Maze War by Steve Colley and Greg Thompson

In the early 70s, the NASA Ames Research Center in California had a couple of Imlac PDS-1. These were the first graphic workstations, with vector-based graphics, a portrait screen, and the equivalent of 8kB RAM. Yes, you read that right: 8kB (4096 16-bit words). These were workstations (minicomputers as they were called at the time) for a single user, and therefore less well equipped than the contemporary multi-user machines. A PDP-10, for example, had more than 1MB (256 36-bit words). For this machine Steve Colley wrote a maze simulation that would develop into Maze War:

In 1973 I was trying to do simple 3D displays on the Imlac PDS-1. The first one I did was a simple rotating cube in which the hidden lines of the wireframe cube were removed. The complexity was in doing the sines and cosines on a slow machine with no multiply or divide.

A little later I had the idea of doing a maze where you were actually in the maze. To keep it simple, it had an array of 16×16 bits to define the maze. I worked out how to display the halls of the maze with perspective (apparently unchanged to today), and it worked quite fast so you could quickly move around in the maze. There was no way to display an overview of the maze, and you were to simply explore the maze and "solve" it. I had several different maze designs, and it was surprising how quickly you could learn them.

Maze was popular at first but quickly became boring. Then someone (Howard [Palmer] or Greg [Thompson]) had the idea to put people in the maze. To do this would take more than one Imlac, which at that time were not networked together. So we connected two Imlacs using the serial ports to transmit locations back and forth. This worked great, and soon the idea for shooting each other came along, and the first person shooter was born. I take credit for adding the "peek" capability to the game at this time. (It was too easy to get shot when you entered a new hall before you could turn to face the person who was already there). Also, due to the delay through the serial ports, it was very possible for two people to shoot each other with their bullets "passing" each other in the hall. We considered this an important feature.

Not long after this the Imlacs were networked to the IBM 1800, and someone else took over the program (I think Greg) and converted Maze to use the network and allow multiple machines to play Maze. A lot of time was used playing Maze (generally at night). Greg then took the program to MIT, and its spread was begun.

At MIT Maze War was rewritten for the Xerox Alto and subsequently sent to Xerox PARC. This version is the best known, indeed for a while it was generally believed that the game had originated from Xerox PARC, as the following excerpt of the docs to Planet of the Feebs, a game from the early 90s, shows:

Planet of the Feebs is a simulation game that is intended as a training aid for beginning programmers in Common Lisp. It is loosely based on the "Maze War" game, written by Jim Guyton, Bruce Malasky, and assorted others at Xerox PARC. In Planet of the Feebs, however, the players do not control their creatures by hand. Instead, they supply programs which control the creatures as they move around the maze trying to zap one another. The game presents an open-ended challenge, and advanced players may find themselves reaching deep into the AI bag of tricks as they try to build ever more clever and adaptable creatures.

This is also the version you see in the screenshot above, which is actually a pixel graphic created after the scan of a printout, made in 1985 or 1986, of Maze War played on a Xerox Star. It is probably not pixel-exact as far as the position of the minimap at the bottom are concerned, but the graphic of the eyeball avatar should be accurate.

It is rumored that there are versions for NeXT and Silicon Graphics workstations running IRIX too, but I could not verify this.

Legacy

Maze War enjoys less notoriety than Space War, which of course was far earlier, but it had the greater influence on subsequent game development. There were some direct follow-ups on home computers, mainly on the Macintosh, where AppleTalk provided a handy base for a networked multiplayer game: Maze Wars (1986), MazeWar + (1987), and Super Maze Wars (1992). Screenshots of MazeWar + still show the same eye avatar as on the Xerox Alto.

Then there was MIDI Maze on the Atari ST, which used the MIDI port to set up a network. It replaced the eyeball avatars with smilies reminiscent of Pac-Man, but was otherwise quite true to Maze War. In 1990 there was a similar game by the same developers for Gameboy, Faceball 2000.

But more important than the direct follow-ups is the indirect influence. The concept of walking through a maze and avoiding or fighting adversaries has been used over and over again, from the already mentioned Pac-Man over Wizardry to first person shooters, which even introduced multiplayer again. For whate is Quake III Arena after all, if not an opulent hi-tec version of Maze War?

Maze War

Legacy

Links