Text Based Adventures: Before the Graphics

Before the graphics, there was the word, and everyone knows words have power. Except in gaming, where graphics have all the power. Nowadays, it seems that if your game hasn’t either licensed the latest graphics engine or created one from the ground up, then it had better have some damned good features to make up for it. Things were not always so. There was a time, before personal computers became capable of producing anything worth the moniker of ‘graphics,’ when text was king. ‘Text Adventure Games,’ ‘Text Games,’ ‘Adventure Games’ and occasionally, ‘Graphic Adventure Games,’ are all terms used to describe a genre of computer games which relied on nothing more than a textual description of the player’s location, and a text prompt allowing the player to control which actions their avatar performed. Today, the genre is known as ‘Interactive Fiction’ or IF, and if you’re sitting comfortably, I’d like to give you a quick look at its history.

Colossal Cave Adventure, later known as Adventure or Advent, was the first IF game. Created by Will Crowther, a programmer at Bolt, Beranek & Newman (BBN). A keen caver, Crowther began to create Adventure in the early 70's as a way to spend time with his daughters after the break-up of his marriage. Much of the layout of the game is a faithful recreation of the Mammoth Cave system in Kentucky, a complex which Crowther had spent considerable time in during the previous years.

The game was originally written in FORTRAN on BBN's PDP-10 computer. Due to the limitations of the operating system, filenames were limited to six characters and capital letters only, leading to the game being known as ADVENT. BBN were one of the lead contractors and developers of ARPANET, the precursor to the Internet, and in 1975, Crowther released a version of Adventure over the ARPANET.

A Stanford University researcher called Don Woods later discovered the game on a computer in the Stanford AI Lab in early 1976. He contacted Crowther and asked for his permission to expand upon the game. Crowther agreed, and Woods expanded the game, adding elements like a troll bridge and a vending machine which dispenses fresh lamp batteries, which feel at odds with Crowther’s original recreation of the Mammoth Caves. That being said, somehow the game does feel like a coherent whole, which is testament to the skill of the authors. The expanded game was then released on the early Internet (as opposed to ARPANET), where it was found by a group of MIT students and staff, the key figures of which were Joel Berez, Mark Blank, Dave Lebling and Albert Vezza.

By turns inspired by Adventure's (at the time) groundbreaking imagination and frustrated by its limited vocabulary, the group began not only to create their own ‘text adventure game’, but to create a programming language of their own, specifically designed for their needs. The game was called Zork, although originally called simply Dungeon, and took two years to finish. At the time Zork was MIT jargon for an unfinished program. By the time the game was done, the name had stuck.

Because of the difference in capacity between the mainframe computer which Zork had been written on (a DEC PDP-10 in the MIT computer labs) and the size of the average household computer at the time, the game had to be cut down into three separate parts. The authors of the game formed the company Infocom in 1979, and Zork I was released for the TRS-80 in 1980, followed by Zork II in 1980 and Zork III in 1981. As noted above, the authors had found themselves frustrated by the limited vocabulary in Adventure. In particular, the parser was limited to a simple two-word, verb-noun structure, “Go North”, “Take Rock”. The parser in Zork was much more powerful, capable of handling full sentences or multiple commands, “Look under the rug”, “Pick up the lantern and light it”. Although the setting of the game seems clichéd by today's standards (a nameless adventurer, an underground empire), we must remember that there had been nothing like Zork before. Not only was the game itself an amazing technical achievement, but the quality of the writing was very high, enough even still to reward today's more distinguishing or demanding players.

The Infocom developers were by no means the only people working on text adventure games during this time. As word spread over the ARPANET, games were being written in UCLA, the University of Alberta and Cambridge University, to name a few, it was the nature of the times that Adventure and other games were picked apart and expanded as they were shared. Indeed, Crowther’s original version of Adventure has been lost, and only Woods’ version is still available.

If today's developers are faced with a challenge porting new games to multiple formats, they have it easy when compared with the developers of the early 80’s. At that time, Apple, Atari, IBM, Amstrad, Commodore, Spectrum, Texas Instruments and more all had personal computers on the market, leaving developers with the difficult choice of attempting to manually port their game to multiple platforms, or stick with one format and hope for the best. The developers at Infocom, or ‘Implementors’ as they called themselves, came up with a solution to this problem which would ensure their lead over their rivals for years to come, and cement a place for themselves in gaming history.

Their solution came in two parts; the first part was the 'Z-Machine', essentially a virtual computer capable of running the game, which would be coded in a proprietary computer language, created by the Infocom team; the second part was the Z-Machine Interpreter Program, or ZIP, a program which would interpret the Z-Machine language and make the computer running the game behave as if it were an actual Z-Machine. The language created by the Infocom team to make this all work was called Zork Implementation Language, or ZIL, itself based on MDL, the language Zork had been written in. In this way, Infocom were able to create versions of their games for multiple formats, without having to create each version from the ground up.

Although Infocom were the most successful interactive fiction company, they were certainly not the only company producing games at the time, and not only in the U.S.

Adventure International, founded by Scott Adams, was actually the first company to commercially release a text game, advertising Adventureland via small ads in computer magazines in 1978 and 1979. Adventure International went on to release over 20 games, including games based on Spiderman, The Incredible Hulk and The Fantastic Four. When the company unfortunately went bankrupt in 1985, the copyright for its games were handed to the bank, but when they eventually reverted back to Adams, he released the games as shareware. Topolgika was formed in Britain in 1987, releasing eleven games for a variety of formats. They eventually stopped releasing text-based games, and now only develop and release educational software for children.

Other notable companies were; On-Line Systems (later Sierra Online) who released Mystery House in 1980, the first text game to include graphics; Level 9 Computing (again from the U.K.) who released their first games in 1983; Australian company Melbourne House, who released The Hobbit in 1982.

The mid-to-late 80’s were undoubtedly the 'golden age' of interactive fiction, with several Infocom titles holding places in industry sales charts in the US. Unfortunately, it wouldn't last. Interactive fiction flourished while PCs didn't have the power to render graphics of any kind, or at least any graphics worthy of the name. As the power of the average desktop PC increased, graphics in computer games became much more widespread, computer games began to feature graphics more and more, and even though most of the graphics in these games were rudimentary, gamers (proving that some things never change) were seduced by graphics, regardless of the actual quality of the games.

Later examples in the genre featured static pictures along with location descriptions more frequently than they had previously, but it wasn’t enough, and once graphics came to be expected in games, rather than looked at as an additional feature, the death knell had been tolled for commercial interactive fiction.

Ironically, the same thing which sounded the death knell for interactive fiction would prove to be the one thing which would allow a small but dedicated community to keep it alive. The Usenet newsgroups rec.arts.int-fiction and rec.games.int-fiction, focusing on creating and playing interactive fiction, respectively, gave people a place to come and meet other fans. While all of these fans were dedicated to keeping interactive fiction alive, one group did more than any others to ensure its survival. The InfoTaskForce, as they called themselves, worked hard to reverse engineer Infocom's Z-Machine. Once this had been done, they created a version of the interpreter in C, which they then released into the public domain.

Prior to this, authors of interactive fiction had been limited to commercial design programs such as The Quill Adventure System (released in 1983), Graphic Adventure Creator (released in 1985), Professional Adventure Writer (released in 1986 and based on Quill) or The Adventure Game Toolkit (released in 1987).

Despite the variety of commercial packages still available in the early 90's, the Text Adventure Development System (TADS), by Michael J. Roberts, was the most widely-used program for creating interactive fiction. TADS has been maintained and updated by Roberts, and was recently updated to version 3. In 1993, Graham Nelson released Inform, which proved to be very popular, and quickly overtook TADS as the preferred development tool for interactive fiction. The most recent version of Inform, version 7, introduced a ‘natural language’ style of coding to the interface. Regardless of the authors’ choice of tool, both Inform and TADS were instrumental in a resurgence of interest in interactive fiction, now that authors had easy access to a variety of high-quality tools which gave them an easy way to create new games.

With the community thriving, Kevin Wilson, an undergraduate at Berkeley University at the time, held the first Annual Interactive Fiction competition in 1995 – originally split into TADS and Inform categories, the ‘IF Comp’ now judges all games against one another, regardless of the programming language used, and focuses on shorter games, playable in two hours or less. Eileen Mullin, editor of the XYZZYnews, announced the XYZZY Awards in 1996. The Awards currently judge ten categories, including Best Game, Best Puzzles, and Best NPC. Adam Cadre, a well-known author of interactive fiction, announced the Spring Thing awards in 2001, intended for games longer than those entered in the IF Comp. The competition went on hiatus in 2004, but returned in 2005 and is now being run by Greg Boettcher.

The interactive fiction genre is unlikely to become commercially viable ever again, but for those of us still playing, writing and above all, enjoying interactive fiction, that isn’t the point. Within the genre of interactive fiction there are enough sub-genres to satisfy almost any taste, from straight-forward dungeon crawls to murder mysteries, from comedy to Lovecraftian horror thrillers. There are games which defy any genre staples, either by flouting known conventions or by blurring the lines between said genres. There are games in the interactive fiction genre which, perhaps more than any other genre, epitomise the ‘games as art’ debate. Some games, like the IF Comp entries, are perfectly suited a couple of lunch breaks’ worth of play, other will keep you hooked for days and still repay repeated visits.

Interactive fiction is not for everyone, but if you ever enjoyed typing your way through dark caverns or futuristic landscapes, scribbling notes as you went, or if you’re simply curious, then I would encourage you to spend some time with the community and try some of the fine examples the genre has to offer.

Recommended and Notable Games

• The Zork Series (Zork I: The Great Underground Adventure, Zork II: The Wizard of Frobozz, Zork III: The Dungeon Master.) by Infocom.
• The Hobbit by Beam Software (Game of the Tolkien book).
• Trinity by Infocom (Based around the Trinity nuclear tests).
• Anchorhead by Michael S. Gentry (Lovercraftian horror game).
• Galatea by Emily Short.

Recommended Websites

• IF Comp. Homepage of the Interactive Fiction Competition. Also contains links to very good introductions to Interactive Fiction. Probably the best place to visit first.
• XYZZYnews. An online Interactive Fiction magazine with a massive back catalogue of articles covering a multitude of categories. Also home of the XYZZY Awards.
• Inform. Probably the most widely-used language for creating Interactive Fiction at the moment.
• TADS.Homepage of TADS, still a popular Interactive Fiction language.
• Brass Lantern. A fantastic website, hosting plenty of articles for beginners, reviews and news.
• rec.arts.int-fiction Newsgroup. A group geared towards discussing Interactive Fiction.
• rec.games.int-fiction Newsgroup. A group for geared towards the playing of Interactive Fiction, including hints and tips.
• Emily Short’s webpage. Emily Short is the pseudonym for a well-regarded Interactive Fiction author. As mentioned above, her game Galatea bears investigating. Her website contains articles, essays, reviews and links all pertaining to Interactive Fiction.
• The Interactive Fiction Archive. A massive archive (unsurprisingly) of articles, tool, interpreters and much more. A little hard to navigate until you get used to it though.
• iFiction. A website hosting online versions of (mostly Infocom) Interactive Fiction games. Well worth visiting in your lunch hour.
• SpringThing. Homepage of the SpringThing Awards.