Three years after a government physicist named William
Higinbotham created the first video game, Tennis for Two
(see page 65), some students at MIT invented a
game called Spacewar! Here’s their story.
If you ever get a chance to see a picture of the Electronic Numerical Integrator and Computer (ENIAC for short), you probably won’t recognize it for what it is. Completed in 1945, the ENIAC is considered to be the first practical digital computer ever made.
ENIAC was the supercomputer of its day. It was as big as a three-bedroom house and weighed more than 60,000 pounds. It contained more than 18,000 vacuum tubes, each one the size of a lightbulb. And because the tubes burned out so frequently (2,000 a month on average), ENIAC was out of order about a third of the time.
Even when it was working, ENIAC couldn’t do very much: Its operators programmed it manually, spending hours or even days flipping switches and rewiring circuits. And ENIAC couldn’t store these “programs,” so each time the operators finished one computational task (calculating the path of an artillery shell, for example) and wanted to start another (nuclear weapons research), they had to flip the switches and rewire the whole computer all over again. ENIAC didn’t have a keyboard or video screen, and it was more than 10,000 times slower than a modern personal computer.
Computers evolved slowly. Computers with video monitors, for example, were extremely rare through the 1960s. Only three universities in the entire United States—Stanford, the University of Utah, and MIT—had one.
So if it took 15 years for computer technology to progress to the point where exactly three American universities could own computers with video screens, how long do you think it took students at these universities to program the first video games into these supercomputers? A couple of months, at most.
What’s another name for a suckling calf? A “bob.”
MIT’s legendary Whirlwind computer, for example, had a demonstration program called Bouncing Ball. Technically, it wasn’t a video game because the viewer didn’t do anything. You could only watch as a ball appeared at the top of the screen, then fell to the bottom and bounced around the screen, with a thwok! sound coming from the computer’s speaker at each bounce. Eventually the ball lost its momentum and settled on the floor, finally rolling off to one side and out of the picture, at which point another ball would drop from the top of the screen.
But Bouncing Ball and the computer “games” that followed weren’t supposed to be taken seriously. They were just things the early programmers dreamed up to amuse themselves and to demonstrate the number-crunching power of the Whirlwind computer. The best games were designed to tax the abilities of the computers to the limit. But other than that, they were “hacks,” as they were called even then—programs with no constructive purpose whatsoever. The people who made them called themselves “hackers.”
Mouse in the Maze was one of the earliest hacks. Designed for a supercomputer called the TX-O, it consisted of a mouse (the animal), a maze, and a piece of cheese. Using a special light pen, the player drew a maze right on the screen and then placed the cheese in the maze. Then the mouse searched through the maze and ate the cheese, leaving crumbs wherever it ate. An “improved” version had the mouse searching for martinis and after drinking the first one, staggering around the maze looking for the rest. There were other games—Tic-Tac-Toe, and a pattern-generating program called HAX—but nothing that would hold the interest of players for more than a few minutes.
Then in the fall of 1961, the Digital Equipment Corporation donated a state-of-the-art computer called the Programmable Data Processor (PDP-1) to MIT. The PDP-1 was smaller than the TX-O, much faster, and a lot easier to use.
Even before it arrived at MIT, the PDP-1 had captured the imagination of the university’s Tech Model Railroad Club (TMRC). Club members had already spent several years “requisitioning” computer equipment from around campus and using it to automate their huge model railroad. They’d also spent a lot of time learning how to program the TX-O, so when the PDP-1 finally arrived on campus, they were already the best computer programmers around. And they were ready to hack.
South-pollywog: According to Sesame Street, Kermit the Frog is left-handed.
The PDP-1 came with no software at all; almost everything had to be programmed from scratch. MIT students were doing much of the programming for little or no pay, so the professors who controlled access to the computer agreed to give them plenty of hack time in return. And what did they do with the hack time? They invented games.
Many TMRC members were science-fiction buffs; so it didn’t take them long to decide what kind of game they wanted to create for the PDP-1: a space game—one that would push the computer’s processing power to its limits. “The basic rules developed quickly,” MIT alumnus J. M. Graetz remembers. “There would be at least two spaceships, each controlled by a set of console switches.… The ships would have a supply of rocket fuel and some sort of a weapon—a ray or a beam, or possibly a missile.”
TMRC member Steve “Slug” Russell wrote the first version of the game, taking about six months and 200 hours of computer time to do it. The game he came up with consisted of two spaceships, one wedge shaped, the other long and thin, which flew around the screen and battled one another by shooting “torpedoes”—dots of light—at each other. Each ship was controlled by a different set of four toggle switches on the PDP-1 console. One toggle switch made the ship rotate clockwise; a second made it rotate counterclockwise; a third switch provided thrust; and a fourth fired the torpedoes. (Ever play Asteroids? The controls were pretty much the same.)
Both ships were controlled by human players. There was no way to play the computer as your opponent, because once everything else had been programmed into it, “there wasn’t enough computing power available to do a decent opponent,” Russell remembered. He named his game Spacewar!
The “average” American CEO’s pay has increased more than 600% since 1990.
As soon as Russell got the game up and running, other TMRC members began making improvements:
• At first the game had no stars in the background, but the blank screen made it difficult to tell whether slow-moving ships were drifting closer to each other or farther apart. So Russell added a series of random dots… but they didn’t last long. Using an astronomical reference book, another club member, Pete Samson, programmed in an accurate map of the night sky, including the relative brightness of each star.
• Another TMRC member named Dan Edwards inserted a sun—complete with accurate gravitational field—into the center of the screen. Now instead of sitting still in empty space, the ships were constantly being pulled toward the sun, and if they crashed into it they were destroyed. That helped to make the game more interesting, because it inserted an element that was beyond the players’ control. It also made strategy more important, because skilled players could figure out ways to use gravity to their advantage.
• Graetz added a feature called “hyperspace.” If a player got into trouble and was about to get killed, flipping the hyperspace toggle caused the ship to disappear for a few seconds and then reappear somewhere else on the screen, hopefully someplace less dangerous and not close to the sun. Graetz also inserted an element of risk—if a player hit hyperspace one too many times, their ship would be destroyed.
The improvements made the game more interesting… which created a new set of problems. Spacewar! addicts played for hours on end, frantically flipping the toggle switches on the $120,000 computer until their elbows hurt. Needless to say, the computer wasn’t designed with that kind of use (or abuse) in mind.
Rather than risk breaking the $120,000 computer, a couple of TMRC members scrounged wire, switches, and other parts from the model railroad and made another innovation—individual game controllers that they connected to the PDP-1 with lengths of electrical wire. Now the players could stand back from the computer and play as furiously as they wanted to, without damaging the computer or getting sore elbows.
Walt Disney World generates about 56 tons of trash every day.
So what do Steve Russell and the developers of Spacewar! have in common with Willy Higinbotham, creator of Tennis for Two? Two things—they never patented their invention; and they never made any money from it. Digital Equipment ended up giving Spacewar! away as a diagnostic program and it became popular with computer engineers and programmers all over the country… including a University of Utah student named Nolan Bushnell, who later founded Atari.
It turns out there was a way to make money off of Spacewar!.… it just involved waiting for the price of computer technology to come down. In the mid-1970s, well after the video arcade craze was underway, an MIT graduate student named Larry Rosenthal decided to build an arcade version of Spacewar! as his master’s thesis project. The game he created—sold as Space War and then as Space Wars—happened to hit the arcades in 1977, the same year that Star Wars hit the big screen.
Space War(s) had nothing to do with Star Wars, of course, but nobody cared. It quickly became the most popular arcade game ever…until a game called Space Invaders came along in 1978.
As for Steve Russell, he not only never made a penny off the game he was largely responsible for creating, he never even graduated from MIT. He relocated to Seattle and got a job with a computer time-share company. One of his responsibilities was hiring local high school kids to come into the office and see if they could get the computers to crash. Lots of kids tried, but, according to Russell, only one kid had enough computer savvy to make the computers crash every time, no matter how hard Russell and his colleagues tried to thwart him.
The kid’s name was Bill Gates. He never graduated from college, either.
The next phase of the history of video games
will take us to the video arcade, so turn
to page 314 and let’s play Pong.
No matter how hard they try, scientists can’t train houseflies to do tricks.