Milliways: Infocom's Unreleased Sequel to Hitchhiker's Guide to the Galaxy

Backstory

Douglas Adams (left) and Steve Meretzky in a promotional photo from 1984.

Written and designed by two legends in their respective fields, game designer Steve Meretzky and sci-fi author Douglas Adams, the first Hitchhiker's Guide game was a tremendous success upon its release in November 1984. It quickly became Infocom's bestselling game, selling over a quarter million copies in the two years after its release. (It ultimately became Infocom's second-biggest seller of all time under Zork.)

Even before they'd finished writing the Hitchhiker's game, Meretzky and Adams were considering two sequels based on the second and third books of the trilogy. In an email to Meretzky, Douglas Adams wrote down some notes from a design meeting, including a list of "some rooms we discussed (some might be kept for future games, which would be blatantly advertised at every opportunity)." The first three places on the list, "Milliways (need reservation to get in), Norway, Krikkit (placed off limits for protection of the galaxy, until you buy Game #3)."

A sequel seemed like such a sure thing, they mentioned it in the game's ending. In the final scene, the Heart of Gold sets down on Magrathea and you exit the ship. "Slowly, nervously, you step downwards, the cold thin air rasping in your lungs. You set one single foot on the ancient dust — and almost instantly the most incredible adventure starts which you'll have to buy the next game to find out about."

But shortly after Hitchhiker's was released, Douglas Adams proposed making Bureaucracy instead, a game inspired by his real-life experience dealing with government red tape after an address change. He proposed the idea to Infocom in late 1984 and Infocom agreed.

But by May 1985, it seems like Adams lost interest. "In light of my discussions with Douglas yesterday and this afternoon at lunch, I cannot imagine finishing this game for 1985," wrote Marc Blank, Infocom's co-founder. "Douglas' attitude seems to be less than I require in order to do a collaboration."

With Adams busy working on Dirk Gently and other book-related projects, Bureaucracy languished. In his absence, Adams suggested a friend for the job, British writer Michael Bywater. Three years later and with no less than 10 co-authors, the game was finally released, crediting "Douglas Adams and the Staff of Infocom" on the cover. (The convoluted story of its development was detailed in an Easter egg in the final release of the game.)

Without Douglas Adams on board, the future of a Hitchhiker's sequel was in limbo.

Developing Milliways

Three huge problems plagued Milliways from its start in 1985 until it was shelved in 1989: no solid game design, nobody to program it, and the backdrop of Infocom's larger economic problems.

Searching the Infocom Drive, the very first mention of a Hitchhiker's sequel is an email from Marc Blank to Stu Galley in May 1985, about replacing Bureaucracy in the production schedule. "My worry is that derailing Bureaucracy might cause the Douglas Adams timetable to be upset," he wrote, "leading to our missing a chance at Hitchhiker's II next Christmas." By July 1985, Marc pushed back the Christmas 1986 deadline a full 18 months to June 1988.

Michael Bywater, as seen on his blog.

In 1987, Michael Bywater was flown in from London to stay at a hotel near Infocom headquarters in Cambridge, MA to work on Restaurant at the End of the Universe full-time . By July 1987, software manager Jon Palace wrote to VP of development Chris Reeve, "I get the impression that Bywater has been making design notes but has not taken advantage of the technical and creative resources here at Infocom. In other words, the work he's doing could be done anywhere, so I'm not sure why he's here. But given that he NEEDS to do what he's doing, I guess there's nothing else to be done." He continues, "I guess that, like Stu [Galley], I feel he came here ill-prepared, and we're paying his rent in Belmont for no good reason. He may as well be doing this work in London, running up his own phone bills."

Michael Bywater produced some preliminary design notes (reproduced in The Story section below), but it seems nobody was happy with it. In December 1987, Jeff O'Neill wrote that he heard "Bywater would be given the opportunity to work on the Restaurant until January," after which O'Neill would be required to work on it.

In January 1988, Plundered Hearts creator Amy Briggs started an explosive email thread after she was apparently asked to helm the project. She wrote to Jon Palace:

Amy Briggs

I just wanted to clarify in writing what we discussed about "Restaurant" last Tuesday -- what I will and will not agree to.

I will not sign a blank sheet of paper: I refuse to take responsibility for "Restaurant" in the state it presently is in -- not knowing who is creatively in charge, how much thinking has actually been done, or how much of a script is written. ...

What I'd like to see is a piece of paper, on which is an exact, REALISTIC schedule of how much creative input the programmer is to receive, and when, that Bywater and/or Adams has/have signed. Then I'll tell you whether I agree to take the project or not. ...

I'll be frank -- If you want me to code "Restaurant":

1. There is absolutely no possibility that I could have a finished product of "Restaurant" by October, 1988, even assuming I had a complete script of the game to work from before the end of January. It is doubtful I could make it by November, either.

2. The longer it takes to get a complete script to this side of the Atlantic, the less likely the game will come out in 1988 at all. If the rumor I have heard of a script in early March is true, then the game is right out.

3. The less script I get, meaning the more creative input I offer, would add time. Working with no script (so I develop the whole game, with a lot of Steve's input) would, I estimate, add 6-9 weeks to the length of the project (and that's a conservative estimate). If I am going to design the thing, I have to know by mid-February that the British contingent are out of the equation.

4. The more people designing the project, the more uneasy I get about the whole thing. A game cannot be written by a committee. If Anita or Bywater or Adams is writing it, fine, but not all three.

5. In any case, whether the game were designed by me or by them/him, I would be open to comments and criticisms. I will not, however, hold up the project waiting on these, nor change things once their window of opportunity has closed. I am trying to avoid the last-minute re-write of "Bureaucracy." This is something that I would like made totally clear to Adams or whichever collaborator is decided on, since I suppose you'd like to avoid lawsuits.

1+2+3+4+5 = I will not agree to this project if you want a product before October, nor if the other time-consuming factors come into play and you want it in 1988. I will not code for some nebulous quantity that may or may not offer comments and suggestions at some indeterminate time during the course of the project.

In early February 1988, while management searched for a project lead, Stu Galley was working on the Milliways skeleton file (playable in the next section) and Michael Bywater delivered a new outline.

But on February 24, management dropped a bomb — development of Milliways would be moved to the other side of the Atlantic, coded by British game developers Magnetic Scrolls and designed by Michael Bywater. Amy Briggs had declared her intention to leave the company and no suitable replacement could be found, so Infocom would not be involved closely in development. The reaction was swift and angry from the original founders of Infocom.

Dave Lebling, an imp (short for "implementer") who worked on several of Infocom's best games (including the Zork trilogy), wrote an angry rant to Jon Palace titled "Flame Alert." It provides some very interesting insights, in very candid words:

Dave Lebling

As you might have guessed, I was pretty depressed by today's events regarding Restaurant. It seems to me that we have been played for fools by Bywater, Anita [Sinclair, co-founder of Magnetic Scrolls], et al.

Anita's theory has always been that MS do H2. She presented this theory to Douglas [Adams] and me when I was in London last Spring. Douglas was clearly unimpressed. We paid Bywater a big chunk of money to design H2 last year, and he (to be charitable) did not come up with a compelling design. He did, of course, cash our check.

We now "have to" have H2 out in fiscal '89. This is understandable, since we have had to have it out in each of the previous four fiscal years. So as usual, we try to pressure an imp into volunteering. Amy, being the only imp not currently working on something, semi- volunteers, contingent on a design existing by right about now. Surprise, surprise: there's no design! Gosh, guess Anita and Michael will have to do it, Batman!

Okay, maybe this is a good idea. One person I talked to said, "When it sells 20,000 copies we'll finally be rid of Douglas Adams." That may be a little cynical. However, let's look at the alternatives.

If MS does it, we have no creative control. I cannot believe that after all this time and effort, we would not publish whatever they come up with, as long as it's reasonably bug-free. Look at how much trouble we are having getting TSP to do what we want, when we are officially collaborating with them, and they are a local phone call away. Anita is no fool; she knows this and will come up with exactly the game she wants. As it will be written in their development system, we won't even have the minimal control we had over Challenge.

Let's suppose for a moment that the game is a big success, sells Dorn's projected 100K copies. Does that benefit us? Sure, some; we get more money than if the game isn't done. It seems to me that it benefits MS much more than it benefits us. It is an official acknowledgment that we think they are as good as we are, that we think they can do this game better than we can, that our vaunted new technology is not needed for a good game, and so on ad nauseum. It sends every possible wrong message I can think of in the PR realm. It is problematic enough to think of publishing any MS game under our banner, but one which we "own" is really wrong.

Now, that's the optimistic scenario. Let's suppose the game is a bomb. Let's face it, Bureaucracy was a bomb. Jinxter is reputed to be a bomb. Bywater wrote both of them. There's Dornbrook's sure 100k seller converted into a $14.95 delisted product. If it's a bomb we get exactly the same bad PR as before, but in addition they say we can't even tell a bomb when we see one. "Infocom couldn't do it themselves, so they had this junk written instead. Boy are they going downhill."

This game is another camel. It's now Bywater, Sinclair, Paul X., and Douglas. If we bludgeon Steve or me into taking it over at alpha from a demoralized Amy, it's even more of a dromedary.

I don't know what the right solution is. I wish I did. The only way this game is going to get written and be a success is if we find someone on this side of the Atlantic who really wants to do it, and an absolute minimum of actual collaborators (preferably consisting only of Douglas [dream on...]) on that side. Until that situation exists, we should not consider doing it. Each solution, as for Bureaucracy before, is more makeshift than the last, more expensive, and more damaging to our reputation.

If we continue on the present path, it's going to be an abortion that will at best damage our reputation and at worst our pocketbook as well.

Joel Berez, president of Infocom, addressed the frustration himself the next day in a message from February 25, 1988:

Joel Berez

I thank you for your opinions, but you obviously appreciate the dilemma that we face. In a time when we are struggling to reach revenue levels that we allow us to maintain some semblance of a reasonable place to work, a sequel to a product that has brought us $5 million so far is extremely tempting.

For better or worse we have brought Bywater into the loop as primary writer because no viable internal alternative presented itself. So far we have failed to get useful work out of him, but Magnetic Scrolls has demonstrated some success. The choice here seems to be to use MS as best we can or to give up on Bywater and start from scratch. The latter option would of course put us back where we were a year ago, having to draft some imp.

Which brings us to the present decision point... It would have been nice if you [Dave Lebling] or Steve [Meretzky] had enthusiastically volunteered to take over from Amy. However, this project is getting fragile enough without having some conscript do the final implementation.

While I am not thrilled about having MS implement the game, it is certainly an available option. I may be naive about this, but I do feel that we can have a significant impact on the game without actually implementing it ourselves. For your information, it currently has the most relaxed schedule of any product we have in development.

I would tend to discount the conspiracy theory somewhat. For one thing Anita has put very little effort into convincing me to let MS implement the game. For another, Amy's decision to leave was presumably not influenced by Anita.

At this point we need to get past the "life is unfair" stage of the problem and select the best solution, even if that solution is not as wonderful as we might like. I would welcome any input on which solution to choose.

Behind the scenes, Marc Blank started working on finding an alternative to Magnetic Scrolls. He emailed Meretzky and Lebling on February 29, fishing for ideas. "Joel tells me that the current theory is that Magnetic Screws create the game," wrote Blank. "Isn't this a big loss for us with lots of bad messages sent to the industry? Not that it's my business, but isn't there some way to get it done in-house, even taking Michael, Anita, and maybe even Douglas out of the loop?"

Lebling responded, "The only way (apparently) that Joel will change his mind is if either Steve or I volunteer to do it, with Amy doing the initial up-to-Alpha implementation of M.S.'s design." Blank replied, "My reading of Joel's comments ... was that he'd be willing to drop all the others out of the picture if one of you or Steve were willing to do the whole thing."

Marc Blank continued the negotiations in a couple followup emails from February 29:

Marc Blank

Joel pretty much said flat out that he'd be willing to have an in-house imp do it "alone" (presumably with Douglas in a minor if any role), but he felt nobody would take it on. I got the distinct impression that he feels forced to go the Bywater/Anita route out of frustration. ...

Stevo said to me that he'd do the game, if freed from collaborators like Bywater and Anita. Joel said that was acceptable (at least, I think he said that.) Perhaps there's some chance of having that happen (which might be the best solution, under the circumstances...)

I really hope we can get this one back from the Brits.

These conversations inspired both Steve Meretzky and Dave Lebling, two of the best adventure game creators in history, to independently offer to take on Milliways, each with their own conditions. Dave Lebling emailed Joel Berez first on February 29:

I am sufficiently convinced that having Magnetic Scrolls do Restaurant will be a disaster that I will volunteer to do it, under the following conditions:

1. That I needn't start on it until after Shogun is finished.
2. That I would be an actual collaborator, not a "programmer."
3. That only Douglas Adams and/or Michael Bywater be involved on the English side.
4. That a relatively complete outline of Adams/Bywater's ideas for the game exist by the time I start work. I would be willing give opinions/advice on this design as it is written, to avoid a totally losing design.
5. That there be no "waiting for Godot" after I start work. Contributions from the English collaborator would be solicited, but that if he doesn't do any work after his initial design, so be it.

In short, I would want the setup to be as much like an ordinary in-house collaboration as possible, such as Marc and I did on the Zorks and Enchanter.

I would expect Anita's involvement to be zero if Bywater is not the English collaborator, and extend no further than prodding Bywater if he is.

The next day, Meretzky outlined his own conditions to Joel:

I recently heard (via Ye Olde Rumour Mill) that Restaurant may no longer be a Mandatory Product for Fiscal '89. If that's true, and if there's a consensus that it's worth delaying Restaurant just a bit more in order to do it in-house, let me list the circumstances under which I'd be interested in doing Restaurant:

a. If I can complete Zork Zero before beginning Restaurant.

b. If I'd be doing Restaurant as a full collaborator. (That is, as a co-designer/co-author/programmer, not just as a programmer.)

c. If the number of collaborators on Their Side was finite. The fewer the better, although I realize that 1 (or zero) collaborators is probably an unrealistic goal.

d. If I don't have to spend any time cooped up with a smoker.

If you substitute "Shogun" for "Zork Zero" in the first item, I surmise that this approximates Dave's position also.

Doing Restaurant is not my Dream Come True, but I do have paternal feelings about the Hitchhiker's IF series, and I agree with those who think that letting Magnetic Scrolls do Restaurant would be, overall, a Big Mistake. I think we'd do a significantly better job, and that's good economically for Infocom and good aesthetically for the Hitchhiker's series.

Joel Berez was grateful to Marc for getting involved. He emailed him on March 5, 1988, "Thanks also for your help concerning Restaurant. I don't know what you said to Steve and Dave, but they've both volunteered to do it (under perfectly reasonable conditions)."

A Hitchhiker's sequel created by Meretzky or Lebling would've been amazing, but it didn't get very far. No other mention of their involvement is in the archives.

The door to the Restaurant at the End of the Universe was finally closed on May 5, 1989, when Activision (renamed to Mediagenic, at the time) closed the Cambridge office and laid off 15 of the remaining 26 Infocom employees. Offering to relocate the rest to their offices in Menlo Park, only five accepted. Infocom was dead.

The Story

What would have Milliways looked like if it had been released? Several design notes by Stu Galley and Michael Bywater give us an idea. The first reference to the storyline is in a list of "Ideas for Next Project" from October 1986 created by Stu Galley.

MILLIWAYS or RESTAURANT AT THE END OF THE UNIVERSE

Takes up where "Hitchhiker's" left off. Manufactured planets, Deep Thought, white mice, time travel, 1001 verb tenses, digital watches, the Frogstar, Total Perspective Vortex, the End of History! (Does Douglas really want to work on this at this time? Does it matter?)

The plot changed significantly over the next three years, but Galley took a first pass in a set of design notes from February 13, 1987. In it, he proposes how the game should start and end, and a couple critical scenes.

1. It seems natural to include a scene in the restaurant, Milliways. Could be a bit of fun: strange parties, unctuous compere, self-introducing food. Perhaps there's an object there that you need to get. (It could be a SPORK, a spoon with sort of forky tines on the end. Or would that be a FOON?) It could be a vehicle from the car park -- Marvin has the keys. If you manage to re-enter Milliways at another time (oops! on another occasion), you will not meet yourself, "because of the embarrassment that usually causes." What about a visit to the Big Bang Burger Bar?

2. Given point 1, you must have a means (or several meanses) of time travel. In fact time travel instead of space travel could be the primary method of changing scene. In the original, the party got to Milliways by accident: in the radio version, a "hyperspatial field generator" overheated; in the book version, Zaphod's great-granddaddy screwed up the works of Eddie, the Heart of Gold computer. Maybe your trip to Milliways would require info from an anti-piracy device in the game package. Once at the restaurant, you can steal a timeship and go anywhen you want.

3. Given point 2, it seems natural for the "best ending" of the game to be your arrival on Earth before it's destroyed, which is the ending of both the first radio series and the second (namesake) book. The original route to this ending was an accidental landing on Golgafrincham Ark B, with its cargo of telephone sanitizers, marketing consultants, etc. (the ancestors of Earth's humans!). I rather like this bit, and hope we can work it into the game.

4. Okay, so what about the beginning of the game? The easy answer: take up the story where the "Hitchhiker's" game left off, namely the arrival on Magrathea. But in the original this arrival is followed by a travelogue of Magrathea and a flashback to the Deep Thought v. philosophers' union story (including the introduction of the "42" joke) and the joke about the true nature of mice. All funny bits, but I have a hard time envisioning how they can be made into interesting interactive versions. Perhaps you could time-travel to Deep Thought and interact with it yourself. The Magrathean catalog of planets on Sens-O-Tape could be useful.

On February 17, 1987, Steve Meretzky provided this feedback on the design notes:

Some thoughts upon re-reading your notes:

The Infinite Improbability Drive acts as a time travel device, as well as a space travel device and an identity changing device. The trips to the party (as Trillian), to Damogran (as Zaphod), and the Earth (as Ford), all involve going back in time. Therefore, the IID could be your time travel device, if you want to avoid the hassle of inventing a new one. (It also provides some continuity with Game One.)

As I think I said before, I think it will work best if you forget about the original stuff, like the Deep Thought flashback, and come up with new stuff instead. References to the original stuff can be used to provide familiarity for "old-timers" and to supplement the humor of the story line.

A month later, on March 13, 1987, Galley's second set of design notes outlined some innovative gameplay possibilities with third-person perspective, multiple viewpoints, text-based visual effects, and a parser with personality.

1. Shall we try to present multiple viewpoints in this version of the story? In the original version(s), the group of travellers became separated while visiting Magrathea, and again later on, with the story following either Arthur's or Zaphod's viewpoint. I have in mind a scheme to narrate the story in the third person, rather than the second, provided that the parser can handle declarative sentences, rather than (or in addition to) imperative ones. Example: ">ARTHUR SEARCHES FOR A GOOD CUP OF TEA. He fails to find any, but he does turn up the spork that he thought he'd lost."

2. If there are multiple viewpoints, how does one change viewpoint? Provisional answer: with the "verb" MEANWHILE. Example: ">MEANWHILE, ON THE HEART OF GOLD. Zaphod is still trying to persuade the computer to unlock the sauna. Ford is chatting up Trillian." Now the story begins to sound more like traditional fiction, with an omniscient viewpoint. What happens to the scene that one leaves when switching to a new viewpoint -- does it go on by itself? Answer: perhaps it doesn't matter. If one returns to the old scene to find that time has passed, one can use time travel to return to the scene as it was left.

3. And now for something completely different. I've been experimenting with textual "visual effects", along the lines of the Boysenberry computer display in Bureaucracy. These effects depend on the ability (available only in Interactive Fiction Plus) to put any character on the screen at any place. One effect I've tried out makes the characters in a message appear in an area of the screen, seemingly at random, until in the end the message is completely spelled out and readable. Another effect "moves" a word up one line at a time, so that it appears to "float" to the top of the screen. (Since the program can't tell what the word has overwritten, the word leaves a trail behind it.) I don't know yet whether visual effects will prove useful or even interesting.

4. I've been talking with Tim Anderson about using the New Parser in this game. It still needs a lot of development, and in the end it may prove to be slow in operating, but it promises to be very capable. Now here's the question: should the game itself make a big deal out of the New Parser? For example, the game could begin with the parser introducing itself to the player, asking the player to type a few sentences to "warm up" the parser, before getting on with the story itself. The parser could take on a personality, explaining that this is its first job, that it means well but it may not succeed. Perhaps it gets depressed and refuses to work at all. Perhaps the parser is in fact Marvin's new aural interface module, depressing him even further.

While Galley was thinking about the game's design, Michael Bywater was making his own design notes independently. The authorship isn't completely clear, but I believe this very rough outline for the plot was created by Bywater on July 23, 1987.

The game has a three-part structure. You begin at the base of the ramp leading from the Heart of Gold; that is, where HH1 ended. As you look forward to exploring this apparently acceptable planet with your faithful companion Marvin and your friends Trillian, Zaphod and Ford, you realize that you are alone. You turn round to find that Marvin has gone back inside the ship. The others are nowhere to be seen.

PUZZLE 1 is to persuade Marvin to come out of the Heart of Gold and help you. WARNING: this is a nasty trick puzzle because if you solve it, you lose the game (although there is a safety-net). Marvin is enticed from the ship by means of an hallucinogenic psychodysleptic. Somewhere close by the H of G you find a broken radio ("a miserable-looking piece of technology" is the clue). Elsewhere, you find a plant. The plant, once picked, begins to deliquesce. You have to put the plant in the radio, whereupon the radio bursts into cheerful music. The plant appears to have lifted its spirits, if such a thing was possible. Marvin will come out of the spaceship to get the plant and immediately go into a state of intoxicated euphoria, whereupon he will fuck off and not be seen again. (Not until the parking lot at Milliways, anyway). You score points for this solution, but it is the wrong one. In fact, the winning score is 470 out of 490; if you get ALL the points, you lose.

The right solution is to TELL Marvin about the plant but not to GIVE it to him; he will follow you fairly faithfully if you have the plant in the radio for the first few plays of the game. After a while the radio begins to drive you mad. Shortly after this, you will be given the opportunity to find the thing your Aunt gave you; you can use this to store the plant. If you don't, the inanities of the radio will drive you insane.

You need to keep Marvin with you until you have encountered Slartibartfast's vehicle, but first you have another problem. While you are fooling around on Magrathea, Zaphod gets fed up and disappears in the Heart of Gold to go somewhere really fashionable. If you still have the radio (i.e., haven't gone prematurely insane) you will hear an announcement to the effect that Milliways is the most fashionable place in the Universe.

Consulting the Guide about Milliways reveals the details. Now you have to get to the restaurant. You will need Marvin to get you there unless you are very, very lucky.

Once you get to Milliways, you have a problem: they won't let you in. You aren't fashionable enough. There are two solutions to this problem. If you fucked up puzzle 1, Marvin will turn up dressed as a parking attendant. You can borrow his clothes and go in as a high-fashion crypto-punk. If you didn't get Marvin smashed on the plant, you can send him in instead, because his pose of fashionable paranoia is regarded as chic by the doorman.

Once in the restaurant, you have to get a response for the dead pop singer to find out where his ship is. There then follows a parking puzzle with Marvin, and you gain access to the ship. The roles of the other characters (Zaphod et al) are not clear here. We may leave them somewhere.

The ship is intercepted and you solve a navigation puzzle to return to Magrathea, where you are snaffled by Slartibartfast. The next major section of the game now takes place, here on Magrathea, and consists of three nested layers of puzzles.

The first level of puzzles are standard adventure stuff, but the objective is to assemble a time machine. Slartibartfast has told you of the great history of Magrathea as the Universal Planetary Design workshop, and says he can show you how it all worked but the machine is broken. It is necessary for you to fix the machine, which you do by pottering about the geography of Magrathea. Next, once the machine is working, you have to get into it WITHOUT Slartibartfast but WITH Marvin. If you get in with S. alone, all he'll show you is Magrathea, boasting about his fjords etc. If however you get there with Marvin alone, Marvin will be so depressed that he'll try and depress you too, which he does by taking you back to Earth. However... the machine only works from the point of view of Slartibartfast, so that's in effect who you become. You observe a small flaw in one of the Norwegian fjords - a design change carried out without your (Slartibartfast's) authority which is responsible for all the awful things which have happened since.

Returning to reality, your job is to arrange for the rebuilding of Earth. You are allowed into Slartibartfast's workshop and you have to give him a huge philosophical puzzle to work on. This, like all the other puzzles in this penultimate part of the game, works on two elvels: it solves itself and allows you to proceed, but it also forms one element of the final endgame puzzle.

The endgame itself consists of a number of elements which are solved by assembling the consequences of solving eight other puzzles within the planetary workshop. To solve these puzzles you have to travel in space around the workshop visiting various locations which turn out to be planets, all of which are in the solar system and all of which are subtly wrong (Saturn has no rings, etc). To do this, you have a Solar System Bug-tester's report as part of the packaging. When you have solved these eight puzzles, you effectively (eight planets plus Slartibartfast's "Sun" puzzle) have all the solar system except Earth, and can then tie all the pieces together to make the adjustment to the fjords. (What we do about THAT, I don't know. There are an infinite number of possibilities).

The FINAL part of the endgame is to propel this new Solar system into its correct galactic location. This is a complex puzzle and to solve it you have to be on the Heart of Gold at the time the Solar System settles into position. At that point the Inifinite Improbability Drive detects that something infinitely improbable has happened and ejects you onto Earth, ready to begin the final part of the trilogy.

Michael Bywater took a final pass at the plot on August 7, 1987, but only gets a description of the first scene before it ends in mid-sentence:

These notes should be read in conjunction with the earlier synopsis in H2... no, they shouldn't. To hell with the earlier synopsis. What a cartload of pinwheel horse-shit. Screw the earlier synopsis altogether. Okay? Okay.

We begin on Magrathea, on the ramp outside the Heart of Gold, which is where H1 ended. You are Arthur Dent -- for the time being. The surface of the planet is cold and inhospitable. The others - Ford, Zaphod and Trillian - have wisely decided to stay aboard ship, in the sauna. You, unwisely but predictably, have forgotten your dressing gown. This poses a bit of a problem to you. Quite a lot of a problem, actually, since you are going to die. You are probably going to die anyway, before the cold gets to you. Irrational, I know, but that's the Galaxy for you: cold, aloof, impersonal and always ready to crap on an organism for no good reason.

The first puzzle is a fairly straightforward, linear, closed-boundary job: to get someone to open the hatch to the Heart of Gold before you die of cold. On Magrathea is a crater left by the impact of the whale. (You don't remember the whale? Then you'll be even more confused by the petunias...) On the edge of the crater is the remains of a intergalactic ghetto blaster left by a touring member of the Arqublustian Space Force, a group of deep-space thugs whose mission is to tour the galaxy playing loud music. In the crater itself, among the other whale bones, is a whale bone. Except it isn't: it's the Thing your aunt gave you which you don't know what it is... and inside it is a Barry Manilow CD. You play the CD and become deeply depressed and realize that only someone like Marvin could truly appreciate this stuff.

Trying to get back into the Heart of Gold, you find that everyone else, being in the sauna, cannot hear you, and Eddie is stupefied with gloom, being engaged in "conversation" with Marvin. Marvin's conversational powers are capable of breaking down any electronic device in the known universe. But Barry Manilow is even more potent. You potter about the ship until you can hear, through the hull, Marvin droning on, then you play the Barry Manilow CD. Marvin enjoys the "music" and frees Eddie to open the door. Marvin comes to the hatch bearing your gown, which contains the Guide and a key-card to the ship.

Now you can survive the cold long enough to explore the whale crater. You find not only the remains of the whale but the remains of the bowl of petunias. When planted in the Thing, the petunias take on a strange quality: if you are holding a...

And that's where the design notes end. Was this the first part of the final outline delivered by Michael Bywater in February 1988? Hard to say, but it was definitely the most coherent. I wish there was more.

Play It!

The project may have been shelved, but the Infocom Drive reveals two early Milliways builds, from May 1988 and another a year later in April 1989. Both are very similar, just simple stubs with sparse descriptions and a small handful of rooms. These appear to have both been written by Stu Galley, but it's hard to say.

You start the prototype game as Arthur Dent, with Marvin, Ford Prefect, Zaphod Beeblebrox, and Trillian standing nearby. It picks up where the last game ended, with our heroes standing "on the ramp leading from the starship Heart of Gold to the surface of the legendary lost planet of Magrathea, which isn't. Lost, that is. It isn't lost because if it were, you wouldn't be here, but you are, so it isn't."

Unfortunately, it doesn't get much further than that. Strangely, the first build seems a bit more polished, with more rooms and a better intro, in which Ford, Zaphod, and Trillian leave you alone with Marvin on the ramp. There are a few rooms with sparse descriptions ("There will be an elevator there someday, but not yet"), but not much to do.

Play Milliways (Release 15, Serial 880512). Java 5 required.

In the later build, you can go down the ramp onto the planet, poke the characters, or try to talk to them, but that's about it. Unlike the first build, you don't have the Hitchhiker's Guide in your inventory and the surface of Magrathea has a large crater next to the ship. Trying to go south-east to view it gives you an error, but in the source files, there's additional text which explains the crater was created by the sperm whale that materialized next to the Heart of Gold.

"The dusty ground rises here before falling away into a crater. The crater seems rather new, as if it had been created by the impact of something huge and confused, travelling downwards at high velocity. It is as if a sperm whale had inexplicably materialized several miles above the surface of Magrathea and immediately plunged downwards, reaching terminal velocity almost immediately, terminal incomprehension soon afterwards, and, finally, terminal impact just as it was wondering whether it was going to have a nice day. This impression is heightened by the shards of whalebone and meat you can see glistening here and there around the crater."

That's the longest description in the prototype, and a tiny glimpse at what might have been.

If you like, you can download the original Z-code files (release 15 and release 184), but you'll need an interpreter that supports Z-Machine version 4 and version 6 story files. I'd recommend Windows Frotz 2002 for Windows or Spatterlight for Mac.

Other Resources

For more information and gossip about Infocom's history, I highly recommend Down from the Top of Its Game: The Story of Infocom, Inc., The Lore and Legends of Infocom, Jimmy Maher's Let's Tell A Story Together: A History of Interactive Fiction.

Several of the bio photos come from Marco Thorek's excellent Infocom: The Master Storytellers.

Special thanks to Wei-ju Wu's Z-Machine Preservation Project, the Java interpreter I used above. It's the only web-based interpreter that would play both prototypes at all. Highly recommended.