Real adventure-solving experts measure their ability by the number of Infocom games, they're solved. These games are not for the uninitiated. Any they are more than just tough. They have a special flavour that is all their own. Eugene Lacey went to Boston, Massachusetts, to meet the aristocrats of the American games business.
When you think of some of the brilliant arcade games that have crossed the Atlantic - titles like Impossible Mission and Summer Games II - you would think I was telling you lies if I said that the main chart toppers in the US are text-only adventure games.
Well, it's no lie. Take a look at America's most widely accepted software chart, the Billboard Top Fifty, and what you'll see there at the top, more often than anything else, is the Infocom range. It's not just that Americans prefer adventures that are really challenging, but they seem to prefer the text-only variety.
Infocom have been the undisputed number one source of these games over the last couple of years, leaving Scott Adams, the early pioneer of adventures, for dead. Games like Deadline, Enchanter, the famous Zork trilogy and, more recently, the Hitch Hiker's Guide To The Galaxy have all sold tens of thousands of copies. Hitch Hiker's Guide alone has sold over a hundred thousand units since its launch in February.
Infocom games are not cheap either, about £30 each, so it is easy to calculate that the adventure experts have earned a small fortune from their software.
The phrase 'adventure game' is not one that fits into the Infocom vocabulary easily. They prefer the grander description, "interactive fiction".
Of course, all adventure writers like to use the hype that they are actually turning you into the author of the action, but the advanced techniques used in Infocom games enable them to justify a term like "interactive fiction" more than anyone else.
The engine room of this interaction is the 'parser' - which is refined and adapted to each new game. All adventure games have a parser of sorts - a device that interprets the English in the command and makes a response accordingly.
Most parsers work by scanning through the computer's memory ot see if any of the words typed in are key words. If they are, the correct response is triggered. The trouble with this is that it gives the player the impression that the computer understands the sense of the English being entered - and this is obviously not the case.
For example, the adventure could begin in a shop. The description of the location describes you, the protagonist, as being dressed in a pair of trousers with deep pockets. The seasoned adventurer would check these pockets to see how much money is in them so that he might buy something useful.
If you typed in "count all the money in my pockets" a keyword parser would search the line to find a word it knows. In this case it would be "count". It can now check this to see what it can count - and comes back with the reply "£5". That makes sense.
The problems arise when you try to count anything else in the shop - say the number of other customers, cuddly toys, or aardvarks. To these questions, and any other at this location using the key word "count" the computer will give the same inane reply - "£5".
The Infocom parser works in a different way. Instead of just scanning for key words, it also analyses the sentence structure, so that it can interpret the syntax, and understand some of the sense behind the words.
An example of this can be found in the parser's handling of adjectives. You might type in "Open the door" - to which you would get the reply "Which door do you mean, the closet door or the bathroom door?"
The parser not only understands the key word 'door' but also that it has been used in an open or non-specific manner. Thus it analyses the context in which the word was used.
Of course not all syntax can be understood. The biggest, most powerful computer in the world could not correctly interpret the infinite variety of possible sentence structures. If it does not understand a word it will "complain". Bt with over three hundred different types of sentences identified and understood, complaints do not occur that often.
The effect of this more intelligent parser is that it gives the impression of having a conversation with the computer. Clever stuff - but not nearly as clever as Infocom's pride and joy - the Z System.
The Z system, pronounced "Zee", is the development machinery used by Infocom to write all their software. Years of programming time and more than a million dollars worth of computer hardware have already been invested in the Z system and they are still enhancing it.
Brian Moriarty, author of Wishbringer told me: "We have more raw computing power here than the average Third World Country".
At the heart of the Z System are two DEC 20 mainframes, housed in huge air-conditioned rooms, with a back-up team of engineers to maintain them. The system itself is best defined as a collection of software development tools with a heavy emphasis on interaction.
They know a thing or two about software development at Infocom - that was the founders' speciality at the Massachusetts Institute of Technology. Some of the costs of the Z System are being offset by a move into the lucrative world of business software with the launch of Cornerstone - an advanced database.
But make no mistake, the main work of the Z System is in developing adventure games. Six adventure writers can work simultaneously writing games from their computerised workstations.
"We can call up any piece of adventure code in a split second... if we see something we like we can take it down and play around with it... the code from all the games is stored in the Z System," says Moriarty.
There is something frightening, clinical almost, in Infocom's hi-tech approach to 'interactive fiction' - rather like the computerised novel writing machines envisaged by George Orwell in 1984.
Infocom would be the first to admit that a good parser and an advanced development system are no guarantee of a good game - sorry, piece of interactive fiction. They scorn graphics and produce their games only on disk, so that, parser aside, all of the combined memory of the C64 and disk drive can be used to store text.
Although the cream of the Computer Science faculty at MIT left to set up Infocom you can't help getting the impression that it's their literary achievements with 'interactive fiction' that they are proudest of.
If there was a Booker prize going for adventure game text you can bet Infocom would not rest until they had it safely in the cabinet.
The high point of these literary achievements was their collaboration with Douglas Adams, author of the Hitch Hiker's Guide To The Galaxy, on the adventure version. According to Joel Berez, Infocom's boss, the adventure is even more amusing and enjoyable than the book.
What the Infocom people are particularly proud of is the degree of Adams' involvement - referring to him as 'Dougie', a personal friend almost, and certainly more than just a straight business licensing deal.
Adams spent several months in Cambridge, just outside Boston, working on the text of the game, personally approving every single word.
Berez would like to do similar projects with other top ranking authors. "But only if we can get proper access to the author. We are interested in working with the writers, using their talents, and mixing them with ours... not simple marketing exercises... anyone can do that".
Albert Veza is Infocom's Chief Executive Officer and a former professor with some pretty impressive academic credentials. Before joining Infocom, as one of the founders in 1979, he was an Associate Director of the Laboratory of Computer Science at MIT - one of the best acaemic posts in American computing.
It was Veza who was the spark that ignited Infocom - persuading some of his high-flying colleagues at MIT to join his new company.
Marc Blank is in charge of product development. Another ex-MIT man and company founder, Blank was the author of the firm's first big hit, the famous Zork. He also developed Deadline on the 'Z System' and designed the famous English language 'parser'.
Softly spoken Joel Berez is the business brain behind Infocom. Like the others he was in at the beginning and has a string of MIT letters after his name. Berez built the company up from a small, tight team, into a major software publisher with dozen of employees. He looks after the day to day running of Infocom and also finds time to act as an executive officer of the Software Publishers Association.