A History in Calculators

Justin Zobel

Page 1 Page 2 Page 3 Page 4 Page 5 Page 6 Page 7

Texas Instruments TI-58

While packing for a house move, I found calculators in drawers, boxes, cupboards. Almost every calculator (or organiser) I'd ever owned. All of these were machines I'd had since they were new, carefully hoarded away over my entire lifetime ... just in case. These calculators reflect the evolution of the technology, from first emergence as a mass-market product to near-terminal decline; the calculator is almost obsolete now, but once was essential. So many calculators. Here they are.

A student at my school, a year above me, bought a serious calculator: the Texas Instruments TI-59, which could load programs from a magnetic card reader. It also had a serious cost, around $400, or about as much as most adults earned in a fortnight. Another student, a friend, bought the cut-down version, the TI-58, identical but half the price and lacking the card reader. It was heavy. It felt well built. It had a lot of functions. The display was bright red, behind a thick piece of plastic. This was the calculator for me.

In the ordinary course of events I could never have afforded the TI-58, but my parents decided to return to England for a couple of months, leaving me in the care of my sister and feeling guilty at not taking me with them. I must have had some skill at exploiting that guilt because a marked-down TI-58 was soon mine, for $115 from the Calculator Supermarket. It came with a plug-in library of 25 suites of software, for applications such as complex arithmetic, each with up to 10 separate functions. In the photographs, the reference card for suite 11 is slotted in below the display.

Though I had used my friend's TI-58, and I'm sure that he evangelised about the fact that it was programmable, I had not appreciated that this was more than a minor feature. Fundamentally — and I have no idea what I thought programming was — it didn't seem important or interesting. However, once I had my own TI, the drama and pleasure of programming dawned on me: the ability to control, to create; the revelation that programming was a kind of superpower that allowed one to do anything. Indeed, the centrality of programming was reflected in how the TI was packaged. It didn't have a user's manual, or a guide. Instead, it had a textbook called “Personal Programming” that explained how to code and introduced programming concepts (stacks, functions, loops, test-and-branch) that were taught through dozens of worked examples.

At that time, 1978, personal computers — very elementary machines that were essentially toys for hobbyists — were just beginning to come into schools and homes. Even five years later, there were few personal computers on campus; computer use was typically via a small number of terminals connected to mainframes on which quotas limited login time to a few hours per week, and in any case most students didn't have accounts. For many people, in this transitional period before PCs became more available, a calculator might be the only practical opportunity to program.

The task of entering programs into the calculator was laborious, requiring the user to carefully write the program down and then methodically enter it. The display only showed digits, a period, and minus signs; there was no indication of last keystroke, and each command had a 2-digit code that had to be memorised (to allow reading and debugging). This same limitation meant that games were rather abstract. The flight simulator cycled through display of the current values of key variables (remaining fuel, current speed, altitude, attitude, wind) while the user entered digits to increase the power or change the flaps. The digit 2 represented speed, so 2.330 meant that the plane was travelling at 330 km/h. If –9 was showing it meant that you had crashed. One of my programs simulated shooting an arrow at a target in the presence of crosswinds, with different types of bow as well as different forces and angles. Numbers in, numbers out.

The crippling weakness of the TI-58 was, once again, power. The rechargeable battery pack was good for 90 minutes or so of continuous use when new, and took a long time to charge; but loss of power meant that all stored information — programs and data — was lost. Occasionally, plugging or unplugging the mains adapter would lead to a momentary power interruption sufficient to wipe everything. With practice even a long program could be entered in 10 or 15 minutes, but this was still tedious and error-prone, so use of the TI at school tended to involve planning an itinerary of power points. (Kindly, the physics teacher would let me leave it charging at his desk.) The first battery lasted just six or eight months, and a new one cost an extortionate $40 or so. This second battery lasted longer, because I was using it less; but when it failed a couple of years later I could no longer find a replacement. I could have upgraded to a newer model with constant memory, but had even less money than when I was in secondary school, and by then was programming on “a real machine”, one of the university's computers.

At university, amongst a much larger group of students who were interested in such things, there were two distinct tribes: TI owners, and Hewlett-Packard owners. The HPs were even more expensive but did seem to be more successful as a practical tool, and were more popular. TI owners rightfully despised the HP people, and vice versa. The poor battery life and lack of access to power points meant that I only brought my TI in once or twice, for specific class activities, but this didn't undermine my loyalty to it.

The TI-58 was well-made; it still works, on mains power. This is despite it being kept on so often for days on end. The Master Library still works too! But in the last 30 years I have only used it very occasionally, for sentimental reasons, or to type in 73802 and tease people.

This is not the story of my own development, but one particular moment has always felt linked to the TI-58: my first encounter with a computer. (The TI was a computer, of course, but was so limited that it was impossible to think of it as the same kind of thing.)

At La Trobe, the university where my stepfather worked and my brother was studying, the experience of programming the TI-58 inspired me to try and see the university's computer, during the annual Open Day. At that time the computer was a DEC10 occupying a dozen cabinets, housed in the lofty isolation of a windowed machine room.

At some point that year, 1979, my final year of school, my school bought a Cromenco computer. This was a desktop machine that ran CPM and had a vast floppy disk drive; I was intrigued by it but never used it. But here, at the Open Day, I found myself sitting at a DECwriter, a kind of terminal that had a keyboard but no screen. The “display” was a line printer where the output was on a continuous sheet of paper, in some ways like a traditional typewriter. A friend, more computer-literate than me (he had used the Cromenco), was there too; we played Adventure and followed some scripts that the demonstrators had prepared.

A journalist from the local paper took our photographs and asked us a few questions. In the next week's issue, a photo of me illustrated the brief item the newspaper ran on the Open Day.

(Source: Diamond Valley News, July or August 1979)

A photo of the first time someone used a keyboard? That photo, used to anchor a story in a newspaper? But there it is, a strange, resonant coincidence. My first computer moment, memorialised; and now I'm a professor of computer science. The TI-58 was influential in my choosing to study computing, and it was influential in that initial encounter, too.

Previous: Page 2. Next: Page 4.

Last updated July 2018.

This webpage was created by Justin Zobel. All images are Copyright© University of Melbourne. Photography by Lee McRae.