by Brian Dear
The long hours turned into long months, and then years. Moria continued to evolve and consume Battin’s and Duncombe’s time. Duncombe would eventually flunk out of Iowa State. He was not the first, and was by no means the last, PLATO-addicted college student to do so. But in the long run, he would discover that flunking was not as bad a thing as it might have seemed at the time. “I learned more on my own doing Moria,” he says, “than probably any of the classes. The games are always pushing the edges of what’s possible. And by doing things that are difficult or are operating-system-like in a lot of respects, there’s all kinds of multitasking things you have to think about, and interlocking things you have to think about, and it’s really, by the time you’ve solved all those problems, you’re working on a fairly advanced level.” Duncombe and Battin both ended up with decades-long software careers.
—
In the history of online computing, one pattern continues to repeat: new programs, websites, phone apps, and games of all kinds will arise as a reaction to previous programs, websites, phone apps, and games—the new developers attempting to apply new ideas to what they’ve seen in existing applications. On PLATO across the 1970s, this pattern would not only repeat over and over, but at an ever more frequent rate. Witness first the Big Board games, then Airfight and the space war games like Conquest and Empire. Perhaps the greatest progression and evolution of programs exemplifying this pattern is that of the dungeon games: Moria was in the middle of this multiyear evolution. Simpler, cruder games had appeared prior to it, and more complex games would come after.
It is instructive to take a look at some of those simpler games, especially ones written by pioneering authors who lacked the luxury of any prior art, who lacked the luxury of being able to react to that which had come before. What one finds is that the original influences were offline: Tolkien’s Hobbit and Lord of the Rings, certainly, but also, in 1974, the release of a tabletop fantasy role-playing board game called Dungeons & Dragons by Gary Gygax and Dave Arneson.
Perhaps the earliest dungeon-related game on PLATO was one by Reginald “Rusty” Rutherford III in a lesson file called “pedit5.” (Legend has it that an earlier game, with the equally obscure name of “m119h,” pre-dates Pedit5, but no one remembers who wrote it or when exactly it appeared—probably 1974. Like Pedit5, it suffered the fate of being disallowed by the higher-ups and was soon deleted from whatever PLATO account it belonged to. No physical evidence of it has been found.)
Rutherford did not fit the usual pattern of a new wave PLATO kid: he was already in his mid-thirties in 1975, seeking a PhD at Illinois, with a family and one child (named Reginald Rutherford IV). He had a job working in the PLATO lab in the basement of the physics building, writing TUTOR code for Professor Paul Handler, who, while nominally a physicist, had taken a great interest in the world population explosion and ran a group called “Population Dynamics” that was pumping out impressive PLATO lessons on the world population crisis. “I got paid effectively an hourly rate, not tremendous, certainly not by today’s standards, but it was very, very interesting and there was a fair amount of free time between assignments.” During that free time he got involved in the brand-new tabletop D&D. “About the spring and summer of ’75, my friends in the gaming group at Illinois had gotten deeply involved in this….It was a fierce game. It was awfully easy to die at level 1. Well, we kind of got really deep into it to the point where not much else was happening in our lives.” They weren’t quite pulling all-nighters, “but 2 a.m. was not that unusual,” he says.
One day Rutherford thought to himself, “Gee, could I put this on a computer, using these wonderful visuals you have available on PLATO? And I sat down and I figured it out.” As Handler’s TUTOR programmer, Rutherford had access to the PLATO account for the Population Dynamics project, and found a number of lesson files created with the names “pedit1” through “pedit5.” He grabbed them and got going. The result was a simple, single-user “dungeon crawler” game, much influenced by D&D. “It is the year 666—the year of the Beast,” the introductory page of the game declared. “In the country of Caer Omn, near the town of Mersad, stands the ruined castle of Ramething. Beneath the castle lie the terrible dungeons of Ramething, an incredible maze of rooms and corridors, occupied by horrid monsters and piles of ancient treasure.”
You were invited to create a character, name him, and enter the dungeon, initially in a corridor. The display was simple: on a PLATO IV terminal, the screen was nearly all black, with the word “CORRIDOR” on top and “EXIT” below. The corridor was depicted in a top-down fashion, with a crude little knight character standing in the hallway. One wall of the corridor had a doorway. You could not see what lay more than a step or two ahead, as if the corridor was dimly lit. Nothing else was seen on the screen. Like so many PLATO games, you “moved” by using the arrow keys. “W” took you forward, “d” to the right, “x” back, and “a” to the left. Upon pressing “w” you advanced and noticed more doorways along the walls of the corridor. Eventually you would reach the end, at which point the corridor split to the left and right, with more doors in the walls. Behind those doors were rooms, or more corridors. At any point you might encounter monsters.
The game was a hit, despite the fact that Rutherford never formally “released” it to the PLATO community. “News of the existence of Pedit5 simply spread by word of mouth,” he says. He then discovered the age-old problem of popular lessons on PLATO: it was hard to make changes and “condense” the TUTOR code because people were always playing the game, and he would have to kick them all out to run the condense, annoying the players. Soon he resorted to adding code to the game that immediately kicked out users with certain signons who were routinely running the game during the day. Another problem: having only two lesson files, Pedit4 and Pedit5, severely restricted how much game and player information he could store away. “The available storage space only allowed for a single-level dungeon with forty to fifty rooms,” he says. “The dungeon design was the same for every user, but the monsters and treasures were random—created at the same time as a new character, and stored with the character record. Only about twenty characters could be stored; when the game became popular, this turned out to be a real hassle.”
In an interview nearly forty years later, Rutherford still laughed at the ruthlessness of his game’s design. Pedit5 was downright lethal. There were no limits to the wrath of Randu, the TUTOR command used to generate random numbers, which determined where the monsters were when your game character was created. “The idea was, go in, make a reasonable score,” he says. “If you get wounded, get the heck out because you can’t take the next fight. Until you get a reasonable level, you don’t want to try for the depths of the dungeon. And besides, I set it up so it got tougher as you went further in. On the other hand, you might run into a dragon in the hall on the first corridor. Tough luck. Unfortunately, that terminated your character.”
Pedit5 would disappear soon after it was created, due to its shady unofficial status as an unsanctioned game created in the serious Population Dynamics account. Then it would reappear, only to disappear again. The very existence of the game became a random, unreliable thing. For the growing wave of PLATO game authors, this would not do.
—
Two such authors were undergrads Gary Whisenhunt, majoring in psychology and political science, and Ray Wood, majoring in electrical engineering, at Southern Illinois University (SIU) in Carbondale. In the 1970s some universities around the country were experimenting on a small scale with PLATO, often leasing two or four terminals for use on their entire campus. SIU was such a place: there was a single PLATO terminal in the basement of the library. “It was common for one person to run the terminal and another five to ten people to be in the room with you watching the terminal,” says Wood. “Gary and I would meet during these ‘joint viewing sessions.’ He seemed to be more normal than the rest of the people—he actually had outside interests….He liked messing around with PLATO, bu
t he knew there were other things to do. Gary is probably the ‘proto-geek’ that all other geeks were designed after. He was into technology, but he also had a lot of interests outside of technology. Gary had a girlfriend who was relatively normal. He listened to rock and roll, drank beer, etc. The other guys didn’t. So we kind of naturally gravitated to each other. It was great to be around someone who may not have had a life outside of computing, but at least knew that there was life outside of PLATO.”
Still, they kept coming back to PLATO. They played games, among them Empire and Pedit5. Like everyone else, they grew tired of the repeated disappearances of Pedit5. Soon, they decided to do something about it: write their own game. The lesson file name for their game: “dnd.” Like Internet domain names decades later, lesson names on PLATO were unique. When somebody grabs a domain today, that’s it, it’s taken. Back in 1975, Whisenhunt and Wood knew Dnd was a good lesson name. They grabbed it.
“I taught myself to program on the PLATO system,” says Whisenhunt, “and liked it so much that I switched majors to computer science.” Soon, they were pulling late-nighters and all-nighters, hacking away at the code. The goal, says Whisenhunt, was fun. They didn’t share Rutherford’s addiction to the D&D board game (though Wood was familiar with the board game and enjoyed playing it), nor was their new game anywhere near as lethal. Instead, it had in-jokes that fellow SIU students would recognize. The game started out with just two levels, the mazes of which they laid out by hand on graph paper. It took hours. Whisenhunt decided to write a “maze editor” online, which gave them a one-hundred-fold boost in productivity. They spent a great deal of time tuning the combat play, making each level as “fair” as possible, taking into account the player’s overall score and skill level, and the level’s difficulty: features the file-space-strapped Pedit5 had lacked.
Whisenhunt and Wood had another advantage over Rutherford. They controlled the account that the “dnd” lesson was in. Whisenhunt had become known to the library staffer who was supposed to be responsible for the presence of the CERL PLATO terminal in the building, but beyond that didn’t know much about the system. Here was this student, Whisenhunt, who hung around the terminal a lot, seemed to know the system in and out, and was even programming on it. The library offered him a job, and soon he became the de facto administrator for the SIU PLATO account, including all of its files, of which Dnd was one. Unlike the travails that plagued Pedit5, “there was no chance,” says Whisenhunt, “that Dnd was ever going to be deleted.”
As was always the case on PLATO, word traveled fast when new games appeared on the scene. Dirk Pellett, who had been attending Caltech but then transferred to Iowa State, found the PLATO terminals there, got hooked like the other Iowa State crew, and eventually found out about Dnd and became not only a fan but also a contributor, offering a long-distance stream of suggestions and ideas for improvement. He never met Whisenhunt or Wood in person, but his suggestions were copious and good enough that in time, Whisenhunt and Wood, heeding the call of their college work (and grades), were happy to give access to Pellett to dive directly into the Dnd TUTOR code. Meanwhile, Dirk’s brother Flint, enrolled in graduate school at the University of Illinois, and soon became an online collaborator as well. In time the Pelletts took over the game and added extensive enhancements.
A sense of mischief and wit permeated the game’s design. Every time you opened a treasure chest, there was a chance you might get injured or fail to open it, at which point the game program might scold you with a “Clumsy dolt!” message. There was a spell called “Kitchen Sink,” which, of course, you could throw at a monster during combat. They named a monster “The Glass” after an obnoxious freshman student they knew. With a nod to Star Trek, the game included “transporters,” which, to add some spice to the experience, were sometimes unreliable, sending you somewhere you didn’t expect, or causing damage to your character’s health. Says Dirk, “At first, I added a lot of magic items with their names and actions taken almost verbatim from D&D, merely adapted to the computer game Dnd, like ‘cube of force’ and ‘horn of blasting.’ Later, I deliberately avoided copying anything from other games, and made up new monsters and items and features that I had never seen anywhere else while keeping the humorous tone.” Two such items were “orange glop” and “roving sludge.”
Another compelling design consideration imbued in Dnd was the idea of consequences. Everything seemed to have a consequence, a price. You might find an incredible amount of gold in a treasure chest. You could be greedy and try to carry it all, but if you did, the game made your character grow tired far more quickly—thus making you more vulnerable to dangers in the dungeon. If you had a lot of money you could buy a “potion of revival” to use as a last resort if your character died. You could then use the potion, but doing so came at a price: you lost all the magic items you were carrying. “A great game gives the players the freedom to make a vast number of choices,” says Dirk Pellett, “some of which are more beneficial than others, and some of which are disastrous, and lets them figure out which is which.”
Dnd is historic not only for being one of the earliest dungeon games on any computer, but also for a particular game design feature that many other games would copy over the coming years and still employ today. In the computer gaming world it’s a concept called “the boss”—it might be a monster, it might be one hard-to-obtain item, or it might be both. In Dnd the boss item was something called the Orb, and finding it was the ultimate quest for a player. Naturally, the Orb was hidden deep inside the dungeon at its bottom-most level. And, naturally, it was protected by the fiercest monster, a classic example of what would become known as a “boss monster”: a Gold Dragon. Kill the Gold Dragon, and you could grab the Orb. (In a similar way, Empire’s short-lived Doomsday Machine was a boss monster, the ultimate enemy requiring extraordinary skill and luck to defeat. A more recent game, DOOM, also has a famous, huge, powerful boss monster. For some games the boss monster appears as a sort of final test of a player’s skill; other more recent games may sprinkle boss items and monsters throughout the experience of the game.)
Early on, Whisenhunt and Wood discovered that more players were successfully grabbing the Orb than they had anticipated, but they could not figure out how they were doing it. So they made a pilgrimage up to CERL to find out. They discovered that players there had discovered a way to cheat: step one toe into the corridor or a room, just enough to find some gold or other valuable item, then step back out to safety. Then step back in, just enough to see if Randu had deemed it time to deposit another item in their path; if so, they’d grab it and step back out. Players were doing this for hours at a time: sometimes many hours, all-nighters. It required a determination and obsession that took the game’s authors by surprise. “We never thought that anyone,” says Whisenhunt, “would have the patience and would spend the time to do such a simple thing over and over again to gain large amounts of wealth. I guess that was the precursor to videogame addiction.”
—
Sometime in late 1974 or early 1975, John Daleske began work on another game at the same time the major effort was under way to develop Empire. His new game was called “Dungeon,” and unlike Dnd and Pedit5 it supported multiple users. Daleske claims it was the first multiuser dungeon (MUD) game on PLATO, perhaps on any computer. Unfortunately, the game struggled to achieve popularity, possibly due to being too complex and not as playable as other games out at the time or under development. Responsiveness was crucial; the gamers would not tolerate anything that didn’t run at the speed of thought. Another factor may have played a role in Daleske not devoting more time to perfecting Dungeon—and Empire, for that matter. In May 1975, a doctor diagnosed Daleske with terminal cancer, “which kind of stifled my creativity a bit,” he says. “Thankfully,” he says, “it was a misdiagnosis, but I find it did affect me for about a year.”
Word continued to spread online around the PLATO system, and around the physical locations lucky enough to have terminals, t
hat more and more games were becoming available. The same process that had attracted Whisenhunt and Wood, and the Pelletts, and Daleske and the rest of the Iowa State gang, was repeating ever more quickly and to ever more people. Not just the fact that there were games to play, but that it was possible to write your own—look how many people were making the leap—as long as you could scrounge up the always rare file space and didn’t mind weeks of late-night coding.
Pedit5 evolved into a new game called Orthanc (another Tolkien reference, to Saruman’s evil tower at Isengard), by Paul Resch, Larry Kemp, and Eric Hagstrom. In late 1977, a new game appeared, Oubliette, written by Jim Schwaiger, with help from John Gaby, Brancherd DeLong, and Jerry Bucksath. David Sides, a teenage gamer at the time, remembers Schwaiger as “probably one of the larger egos of anyone I had met at PLATO. He used to call everyone a ‘pud,’ that was one of his big words as I recall. He would walk around with a beret, he was a tall thin guy, real unusual character.” Oubliette, another MUD, like Moria showed a small 3D maze on the screen (small because the maze walls could plot more quickly, livening up the game) and offered multiplayer capability. Like Dnd it offered a method of teleportation. However, if Randu was in a bad mood that day, it was possible to teleport a player into solid rock, killing off the player instantly. (Players referred to the situation as “getting stoned.”) Says Brian Blackmore, a UI student during the era, “If you teleported into rock, that was the end of your character. And you couldn’t even get your items up. I mean, it was doomed. So teleporting into rock was like the most feared thing that could happen to you.”
