Editor's note: This is a surprise out of the archives … meaning a box in the garage tipped over and these typewritten pages spilled out. Jon wrote this story in 1982 and we thought it was a fun read for computer veterans. It's easy to forget that there was a time when word processors were very expensive and specialized pieces of equipment requiring trained operators … okay, not much training but still, word processing operator was a separate job description from typist. And, a word processor is a long way away from a TRS 80.
As a long-term advocate of automation and labor-saving devices, it suddenly occurred to me that I should practice what I preach. Having designed, sold, and written about complicated computer-based data acquisition systems for the past 12 years, I felt perfectly comfortable purchasing the components necessary to build a word processing system. I had the perfect justification: a new project coming up would require lots of writing and several iterations. It was very logical to have a word processor so that the reiterations could be done conveniently and quickly. Money was somewhat of a limiting factor. However, I felt that my cleverness with equipment should allow me to construct a word processor at the least possible cost. After all, I knew what those parts really cost, and I wasn't going to be fooled by some fancy advertising from some large computer company.
At the time, I made this momentous decision I had in my possession, a 16K Level 2 TRS 80. I had longed to add a printer to this because I was tired of hand-copying the lists of programs that I constructed. I knew, however, that a Dot-Matrix printer would be totally unacceptable for the type of report generation I had in mind, so I set about searching the various journals for an inexpensive high-quality printer.
It didn't take long for me to identify several sources of typewriter quality printers. The only question now was to find the least expensive one. While thumbing through an issue of a TRS 80 journal, I discovered an ad by Microtronics for a TRS 80 based word processor system. The ad featured an AndersonJacobsen terminal (at a very reasonable price for such a device).
I decided to begin my investment by first making a long-distance phone call (from California to Philadelphia). I called Microtronics to get a little more information on their system and to find out how quickly they could deliver one. I had already accepted the project to generate a report and the clock was running. Therefore, I wanted to get a printer just as quickly as I could.
The people at Microtronics were very friendly and very sympathetic to my problem. They explained that the AndersonJacobsen terminal would take ninety days to deliver. And that, no matter how they tried, there was no way that they could get one to me sooner. They did suggest, however, that I consider a TrendData 1000. The TrendData, they said, was everything that the AndersonJacobsen was and more. But the people at Microtronics, being sympathetic to my plight, gave me a discount on the list price of the TrendData.
I called Microtronics a few more times to get some additional technical information and discovered that, to make the printer work with my TRS 80, I would have to have an RS232 interface and a word processing software package. None of this information was startling to me. I knew that some type of interface would be needed and obviously a software package.
Another phone call to Microtronics revealed that the TrendData 1000 possessed an additional feature that I wanted almost above anything else. Namely, variable spacing. After another telephone call to confirm that the printer would do variable spacing and that it would be shipped on time, I placed the order for the TrendData 1000, RS232 interface, and an “Electric Pencil” word processor software package. Microtronics said everything was fine and they would ship the equipment.
When the day came that the equipment should arrive, I called Microtronics to find out where it was. At that time, they informed me that since they didn't know who I really was they weren't going to ship the equipment on credit and that if I wanted it to be shipped, I would have to send them the money. This was frustrating, but manageable. I quickly ran down to my local bank and had the money wired transferred to them. I then called a few days later to make sure that they had received the money and that the equipment had been shipped. This time I learned that the equipment had been shipped and they had received the money, but they had made a slight mistake in their calculations and the discount that I thought I was going to get vanished. Therefore, I would be obliged to pay a hundred dollars C.O.D. when the equipment arrived. I didn't feel too bad about that. A hundred dollars didn't seem like a large of an amount of money and Microtronics had been very cooperative in helping me obtain the equipment in as timely a manner as possible. I quickly agreed to the terms.
Finally, the equipment arrived. However, it didn't quite arrive in the same condition that it was shipped. I had requested from Microtronics that they air freight the equipment because I was in such a hurry, and in their usual obliging manner they did exactly that. However, the air freight company misplaced a handle or something and dropped it or kicked it or whatever.
I had had experience in dealing with freight forwarders before, so I quickly called the carrier, advised them of the damage and told them I would like to have an adjustment made. The air carrier was extremely cooperative, very understanding and prompt in their response to my request. After the air carrier sent out an insurance inspector, the ball was back in my court to find someone to fix the machine.
Looking through the information provided with the terminal, I found a list of TrendData repair centers. And, joy of joy, there was even one in my local area. I called the TrendData repairmen and explained my story and asked them what I should do. He told me that I should bring the machine to him. “To him” was approximately 70 miles away in a nearby town called San Jose. I explained that the machine was larger than I could handle (almost larger than my car) and certainly not possible for me to haul it down to them. The TrendData repair center proved to also be very cooperative. They volunteered to send a truck with two experienced pickup men who would handle the machine quickly and correctly. I jumped at the opportunity to have them do that and encouraged them to come and get it immediately.
While speaking to the TrendData repairmen, I asked them how one went about getting variable spacing on the machine. I then got a very quick and painful education in semantics and definitions. It seems that there is no such term as “variable spacing” in typewriter terminology. You can have multiple spacing which consists of one-, two- or three-line feeds between lines or you can have variable pitch. Variable pitch was what I wanted. It was then that the TrendData repairmen explained to me that variable pitch was not possible with the TrendData 1000.
A TrendData 1000 is basically a modified IBM Selector 730 typewriter. IBM had never intended that typewriter to be a variable pitch device. They were saving that trick for their Mag Card system.
Simultaneous with my education in typewriter terminology, a friend of mine, who is in the computer peripheral business, advised me of the availability of some used Qume terminals. The Qume units, he said, did have variable pitch and were only a few hundred dollars more than I had paid for the TrendData.
I quickly called Microtronics and asked them if they would consider taking back the TrendData and return my money. I explained that it wasn't what I wanted, and I felt that they had some obligation to take it back since they had told me that it would do what I wanted it to do. They denied ever having told me that it would do variable pitch (and I never did use that term, unfortunately, but I did explain what I wanted it to do). And, in any case, it was against their policy to take back equipment for any reason. I was really getting frantic at this point, and I pleaded with the people at Microtronics because I absolutely had to have variable pitch printing. Microtronics tried to help me by offering to take back the TrendData in trade for a Qume that they had; that was the good news. The bad news was that their Qume was twice the price of the one that I could get through my friend's office and clearly out of my budget. Further pleas with Microtronics to return the equipment fell on deaf ears and I abandoned that approach.
I decided that since compromise was the backbone of our political system, it would serve me well also. I, therefore, set out to live with the non-variable pitch TrendData terminal and start my project.
While the typewriter was being repaired, I started to investigate the RS232 interface that came with the TrendData from Microtronics. It was then that I learned that the RS232 interface that Microtronics had sold me required an expansion interface from Radio Shack. I had longed for an expansion interface anyway. I had dreams of one day obtaining a disc drive and I felt that the real timeclock in the expansion interface would be beneficial to me in some other plans I had for a computer. However, I wasn't prepared to spend that extra three hundred dollars necessary to get an expansion interface currently. The cost of my word processing system was increasing on a very rapid and unpredicted rate. Nevertheless, if I was to complete this project, an expansion interface would have to be purchased. So, I went and bought one.
The day I got my expansion interface, the TrendData terminal was returned. The TrendData people did just about everything they said that they would do and on time. I decided the time had come to interface the TrendData to the RS232 interface, which was now in my brand new and shiny interface box. It was then that I discovered that the RS232 cable on the TrendData also had a male connector. This was not a difficult problem for me because of my experience with computers and RS232 in general. I went out and purchased a couple of female connectors as well as a few feet of wire to make a cable for the TRS 80 to talk to the TrendData.
Then I built the RS232 cable. As I was doing it though, it occurred to me that if someone without my experience had sought to put together a system like this, they would have been totally frustrated at this point. Not that building a cable is any big deal but if you have never done it before, it can be a little confusing because you must reverse some pins and not others. I finished my cable and set out to run a test between the TRS 80 and the TrendData.
I loaded the “terminal” program supplied with the RS232 interface into my cassette. The TRS 80 cassette has been well documented in various journals for its idiosyncrasies and bad attitude…I discovered that to get the tape to work, I had to load it three times. I never found out what the magic of three was about that tape, but it did seem to be an irrevocable law. In any case, having established a loading procedure, I was now prepared to start transmitting data from the TRS 80 to the TrendData.
It was at this point that I connected everything together and tried to send data from the keyboard of the TRS 80 to the TrendData 1000. What a surprise that was. I didn't know the TrendData had some of the characters in it that I saw. I then tried transmitting data from the keyboard of the TrendData to the screen of the TRS 80. Another great surprise. By this time, I was used to things not working or being as expected. I called the TrendData repair office and asked them if there was anything that they could tell me to do that would not require sending a serviceman out. The service manager told me that the system had been thoroughly checked and there should be nothing wrong with it and since it was still in warranty from the repair, he would dispatch someone out to see me right away. I was happy to hear that because there was one little problem with the carriage return giving me an extra line feed occasionally.
In desperation, while waiting for the repairman to show up, I removed the side cover and examined the circuitry associated with the TrendData. And quite a bit of circuitry there was. I discovered a package of slide switches on the Main PC board. Then I remembered a description about some switches in one of the sheets of paper that accompanied the TrendData. In comparing the switch positions that were on the printed paper with the switch position that were in place when I first examined it, I found no correlation. I also couldn't really understand what TrendData's coding was concerning the switches. So, I resorted to the trial-and-error method. Try this switch and try that switch. Try the other switch, etc.
My procedure was to try a switch position, go to the keyboard, type a few things, try a switch position, and on and on. By doing this, I ultimately came up with a combination that allowed me to actually transmit data. At this point, I was almost ecstatic. Now I was able to transmit data from the keyboard of the TRS 80 to the printer of the TrendData and from the keyboard of the TrendData to the CRT of the TRS 80. I knew that my word processor was just a step away.
Anxiously, I loaded the “Electric Pencil” program into my cassette. It didn't work the first time or the second time, but after giving it enough tries and adjusting the volume controls and cleaning the heads again, and saying a few incantations over the machine, I did get the program to load. I was ecstatic when the menu for the Electric Pencil came upon the screen. Quickly, I typed in some garbage on the screen, entered the necessary constants and joyfully, pressed Shift P (which is the instruction that transmits the page of data to the output printer). Nothing happened. I repeated the process three or four times. I tried different combinations. Still, nothing happened. So near yet so far away. What utter frustration. I knew that my computer would talk to my printer. I knew that I had the program loaded; I knew that I was following the instructions to the letter. Why wouldn't this blasted thing transmit the data that was on the screen?
After rereading the manual for the sixth time, I noticed in the back there was a reference made to a fellow in Southern California that had something to do with the Electric Pencil (not the author Michael O'Shay). Assuming that the author would be totally unavailable to me, I took a chance and called the fellow in Southern California. I explained to him what I had done and what I was doing and asked him if he could help me. He was very kind, very cooperative, and very helpful. He explained that the Electric Pencil was designed to work with either Radio Shack's printer or his TS232 interface, but not Radio Shack's RS232 interface. I was stunned. Now I was really at the end. I couldn't imagine what else I could do.
During our conversation, he had referred to the fact that one I/O port was used for the printer and a different one was used for the TS232 interface, and still a different one for the RS232 interface. After I had hung up and thought more about it, I decided to call him back. I asked him if it would be possible, but it would not be a trivial task because the program was written totally in machine language. Without some type of a developer system or an assembler package, it would be almost impossible to find the locations in the program where the output board was fouled up. I asked him if he would be willing to do it and he said that he would not at this time since he had too many other projects going. Then he told me that Michael O'Shay had called him and asked him to do the same thing and he might undertake such a project in the future, but nothing would happen in the next three to six months.
So, there I sat with an expansion interface, an expensive typewriter, and an RS232 board, and a computer, and all I could do was type on one keyboard and have it print on the other. I then tried calling one of the TRS 80 Journals that advertises a word processing program with a subscription. I explained to the gentlemen there that I was in an extreme hurry and that my project was now long overdue and that my configuration was an RS232 interface to a printer. He assured me that his program would work so I quickly took a subscription. The only way that he would send the program to me right away, however, was for me to give him my credit card number, which I did. I felt that although this was not my original approach, I did at last to have a solution to the problem. A week and a half later, I received a note in the mail from the Journal indicating that some mistake had been encountered with my credit card number and he was not able to send the tape. At this point, I threw in the towel.
The next day, I went to a typing service, submitted all my papers, and had them manually typed. Then I took them to my client, edited them, took them back to the typing service, and had them manually typed again. Then I took them to my client again, edited them, and took them back to the typing service, and so on. God, if ever there was an application for a word processor, this was it. Its totality came to over 100 pages, and they all had to be numbered.
Next, I decided to get out of the Radio Shack computer business. A friend of mine had a friend who wanted a small computer. A brief negotiation and I sold all the Radio Shack equipment that I had and gave away all the various game programs and things that I had accumulated over the years. At this point, I simply wanted out.
I decided to keep the typewriter because if nothing else it was a good typewriter (even though the repairman never came). After the computer was gone and after the report was finished, I had a lot of time to sit back and contemplate what occurred.
I realized in retrospect that I had violated some of the basic principles that I preach to others. First, I tried to put together an unfamiliar system composed of alien parts in an unreasonable time frame. Instead of taking a few weeks as I had originally anticipated, it took me several months and it couldn't be done. Secondly, I dealt with unknown vendors. Normally in a procurement cycle, you make certain checks on vendors. You find out what their terms and conditions are…what their delivery schedules are and ask to speak to a few of their customers.
Thirdly, although you should always look for the best deal that you can get, very often the person who is the cheapest is the cheapest for a good reason. Fourth, I didn't bother to get written guarantees from the various vendors of the compatibility and prerequisites of the various equipment. Fifth, I didn't bother to take into consideration the lack of a software guarantee. It is now my painful knowledge that no personal computer software writer will guarantee his work. I find this to be not only confusing but also very curious. I suppose the explanation is because they can't guarantee the condition of the computer, or they are not familiar with the computer or one of the other variables that they will use as excuses. Buyer beware. So, in retrospect, I probably broke every rule in the book trying to put together a word processing system. The main reason, of course, being my requirement to do it right away, lack of preparation time, and lack of investigation time.
I am, however, by no means, turned off on small computers or word processing machines. I recently witnessed an extraordinary demonstration of a word processor at Commodore where the Pet computer is manufactured. I decided at that time that I would try a Commodore system. This is my current adventure. I now have a 32K Pet computer with a dual floppy disc. I am trying to find an RS232 interface for the Pet computer to use with my TrendData 1000. Next, I will try to find a word processing software package. I understand Commodore has one. Perhaps, another article is in the offering.