The Days of Text and Parsers

Once upon a time, text adventures ruled the computer universe. Life easily condensed into simple verbs, takeable objects, and two-word commands. Passages were twisty and narrow (and frustratingly all alike). Babel Fishes came in vending machines. So what happened? Where have all the good Grues gone? 

Lara Crigger


This article originally appeared in Computer Games Magazine #185

“You are standing in an open field west of a white house, with a boarded front door.”

These are the immortal opening words of Zork, the most famous of text adventures—or “interactive fiction,” to go by the fashion of the day. The success of Zork transformed its developer, Infocom, from a hodgepodge of crab-racing computer geeks to a multimillion-dollar company. The era birthed some of the greatest games ever written, such as A Mind Forever Voyaging, Deadline, and The Hitchhiker’s Guide to the Galaxy.
Then, abruptly, it was gone.
But not for long. Fittingly, the genre’s resurrection was much like its inception, over 30 years ago. Text adventures resurfaced on the Internet, slowly regaining their fan base and momentum.

An Adventurer Is You
It began, as most things do, with trying to impress a girl. Two of them, in fact.
In 1972, ARPANET developer Will Crowther faced a painful divorce. Feeling alienated from his two young daughters, he decided to create a game for the three of them. Crowther, an avid caver, crafted a computer simulation of Kentucky’s Bedquilt Cave from survey data he’d taken himself. Although his creation wasn’t much more than a plotted-line tour of the cave, his daughters loved it. He called the game Colossal Cave Adventure.
In 1975, Crowther created a revised version for his PDP-10 computer and released it on the infant ARPANET, a progenitor of the current Internet. Despite the then-staggering 300KB memory requirement, the game quickly spread on the network.
Eventually, it appeared on the computers at the Stanford Artificial Intelligence Lab, where a resident researcher named Don Woods stumbled on it. After finishing the game, Woods asked Crowther for permission to expand it. Crowther readily agreed. Woods added a substantial fantasy element, giving the game trolls, elves, and even a Mount Doom-like volcano. He named the newly revised edition Adventure.

Anything Can Happen
Crowther and Woods’s creation continued to spread, and in 1978, the young Scott Adams discovered Adventure. Hungry for more, he wanted to create a similar game for his TRS-80; however, given the tiny memory of the computer (only 16KB), the task seemed impossible. Adams was undaunted. “I decided to just invent my own adventure writing language and compiler,” he recalls. With his new language, he wrote Adventureland.
Later that year, Adams ran an ad for Adventureland in a computer magazine. “When I saw no one was actually selling computer games, I thought it would be a good idea to help those, like myself, who wanted to play,” he says. The ad made history, as Adams was the first person to commercially distribute a computer game. To handle demand for Adventureland, Adams and his wife Alexis formed their own computer game company, Adventure International, which would put out games for the next seven years.

Let There Be Grues
Adams wasn’t the only one captivated by Adventure. Crowther and Woods’s game had migrated onto the computers at MIT’s Laboratory for Computer Science, where it found many fans.
Among them were Marc Blank, Tim Anderson, Dave Lebling, and Bruce Daniels. But their enjoyment soon turned to irritation. To move through the world or manipulate objects in Adventure, you had to type directions into the command line for the game to “parse” or decipher. Frustratingly, Adventure could only handle two-word inputs, precluding more natural and complex commands. So, in 1977, the four of them started work on an improved version of Adventure, and Zork was born.
Although Zork kept the command-line interface from Adventure, its parser was much improved. It could sift through an English sentence and diagram the phrases. Now, instead of having to type stunted directions such as “HIT THIEF,” players could give longer commands like “Hit the stupid thief with the empty bottle.”
Zork spread like wildfire. Because it ran on the lab’s mainframe, which was open to anyone, people all over the world were soon playing the game.

Infocom Begins
In 1979, a few idealistic MIT students, including Anderson, Lebling, and Daniels, decided to form their own company named Infocom. There was no business plan, no model, not even a product, but the founders knew they wanted to continue working together outside MIT, possibly by building business software.
To stay afloat, the company needed a product, and it needed one fast. Naturally, the company looked to Zork. But, like Adventure before it, Zork had memory requirements hopelessly beyond what any home computer at the time could handle.
Just as in Scott Adams’ case, their solution was to design a new language and a new virtual machine just for text adventures, named the Z-machine. In addition, Infocom created a new compiler called ZIP (Z-machine Interpretive Program). To make a game run on any specific computer, programmers simply had to write a ZIP for that system. Since many different computer models competed in the budding home PC market, this portability would be crucial to Infocom’s success.

Grues Gone Wild
Over the next five years, the games from Infocom swelled in popularity, and the company became the biggest name in text adventures. Early games like Deadline, Starcross, and Planetfall met wide critical acclaim and commercial success. Later ones, such as The Hitchhiker’s Guide to the Galaxy, A Mind Forever Voyaging, and Trinity, were also considered brilliant works of fiction.
From the beginning, Infocom used unique business tactics. For instance, games appeared not just in computer stores but in bookstores as well. Since newer text adventures didn’t make the older ones obsolete, both stayed side by side on store shelves. Games came with “feelies,” or extra items like pens, pamphlets, coins, maps, even sample drugs. These feelies often were crucial to beating the game (thus acting as a subtle defense against software piracy).
Infocom’s success had just as much to do with the people in the company as it did with quirky marketing strategy. Most of the employees working at Infocom were in their mid-twenties, without families. “The Infocom group wasn’t just our work base, but also our after-hours friends,” recalls former staff writer Steve Meretzky (author of dozens of games, including Planetfall, Spellcasting 101, and The Space Bar).
The atmosphere was one of creativity, ambition, and fun. Infocom programmers were known as “Imps,” short for Implementors (that’s how they spelled it). They held hermit crab races on a makeshift racetrack named “Drink ’em Downs.” When the goldfish in the courtyard pond mysteriously went belly-up, employees conducted a murder trial to catch the suspect.
“At Infocom, I had conversations you could never expect to have any place else in the world,” says former Imp Bob Bates. “It was great.”

Cornerstone for Failure
As the game business soared, Infocom looked to try something new: making business software. In 1982, it started work on a database package called Cornerstone.
Selling business software sounded great on paper. Profit margins were much higher: Infocom games generally sold for $30 to $50, but database packages retailed for hundreds. However, from the start, Cornerstone was racked with problems.
“The main problem was basically all the money spent marketing it,” says Meretzky. Personnel and resources for Cornerstone were funded by money Infocom hadn’t yet made, and cash for new offices, employees, and advertising was spent at an alarming rate.
Money might not have been a problem had Cornerstone been successful. But it wasn’t. Although it debuted in 1985 to rave reviews, Cornerstone only sold 10,000 copies. “If it had come out maybe a year or two earlier, it would’ve been a different environment,” says Meretzky. “It was mostly just a case of bad timing.”
Even worse, the industry experienced a painful crash in 1985. Costs skyrocketed, but game sales slowed. Financially, Infocom was screwed, and debts and layoffs were inevitable. But even layoffs didn’t help. “[Cornerstone] put Infocom in a situation where it just wouldn’t have been able to survive independently,” says Meretzky.

Infocom: Game Over
At the end of 1985, Jim Levy, the CEO of Activision, offered to buy the struggling company. Infocom had little choice but to accept, but by most of the Imps, the merger was seen as a godsend. Activision promised to save Infocom from certain annihilation, and, better yet, it swore to stay out of the company’s internal affairs.
Six months later, Activision ran into its own money troubles. The board fired Jim Levy, bringing in Bruce Davis as CEO. “There was a total climate change when he took power,” Meretzky says. Davis had been the only member of the board who opposed the Infocom merger, and he quickly inflicted a series of ill-suited, heavy-handed decisions on the struggling company.
Davis required Infocom to use the Activision packaging plant instead of its in-house one, doubling packaging costs. Newer games now replaced older ones in stores, cutting the life expectancy of Infocom games. To fill the empty shelves, Davis ordered Infocom to produce eight games a year, instead of its usual four or five.

Graphics Killed the Text Star
Moreover, computer graphics had finally matured, giving Infocom’s competitors an edge. Without time or money to spare on experimentation, Infocom slowly fell behind. More importantly, the Imps themselves resisted the change to graphics. “I don’t think they embraced the idea of graphical adventures,” says Bates. The company waited until the late 1980s to produce quality graphic adventures, but by then it was too late.
Infocom desperately struggled to make money under Activision, but its games just weren’t selling as well as they once had. By 1989, enough was enough, and Activision shut Infocom’s doors for good.
“It was a dark day,” says Meretzky.
After the demise of Infocom, Bob Bates founded Legend Entertainment with his friend Mike Verdu. Together, they ran Legend for the next 15 years, releasing titles such as Eric the Unready, Timequest, and Meretzky’s Spellcasting series.
Legend was lucky. With Infocom gone, it seemed like certain death for interactive fiction.

The Internet Age
Although commercially defunct at the end of the 1980s, text adventures were still very much alive on the Internet. Fans quickly nailed the mechanics of making games. A collection of amateur-designed programming languages surfaced on the Web, such as TADS, Hugo, and Inform. The languages allowed anyone with enough time, persistence, and coffee to create interactive fiction.
And soon they did. The Annual IF Competition, where writers like Emily Short and Andrew Plotkin got their start, began in 1995. In 1997, Scott Adams released all of his old SAGA games on his website. Sites dedicated to text adventures mushroomed.
Interactive fiction started to move in new directions. “During the 1980s, Infocom couldn’t have put out a game like the ones today,” suggests Nick Montfort, author of the book Twisty Little Passages. Brief experimental games like Galatea, an interactive conversation with a statue, and Shade, a one-room apartment horror game, revolutionized the genre.
Recently, some authors have attempted once more to make money from their efforts. Short’s City of Secrets, commissioned by the band Secret-Secret, was the first paid interactive fiction in years. Peter Nepstad boldly attached a $20 price tag to his 1893: A World’s Fair Mystery, which has met with small but steady success. Yet for most authors, making money isn’t an option.

Back to the Future
“Interactive fiction still has many directions for development,” says Short when asked about the future of the genre. “So many techniques have to be refined, and so much remains to be invented.” As long as authors have time, ideas, and text editors, there will be interactive fiction.
In 2000, if someone had asked whether 2D gaming had a future, the answer would’ve been a resounding no. However, handhelds like the Nintendo Game Boy and DS have helped keep alive the 2D side-scroller. Perhaps something similar will happen with interactive fiction.
In the meantime, there’s always Zork.

It began, as most things do, with trying to impress a girl. Two of them, in fact.

This article originally appeared in Computer Games Magazine #185