply, Integer is a programming language that was developed for use on the Apple II computer. It has largely been superceded by its cousin, Applesoft. All Apple computers that are sold now come with Applesoft BASIC resident in the machine and, in order to run programs that are written in Integer, you need to load this language into memory at the appropriate spot. Apple recognizes that it has a responsibility to its owners and so it provides a version of the language that loads automatically when you boot your SYSTEM MASTER or BASIC disk. For most users of the Apple, then, accessing Integer is not a problem. They can flit from one language to the other by simply typing a command before running a program. Those of us who want to use a speech synthesizer are the ones who get into trouble. This is because the Integer and the Textalker store at the same memory position. When you load the Textalker, the Integer gets zonked out of memory. So what can we do about it? The solution that I am going to suggest will seem a little cumbersome, but it works. I am suggesting this approach because it is a good all-purpose solution to a whole range of problems and, though there will be those among you who will argue that it is wasteful of both time and energy, I suggest that it is the simplest long-term solution to the problem of making Integer programs easily and quickly available.

First, what will you need? You will need a twenty-two sector Integer program that is sometimes called INT and sometimes INTBASIC and sometimes INTEGER BASIC. The main thing is the size of the program. If you can find a twenty-two sector program that says something about Integer, you've got it made. If you don't have one, check with a local users' group or with your dealer. Next you will need two programs from your textalker disk or from your BRAILLE-EDIT disk. These are TEXTALKER.RAM and TEXTALKER.RAM.OBJ. Next you will need a FID program. If your system master has gone the way of all masters, there is one on the BRAILLE-EDIT disk. Last, you will need a HELLO program. In an effort to prove that the last shall be first, we will deal with the HELLO program first.

When you initialize a disk, outside of BRAILLE-EDIT, you have the option of including as part of your initializing process, a little program that gives the computer some instructions to carry out every time that disk is booted. This has come to be known as a HELLO program. The one that we are about to write will do several things. First it will bring up speech. Next, it will set the speed of the speech. Next it will set the pitch. Then it will catalog the disk onto which you put it and, finally, it will run your small Integer program so that you can immediately run any Integer program that is on the disk. Obviously accurate typing is essential for this little exercise so you might want to turn on all punctuation on the speech synthesizer with a control e followed by an a. The Integer program that I use is called INT. Where I have INT, you should substitute the file name that corresponds to the Integer program you are using. Incidentally, I have written out the word PRINT each time it is required in the program but you can substitute a question mark when you are typing your program. Just put the line number, a space and then a question mark and then another space and the Apple will translate that code to mean PRINT and, when you have saved the program and reload it, your question marks will have been replaced by the word PRINT. Well, folks, here goes. Below is listed the HELLO program. It should be typed exactly as it appears here except for those elements noted above and for the fact that you can set the pitch at whatever level you prefer.

From now on, I will give you instructions one step at a time. If you follow them, all should go well. 1. Load speech and get into Basic. 2. Type NEW to get rid of any programs that may be lurking in the memory. 3. Place an unused disk into the drive that you are using. 4. Type the HELLO program listed above into the computer. Remember to press RETURN after each line. 5. Type INIT HELLO and wait while the disk drive formats the disk and puts the program you have just typed onto the disk. 6. Using the FID program, move the TEXTALKER.RAM, the TEXTALKER.RAM.OBJ and the INT or equivalent program onto your newly-created disk from wherever they are. This will require the use of seventy-four sectors of your disk so you will have less space to work with. 7. Using the copy program on BRAILLE-EDIT or from somewhere else, make several copies of this disk. 8. Now you are ready to put whatever programs that you want to onto the disks that you have prepared for them. Once again, you should load your FID program. Use option 1 COPY FILES. Fill in the disk drive information as prompted and, when asked for a file name, hit the equals sign (=) and then RETURN. You will be asked if you want prompting and you should probably answer Y followed by RETURN. The rest is a piece of cake. It will tell you what file it proposes to put onto your disk and you have the option of saying Y or N. If you say Y, it is transferred. If you say N, the computer will say CANCELLED and go on to the next file.

This process will take some time to set up. However, it will save you lots of time in the long run. Once you have gone through the process of typing the HELLO program once, you will never have to type it again. By using the copy program on your HELLO disk, you can save yourself having to move those four files more than once and can initialize and copy at once. There are easier ways of doing what I have described but they will not always work. Integer disks that have Integer HELLO programs are one of the major problems that this system circumvents. It is not easy to change an Integer HELLO program into an Applesoft one. This way, you can use a standard HELLO program for everything. Good luck with this system. If there are questions, send them into the newsletter and either I or David will try to answer them.

A PROMISE ALMOST KEPT -- Paul Edwards

Approximately a year ago, now, the first self-contained talking data base came onto the market. This program was called INFO, was developed by William Grimm and marketed through his company, Computer Aids Inc. for approximately two hundred dollars. I was able to borrow a program from a friend and work with it for a period of time and have read several different versions of the manual as they have appeared. What follows are impressions and, while they are obviously colored by the fact that I have not yet been able to purchase the program, I think that the conclusions that are reached have validity. Let me begin with the good points. Info certainly constitutes a significant step forward for talking programs. It offers opportunities to do things that have, prior to its appearance, simply not been possible for the blind. Of particular value is the ability to use DOCUMENTS, another Grimm program, in conjunction with INFO to create forms, lists and directories. However this article and all reviews that I write must assume that the product under consideration passes more than this kind of a test. All products that are developed for the blind must compete not among themselves, ideally, but against those products that are available to the sighted population. The blind are consumers. As consumers, they have the right to expect that the programs that are being marketed for them contain the same flexibility, range and innovation that competitive software offers to their non-blind counterparts. Judged against such standards, INFO does not fare so well. More significantly, it fails a second test. Every program that requires the kind of expenditure that INFO does ought, in my opinion, to provide effective support services. I have had the opportunity to speak to several owners of Computer Aids programs and the story is essentially the same. INFO, DOCUMENTS and TRANSCEND owners all complain that there are glitches in the programs that remain uncorrected for long periods of time. When glitches are fixed, owners are often required to pay for updated programs; and, just as criminal, programs are rushed onto the market before adequate of effective documentation is available. Incidentally, I looked at the latest INFO manual and found that it was rife with spelling errors and had at least one computational error in it as well. I would like to commend Mr. Grimm and his associates at Computer Aids for designing an extremely user-friendly program. It is simple and straightforward to use and the command structure is simple to learn and easy to remember. Its limitations lie in areas that I understand have been corrected in more recent updates of the program. I would be interested in seeing someone from Computer Aids comment on the deficiencies that I point out. In the first place, there is no way of deleting a file record once it has been entered. The only way that a data file can be purged is to write over its contents. Second, twenty fields is simply not enough. Third, the COMPUTE function is both limited and very slow. This is partially due to the use of sequential text files. This decision also makes it impossible to do alphabetic sorts with the data base. What this does is to make it less useful than it might have been as a generator of printed materials. A simple computation function can take as much as three to five minutes which is unacceptably slow in this day and age.

Competitors such as D B MASTER offer features that INFO does not include. Among these are a more effective and meaningful definition of field types that allows the performance of a wide range of computational functions. Confidentiality, of course, does not exist with INFO. There is no system of passwords which are commonplaces on commercial products now. D B MASTER allows for the use of the records in a data file on more than one disk and makes their transference easy and their reorganization and reformatting simple.

Another deficit of INFO lies in the fact that it places a finite limit on the size of the data base. This is true because it makes no provision for multi-disk files and, if some of the stories that are current are to be believed, it will destroy all the data if the disk is over-filled. I cannot confirm or refute this last statement but have heard it from two users. I tried to create a fairly complex template using INFO and DOCUMENTS and had trouble if I tried to put more than two data base elements on the same line. I am prepared to accept that the errors were mine but the documentation should have explored how such problems as data elements that are longer than the space allotted for them, will be handled. Another disadvantage of the system is that, if you buy the II+ version, you must upgrade it when and if you have access to a IIE. I produced work on a IIE which was not legible on the screen using a II+. This is a real drawback as far as I am concerned.

After all of this, how would I advise the blind person in search of a data base to operate? First I would advise him to take a hard look at the talker/lister program which is less expensive and appears to have more flexibility than does INFO. Since I have not seen the former program, I am not in a position to make effective comparison. Again, if you are looking for a relatively easy program which can perform a wide range of functions, INFO will provide good service for you. I suppose that this review has focussed on many of the bad points of this program but that is as it should be. As I implied at the outset, we cannot afford to begin evaluating the products that are available to the blind separately from those that are accessible to the sighted. The former must accept the responsibility of competing with the latter. We, as consumers, must demand from software designers the same excellence as is required of software writers for the community, in general. We must accept the responsibility of exposing shoddiness where it occurs and of heralding excellence where it is found.