Computers

I started working with computers in 1968. The University of Maine in Orono had an IBM 360 computer. It had a climate controlled room all to itself. We never got to see the computer, only those who operated it were allowed. We just went into a separate room with key punch machines and keyed in our program (in FORTRAN) onto punch cards. A hole in the card stood for 0 and no hole was a 1. You then passed our stack of cards to the computer operator. The next day you would come by the computer center to get your cards back along with a printout of your results. Looking at your results, many times you would get an error message. Then you would have to figure out the error and pull the bad cards and insert new ones. If you accidentally drop the cards, you had to go through them and put them in order again. If you had a long program that meant going through 100 cards.

There were no personal computers until later. When I had my own business (Herrick and Assoc) in 1976 to 1979 I purchased the Radio Shack computer (in October 1977) to do surveying calculations.

The Radio Shack TRS-80 Microcomputing System, later known as the TRS-80 Model I, was introduced by Radio Shack on August 3, 1977.

It made computer history as one of the first mass marketed, fully assembled microcomputers. The initial price was $599.95, which included a typewriter-style (not membrane) keyboard, monitor, and cassette recorder. No disk drive, just a tape recorder.
The TRS-80 (which became known as the Model I after the Model II was introduced in 1979) was created by Don French and Steve Leininger. It ran a Zilog Z80 processor at 1.77 MHz and came with 4K of RAM, and a 4K ROM on which was called Level I BASIC. You could get another 8K bringing the price to almost $1000. Only uppercase characters could be displayed, and the only means of storage was a cassette recorder.

“Storage”, I put it in quotes because it really did not work well. I would create a program or type a letter. I then would hook up the tape recorder. I would type in the command to save to the tape. I then would wait the five minutes or so for the tape to turn and the computer to transmit the data to the tape. After the five minutes the computer would flash the message, “Did not copy”. The procedure then was to change the volume setting and try again. Maybe after four or five tries it would work. You just spent a half hour of your day accomplishing nothing. And you had one letter saved.

Expanding the Model I further required a device called an Expansion Interface for about another $1000. This provided:

32K of memory, for a total possible of 48K
a single-density floppy disk controller
a parallel printer port
an RS-232 port

Shown with the expansion interface below the monitor. The first 4K of RAM was in the keyboard, 32K was in the expansion box.

Wow, we got rid of the tape recorder! And could attach a printer and a modem.

SURVEY SOFTWARE
The TRS-80 had BASIC built in, but very little other software.

I had anticipated that software would be available to do the calculations for surveying on the market, but it was not. Finally I found a surveying package, purchased it and loaded it up. I had been surveying a three square mile tract of land and wanted to prove out my calculations on the computer. I spent the time entering into the program every distance and angle turned by my survey crew around the tract. Three miles north, a mile west, three miles south, and a mile east to the point of beginning. I had the latest electronic measuring equipment and a state of the art transit to turn the angles, and my hand calculations showed that the 8 miles of the course had only a tiny margin of error, a matter of about three inches. Well within surveying parameters for those distances and number of angles.
After entering the data into the computer, the result indicated that there was an error of 20 FEET. This was unacceptable. I tried entering the data again and again, with the same error resulting.
I tried entering 90 degree angles and distances for a perfectly rectangular tract. The computer still came up with the rectangle not being square! Who would sell such software? I finally figured out that one problem was that they had used “single precision” when doing the calculations. In other words, a number was treated as 8 digits. No mater what. If I had entered +15851.194523, the computer saw it as 8 digits, 15851.1 (the period and the + were treated as digits). When the numbers were multiplied out, errors would quickly creep in.
I was flabbergasted, I had spent good money on a computer and software that was worthless. My reaction was to build my own survey program in BASIC, but using double precision (16 digits) for every number. Once I did that, the survey came out just as I had by doing it by hand.
I was putting finishing touches on the program in December 1977 when HP came out with the hand held HP-65 with a surveying software pac. The HP-65 was supported by huge program libraries. As of 1977, the application pacs consisted of:

EE Pac
Business Decisions Pac
Clinical Lab and Nuclear Medicine Pac
CE Pac
Navigation Pac
Surveying Pac
Stat Pac
Math Pac
ME Pac
Games Pac
Standard Pac

My survey pac on the computer was just no longer needed, but it helped me out at the time, and I enjoyed the challenge.

OTHER SOFTWARE
VISICALC Dan Bricklin came up with the idea for an “electronic spreadsheet” while still a graduate student at Harvard Business School. He and Bob Frankston founded Software Arts, Inc. in 1979 to explore the idea and VisiCalc was the result.
VisiCalc for the Model I (catalog number 26-1566) was released in late September 1980 and cost $99.95.

Scripsit 1978 Model 1 word processing program. Scripsit was a rudimentary word processor. It had basic text entry and margin controls, as well as word wrap. The original version was tape-based and had no ability to read or write to disk.
Scripsit had a number of significant bugs that could result in the loss of work. If the machine turned off or were reset while a document was still open, the software could not open the document ever again. Scripsit had the counter-intuitive step of “closing the file” which required a special operation before saving and exiting the file. If this step was omitted, the file could not be opened again. No warnings were issued beforehand.

In addition, the disk drive needed to be shut down before powering off. If there was a power outage, even for a split second, the read heads would drop onto the disk. That disk would not work ever again, you had to pay $20 for a new disk.

MODEM

The modem in those days was 100 bits per second, today we normally transmit a million times that (100 MBPS) wirelessly!


The modem 40 years ago only connected to a few colleges and a few US government computers. It was exciting to hear the dial tone on your phone, then place the hand set into the cradle and hear the handshake tones being shared. Then you made your choice between a number of files who had names that were only 8 characters long. Then after an hour or so, there it was, the document you did not want. Not very user friendly, But if you were persistent, you could get something.

TODAY

Skipping ahead, my watch has much more computing power than what we had back then, I get my pulse rate graphed over the past day, the intensity of my exercise, the number of steps I took today, how many calories I expended, how much deep sleep, light sleep and REM sleep, and if I turn it on, it tracks through GPS the route I take.

History of the Radio Shack computer:

John Roach. Born in rural Texas, the Roach family moved to 150 miles east to Fort Worth when John was 4 — and he pretty much never left, not counting a stint in the U.S. Navy. He graduated from Texas Christian University there, majoring in physics and math, and returned two years later for his MBA, which he received in 1965. That’s when he got interested in computers. In 1967, Roach was hired by Tandy Corp. as its data processing manager. “At that time,” he said later, “neither the concept nor the thought of a personal computer had even been conceived.”

Fort Worth-based Tandy started as a leather craft company, but in 1963 Charles Tandy branched out in a wild way: he bought out a small, failing electronics chain called Radio Shack. Roach had ideas about where the new “home computer” market was going: the Apple 1 had hit the market in April 1976. Employee Don French showed Roach the MITS Altair computer kit he had bought. Roach wasn’t impressed, and had a team come up with something better — full-built, not a kit. In January 1977, Roach team took their idea to CEO Charles Tandy and Radio Shack president Lewis Kornfeld: a box of electronics with a full-size embedded keyboard, a TV set as a monitor, and a cassette tape player for program storage. He called it the “TRS-80” (for Tandy Radio Shack, and its Z80 microprocessor). “Charles blew a little smoke,” Roach recalled later, “and said, ‘Build a thousand and if we can’t sell them, we will use them in the store for something’.” Roach countered with 3,500 — one for each Radio Shack store. The computers finally hit the market in September 1977: the company “shipped 5,000 that year,” Roach said. But only that many, because it was “all we could assemble.” They cost $599.95, vs the Apple I at $666. The software for the machine came from a struggling startup: Microsoft.

That was probably part of the success of the TRS-80 Model I: the company manufactured it itself. Once it geared up it could produce 18,000 machines per month. The first model sold 55,000 in its first year, and over 200,000 during the its lifetime. An incompatible Model II soon followed. By 1978 Tandy promoted itself as “The Biggest Name in Little Computers,” and by 1979 needed 1,600 employees in six factories just to build computers. Charles Tandy died in November 1978; Roach was promoted to president in 1981, and CEO in 1983. He retired in 1999. Roach “helped create a market that so many people and companies benefited from as the personal computing industry took shape,” said Bill Gates. “John’s vision and his ability to get early computers, like the TRS-80, into people’s hands through Radio Shack made him one of the true pioneers of this industry.” John Vinson Roach II died in Fort Worth on March 20, 2022 from diabetes. He was 83.