Enter the KIM-1

I continued polishing and tuning the game on the 1130, but I was already turning my attention to a strange new technological development: microcomputers. It's difficult for people these days to appreciate just how exotic microcomputers were back then. Only a tiny number of people, perhaps a hundred thousand in all, were even aware of the technology. But those of us who caught the bug were passionate in our optimism for this technology. I realized that I'd have to make the jump, so I set to work teaching myself digital electronics and microcomputers.

After a few months, I bought myself a KIM-1 single board computer for $249. That was a lot more money than it seems; adjusted for inflation, it comes to $750 in today's money. This baby was packed with features: an 8-bit, 1MHz 6502 CPU with all of 1K of RAM. It had two 6520 Peripheral Interface Adapters, a machine-language monitor, a 21-key keypad, and a 6-digit, 7-segment LED display. It took me a few weeks to learn 6502 machine language; pretty soon I was ready to transfer Tanktics from FORTRAN on the IBM 1130 to machine language on the KIM-1. Unfortunately, 1K of RAM wasn't enough for the program, the tanks, and the map, so I had to trim it down. I ended up with a tiny map, just 16 hexes high by 16 hexes wide, with a single type of terrain: forest. Each side had one tank. And this wasn't human versus computer: It was human against human! I carefully divided the keypad so that one player used one side of the keypad and the other player used the other side. Each player read three of the six LED digits. This was enough, however, to communicate the game's events.

The best part of the game was that it provided blind play: Neither player knew what the other was up to. This, however, did create a problem. It's pretty hard not to notice what the other player is up to when you're sharing a calculator-sized keypad and display. My solution was, as Mr. Spock used to say, crude but effective: I draped a bedsheet down from the ceiling, taping it down to the keypad so that each player could see only his half of the computer. The two players had their own little mapboards that they used to represent the game. Being left-handed, I took the right side of the computer; my opponent had to be right-handed to use the left side of the computer.

This two-player version of Tanktics was fun, but I soon lusted for more power. And so, in the summer of 1977, I ordered my first memory expansion kit. For just $149 I got a card with 4K of RAM! Of course, it was a kit, not a complete card. I spent hours and hours soldering RAM chips onto that board. There were 32 of those little chips, each of them with 16 pins. Along with all the addressing circuitry and the capacitors on each chip, there were more than 600 joints that had to be soldered and a single mistake could fry the board. Nowadays, the same amount of money will buy you 64,000 times as much memory, and you don't have to solder anything. Yes, those were the good old days.

After building the memory board, I built an interface circuit for it and built a big box for my KIM-1, including a power supply. (I nearly burned down the physics lab while testing it the lab had defective wiring.) I designed and built a pair of "tiny terminals," each with its own calculator-style keypad and a set of sixteen seven-segment digits, as well as an interface board for them. Now I was set! By early 1978, I had programmed KIM-Tanktics, a huge leap over previous efforts. This was a two-player, blinded game (each player could see only the data on his own tiny terminal). Each player had his own map and his own game pieces, and a small screen sufficed to block his view of the other player. With eight tanks per side and a large map, there was plenty of opportunity for interesting games. I expanded the game, permitting a variety of tank types and scenarios.