Advent of Computing - Episode 11 - Spacewar! (the Game)

Episode Date: August 25, 2019

It really seems like in the last decade video games have gone from a somewhat niche hobby to a widespread part of our culture. Nowadays, there are a multitude of ways to get out gaming fix. Consoles, handheld game systems, and even smartphones make video games more accessible than ever. But when and how exactly did video games start to creep into the modern consciousness? In this episode we look at some of the earliest video games and how they came to be. Like the show? Then why not head over and support me on Patreon. Perks include early access to future episodes, and stickers: https://www.patreon.com/adventofcomputing Important dates in this episode: 1962: Spacewar! Developed

It really seems like in the last decade or so, video games have gone from a somewhat niche hobby to a widespread part of our culture. Nowadays, there's a multitude of ways to get our gaming fix. Consoles, handheld game systems, and even smartphones make video games more accessible than ever before. But when and how exactly did video games start to creep into the modern consciousness? Now, a lot of people would probably think back to arcades, or companies that produce games in the home like Atari or Nintendo. But it turns out that the first arcades only started showing up

around a decade after the first video games, and for home systems even later than that. It also turns out that the first generation of these games were all developed and ran on expensive mainframe computers. So what were these games like? And how did programmers tinkering in their spare time lead to a multi-billion dollar industry in the modern day? Welcome back to Advent of Computing, I'm your host Sean Haas and this is episode 11, Space War, the game.

Now we're going to be diving into some of the first video games ever made, so I think it really behooves us to have some way to distinguish between what is a true video game and what's just an idle demo or a toy. And I guess the best way to start this episode is to figure out that working definition. Now, you may be tempted to say that a video game is any game that you play on some kind of video display. And while in the modern day that's definitely the case, there are early examples that use a teletype or paper tape in some instances as an interface, and those examples are definitely video games, despite not being connected to any TV or screen. We could say that any kind of game that's controlled by a processor or a

computer would count as a video game, but that also runs into some big issues. Some early home systems such as Pong were implemented without a CPU, and Pong is definitely a video game despite that. Instead, I'd like to propose a definition along the lines of this. A video game is a game that is interacted with and functions via some kind of digital means. I know that's broad and not very satisfying, but I think it gives us a good working definition for what's to come. And with that definition in mind, it may not come as a shock, but many of the earliest games were designed on and written for digital computers. But I'm not talking about PCs. In this sense, I mean mainframes. It turns out, the computers ended up being a great platform

for game development, even going as far back as the 50s. So today, I want to talk about some of those early mainstream video games, the culture that springs up surrounding them, and the eventual first video game. Computer games, or at least something approaching that, would start appearing as far back as 1950. The first example that we know of being Birdie the Brain, a single-purpose computer designed and built by Joseph Cates to play tic-tac-toe against a human opponent. Sometime even earlier than that, the Cathode Ray Tube Amusement Device, or as I like to call it, the best name ever developed in gaming, was patented. This was a device capable of simulating the arc of a projectile on a glass screen. But I wouldn't really consider these fully video games in their own right. First of all, both these games were analog, and well, there's nothing wrong with that, but it doesn't adhere

to my earlier stringent definition. Also, I just don't understand analog computing so it kinda spooks me. Secondly, and more importantly, both of these games existed in somewhat isolation. In the case of Birdie the Brain, it was only shown at the 1950 Canadian National Exhibition. Even worse, the cathode ray tube amusement device was never manufactured, only living its archaic analog life on paper and in the patent offices. And these are just two examples out of a handful, maybe a dozen or so, of these very early proto-video games. While early analog games are interesting, they never really went anywhere. These machines were cut off from the cultural mixing pool that

would eventually lead to the creation of something that we'd recognize today as a fully realized video game. And sometime in the late 50s or 60s, the stars would align and create the perfect home for that birth of the video game. MIT. Yep, that MIT, the university in Massachusetts. This may initially sound strange, but let me explain a few key factors that made MIT a good environment for this type of innovation, and then maybe this will make a little more sense. On the basic level, MIT was one out of a handful of places in the country where someone could get access to a computer. We're still in the era of

mainframes here, so any computer would be enormously expensive to purchase and exceedingly difficult to use without extensive training and know-how. Only large businesses, the government, research labs, and universities really had the needs and means to have computers around in any capacity. And in the case of MIT, it turns out there were computers to spare at the time. The college has always been on the forefront of computer development. Some of the first computers were built in labs at MIT, and researchers there had to keep pace. In 1955, the newest computer to come out of the lab was the TX0, or TICSO as a lot of

workers at MIT pronounced it. This was one of the first transistor-based computers. Rather than using bulky and unreliable vacuum tubes for logic operations, TIXO used the smaller and more reliable transistors. There's a lot more technical details to this, but suffice to say that the new cutting-edge computer was smaller, faster, and easier to maintain than the previous generations of machines. But importantly, or more importantly for our story, the Tixo had a screen. It was a 12 inch vector display made from a modified oscilloscope. Not something that you really see nowadays. The screen itself was round with a green phosphor on a black background.

Really seeing one today, it looks more like it should be a sci-fi prop than cutting edge technology. A screen may seem unimportant, but in the 50s it was anything but. Most computers in this era had no graphics to speak of, they were either interfaced with using buttons, punch cards, or a typewriter terminal. To be fair, Tixo wasn't the first computer to have a screen attached, but it is definitely an early and important example. While the hardware side of Tixo was impressive for the time at least, what really matters is how it was used. The computer could be operated by one person. Earlier machines had to have a full staff of engineers and programmers just to keep them

humming along and feed in new jobs. But TIXO was reliable and small enough that it could be run independently. Initially, that wasn't always the case. The machine was being used for research, after all, so it was still kept behind a slight bureaucratic curtain. But by 1958, TX-2 would roll along, and the earlier model would quickly become obsolete. So that year, it left its original home at MIT's Lincoln Lab, where it was designed and built.

But it wouldn't go far, as it was transferred to what would soon become MIT's Artificial Intelligence Lab, just a little bit across campus. And in this new space, it was open for use to really any student that was working in that lab. J.M. Graetz was one such student researcher. In his essay, The Origin of Space War, he describes working with TIXO like this. Quote,

Instead of being sealed into its own special chapel, it sat at one end of a typically large, messy MIT research space. With its rack of exposed circuitry, power lines and meters, and its long, low, L-shaped console, the TIXO looked for all the world like the control room of a suburban plumbing station. And the thing of it was, you were expected to run it yourself. End quote. So here's the winning combination. A computer that's more easily accessible than earlier systems, pulled out of the chapel so to speak. A fancy display that you can draw graphics on.

And a lab full of computer science students that are given more or less unsupervised run of that system. Now anytime you get a bunch of students, especially in a STEM field like computer science, and give them access to a fancy new tool, they're going to put it through its paces and see just how much it can do. In the case of Tixo, something special would start to spring up. The programmers in the AI lab would spend countless hours and sleepless nights working on the machine, partly for actual research and university projects, but also trying to pull off fancy tricks and feats of programming to show their peers. Inside the lab, all the side projects and computer trickery

were called hacks, and the people carrying them out would end up being called hackers. Some of the earliest computer video games would be put together on this machine by AI lab hackers. So what did these games look like? The most easily recognizable was, well, Tic-Tac-Toe. Again. This game seems to have been popular in the demo scene in this era, and what I find interesting about the Ticso implementation is that it didn't actually use the machine's fancy display screen. Instead, it displayed on a paper printout. Tic-Tac-Toe is cool, especially if you're playing against a computer opponent, but it's the kind of game that you can just play at a normal desk

using a pen and paper without the computer. It's not really using the new electronic medium to its fullest potential. What I think is the more interesting game developed on Tixo was called Mouse in the Maze. This game took full advantage of the computer's screen, as well as a pointing device called a light pin, which allowed users to interact with the program by pointing at and clicking on the screen. The gameplay was simple, you construct a simple maze on the screen,

then place cheese around that maze. When you're ready, a mouse controlled by the computer will travel across the maze and try to find and eat the cheese. There was no score kept and no timer, so there wasn't really a way to win the game, but this is a lot closer to a recognizable video game. Importantly, this is a game that you can't really play outside of a computer simulation, unless you work in a lab with mouses and mazes, and it uses the specialized hardware of the Tixo to its advantage. Luckily, this was just a start. In the coming years, Tixo would become obsolete even to the AI lab. In 1961, a new computer would hit MIT, Dex PDP-1. This new machine was actually based on Tixo, even coming with its own display similar to

the earlier machine. So it was easy for programmers to switch over to this new computer. But the PDP-1 wasn't just a straight clone of Tixo. It was faster and smaller as you'd expect for a new computer. But beyond that, it was just plain more reliable. The Tixo could be operated by a single person, but it was a long and complicated process to turn the machine on and then turn it off at the end of the day. process to turn the machine on and then turn it off at the end of the day. In the case of the PDP-1,

you just had to flip a single switch and it was up and running. To turn it off, just do the same thing. The bottom line is that the PDP-1 was a familiar computer, but more powerful and more accessible. And the access was key, since it made it easier than even before for hackers at MIT to work on their software. And, of course, one of the top priorities was to work up some way to show off the new computer. A demo for all the non-programmers that would be passing through the lab, something to drum up interest in the department and help keep funding flowing. help keep funding flowing. Going back to Graetz, he described in his essay the idea of three key points that would be used as a guideline to make an ideal demo. To quote, 1. It should demonstrate as many of the computer's resources as possible, and tax those resources to the limit. 2. Within a consistent framework, it should be interesting, which means every run should be different.

And 3. It should involve the onlooker in a pleasurable and active way. In short, it should be a game. End quote. To modern ears, that pretty well describes a video game. You have a showy program that's interesting and inviting. So, what did that look like on the PDP-1? Well, pushing the computer wouldn't be that hard. The machine came with only 9 kilobytes of RAM. To put that into perspective,

the file for this podcast is going to be about 25 megabytes once I'm done with it. So to store that all in RAM, you'd have to use about 2700 PDP-1s. On top of that, it didn't have any hardware operations for multiplication or division. This may seem like a small issue, but this means that you have to use a slower software operation to do more complex math. Over a lot of divides, that adds up to a big drop in performance. Point 2 and 3 are a little harder to quantify. What makes something engaging and interesting? That's a tough question to answer.

One way to do this would probably be to provide a compelling theme. And to the team at the AI lab, that meant, well, space. Most of the programmers on the team were already obsessed with pulp sci-fi novels, and the space race had just recently kicked off, so why not get in on the hype? And just like that, the idea for the first recognizable video game was born. The world would come to know it as Space War! With the exclamation mark. The primary creator of the game was Steve Russell, a researcher in the AI lab just coming off developing the first Lisp interpreter. Russell was the one to come up

with the idea of Space War, and did most of the programming. But the game was really a team effort, with code from many other programmers making its way into the final product. So what made Space War so different from earlier games? Really, it just came down to the fact that Space War was a more developed video game. Before I get too far into the weeds, let me explain what the game was actually like. Space War is a game for two players. Each player pilots a spaceship around the screen while trying to shoot the other player with torpedoes. Each player has a limited amount of fuel and ammunition.

At the center of the field is a quote heavy star, so the whole game you have to navigate your ship under the influence of that body's gravity. Get too close to the star and fall in, you lose a life. Scoring was simple, you hit your opponent's ship, you get a point. Once everyone has used up all their lives, the game's over, and whoever has the most points wins. To add to the challenge, each player also has a limited number of hyperjumps that teleport your ship to a random place on screen. Now, that may sound simple enough, but it's still engaging even today. One of the key points of Space War is that, well, you can win the game.

There's an end goal. And that makes it engaging and addictive, since you can quantify and improve your skills of the game. Many earlier games lacked this, really making them more like large overblown toys than full video games. Another point in the engaging category was the theme of Space War. Games like Mouse in the Maze were only a step away from just plain simulations of existing games, but Space War pushed the limits in this regard.

I think this is an important note, because it's a key example of using the medium of video games to full effect. No one, especially in 1961, was ever going to be able to experience battling with a spaceship in real life, but for a few fleeting minutes, you could experience it at a computer console. Russell and others recognized this and took pains to make the game as immersive as possible, well at least as much as possible, well, at least as much as you can do with 1960s technology.

On the lighter side, this went as far as having backstories on in-game rules. For instance, the hyperjump feature was explained as an experimental device, hence why you only had a limited number of jumps before it broke down, and why you couldn't control where you exited from hyperspace. This philosophy extended into the actual programming of the game as well. Special routines were developed to display convincing explosions when a ship was hit. On the extreme end was the background of the playing field itself. If you look at the game, you can see that it takes place in space, so the background

has to reflect that. But a random star field wasn't good enough. Instead, Russell used an existing project at the AI lab called Expensive Planetarium to create an accurate star map of a swath of the real night sky to play on. Space War does sound remarkably similar to a modern game. Sure, the graphics have been upgraded, but the core is there. And while a lot of what made it a fully realized video game was planned out, many choices were due to sheer necessity. One example was the game's controllers.

Earlier computer toys and demos were controlled via switches, text, or other input devices that were normally used to control the computer. There wasn't such a thing as a specific game controller. text, or other input devices that were normally used to control the computer. There wasn't such a thing as a specific game controller. In another departure from the norm, Space War was played using its own custom input device. Each gamepad had two toggles for moving the ship and engaging the hyperdrive, and a button to fire. So, why the change from earlier input methods to a custom controller? Originally, Space War was played on the PDP-1's front panel switches, but it was too cramped for two people to comfortably sit and play. That, and the fear of damaging the expensive computer,

made separate controllers the order of the day. The idea of a scoring system with set rounds also came out of necessity. Remember, Spacewar started as a demo to show off the PDP-1 at MIT. Part of that would be having the game running at open houses. The problem here was that the team needed a way to keep visitors from hogging the game. Scoring and limited lives made it easy to kick people after their round was over. So that's the broad brushes of Space War. Russell finished most of the programming for the game in time for the 1962 MIT Science

Open House, where the game was a huge hit. In fact, the game was a huge hit all over campus, and in retrospect, it's little wonder. Space War does look a lot different than the games we may be more familiar with, but the feeling and fun of a more recent video game is still there. So the next question I want to tackle is how did Space War and the broader idea of video games first start to spread? I mentioned at the top that Space War was special because it didn't exist in isolation, and I mean that in a few different ways. The most direct example I can think of is the spread of Space War itself.

And, well, the original game was originally developed by hackers and academics, so a lot of its early spread was contained to that subculture. Partly spreading by word of mouth and partly by papers and articles, Space War would start to seep into the computer scene in the early 60s. And when people heard about the game, they wanted to try it and they wanted to learn more about it. This was spurred on by the openness in the hacker community at the time.

Russell and his colleagues happily shared Space War, sometimes sending printouts of their source code for the game, or even compiled versions of it stored on paper tapes. And as programmers left the AI lab for other institutions, they'd take the game with them. In a matter of years, Space War would spread from coast to coast, nowhere with a computer was safe from the new craze. And craze isn't an exaggeration. computer was safe from the new craze. And craze isn't an exaggeration. Many colleges tried to ban space war at one time or another, mainly just due to the distraction it caused on campus. MIT wasn't even safe.

Shortly after its release, the AI lab had to ban playing the game on their PDP-1 during regular lab hours. There were even reports from some labs that compulsive players were wearing out and breaking computer equipment during space war binges. Even DEC, the parent company that made the PDP mainframes, got in on the action. Sometime in 1963, they started to ship space war with new PDP-1 systems as test software. The game was designed to push computers to its limits, so it made a good test of new installations. But think about that for a minute. In 1963, you could buy a computer for $120,000, or roughly a million dollars today,

and it would come with a pack-in game. But don't be mistaken, there was more than just a hype around Space War. An important part of the proliferation of this game was the fact that the source code was so readily available. Today, we would call a program like this open source, and in the hands of hackers, this meant that new versions of Space War started to spring up pretty quickly. Some modifications were simple, changes to the playing field, hyperjump behavior, or graphics, but outside of the small tweaks were full ports, conversions of Spacewar! for other computers besides the PDP-1. To quote Greats again, Programming tapes were already showing up all over the country, not only on PDP-1s, And just like that, in a matter of years after the first conception of a video game,

it would spread out all over the nation, and in the coming decades, its reach would only grow deeper. And in the coming decades, its reach would only grow deeper. So I think it's just about time to close this story out. Space War would continue to spread and spawn clones and updates for decades to come, but its reach wouldn't be limited to computers. In 1971, nearly 8 years after Russell and Co. completed the first game at MIT, Spacewar! would be released in a more publicly accessible form, as Computer Space, also known as the first arcade video game. At the time, the company that made Computer Space was called Syzygy Engineering, but the very next year, they changed their name to something more familiar. Atari. So next time you sit down to play a video game, remember that you have a small group of hackers at MIT to thank. And thank you for listening to Advent of Computing. I'll be back

in two weeks time with another episode, and since the last few have been based on Big Iron, I'm planning on covering a topic more micro in scale. Until then, if you like the show, take a minute to share it with a friend. You can also rate and review on Apple Podcasts. If you have any comments or suggestions for a future topic, go ahead and shoot me a tweet. I'm at Advent of Comp on Twitter. And as always, have a great rest of your day.