RSS

Tag Archives: zork

Return to Zork

Where should we mark the beginning of the full-motion-video era, that most extended of blind alleys in the history of the American games industry? The day in the spring of 1990 that Ken Williams, founder and president of Sierra On-Line, wrote his latest editorial for his company’s seasonal newsletter might be as good a point as any. In his editorial, Williams coined the term “talkies” in reference to an upcoming generation of games which would have “real character voices and no text.” The term was, of course, a callback to the Hollywood of circa 1930, when sound began to come to the heretofore silent medium of film. Computer games, Williams said, stood on the verge of a leap that would be every bit as transformative, in terms not only of creativity but of profitability: “How big would the film industry be today if not for this step?”

According to Williams, the voice-acted, CD-based version of Sierra’s King’s Quest V was to become the games industry’s The Jazz Singer. But voice acting wasn’t the only form of acting which the games of the next few years had in store. A second transformative leap, comparable to that made by Hollywood when film went from black and white to color, was also waiting in the wings to burst onto the stage just a little bit later than the first talkies. Soon, game players would be able to watch real, human actors right there on their monitor screens.

As regular readers of this site probably know already, the games industry’s Hollywood obsession goes back a long way. In 1982, Sierra was already advertising their text adventure Time Zone with what looked like a classic “coming attractions” poster; in 1986, Cinemaware was founded with the explicit goal of making “interactive movies.” Still, the conventional wisdom inside the industry by the early 1990s had shifted subtly away from such earlier attempts to make games that merely played like movies. The idea was now that the two forms of media would truly become one — that games and movies would literally merge. “Sierra is part of the entertainment industry — not the computer industry,” wrote Williams in his editorial. “I always think of books, records, films, and then interactive films.” These categories defined a continuum of increasingly “hot,” increasingly immersive forms of media. The last listed there, the most immersive medium of all, was now on the cusp of realization. How many people would choose to watch a non-interactive film when they had the opportunity to steer the course of the plot for themselves? Probably about as many as still preferred books to movies.

Not all that long after Williams’s editorial, the era of the full-motion-video game began in earnest. The first really prominent exemplar of the species was ICOM Simulations’s Sherlock Holmes Consulting Detective series in 1992, which sent you wandering around Victorian London collecting clues to a mystery from the video snippets that played every time you visited a relevant location. The first volume of this series alone would eventually sell 1 million copies as an early CD-ROM showcase title. The following year brought Return to Zork, The 7th Guest, and Myst as three of the five biggest games of the year; all three of these used full-motion video to a greater or lesser extent. (Myst used it considerably less than the other two, and, perhaps not coincidentally, is the member of the trio that holds up by far the best today.) With success stories like those to look to, the floodgates truly opened in 1994. Suddenly every game-development project — by no means only adventure games — was looking for ways to shoehorn live actors into the proceedings.

But only a few of the full-motion-video games that followed would post anything like the numbers of the aforementioned four games. That hard fact, combined with a technological counter-revolution in the form of 3D graphics, would finally force a reckoning with the cognitive dissonance of trying to build a satisfying interactive experience by mixing and matching snippets of nonmalleable video. By 1997, the full-motion-video era was all but over. Today, few things date a game more instantly to a certain window of time than grainy video of terrible actors flickering over a background of computer-generated graphics. What on earth were people thinking?

Most full-motion-video games are indeed dire, but they’re going to be with us for quite some time to come as we continue to work our way through this history. I wish I could say that Activision’s Return to Zork, my real topic for today, was one of the exceptions to the rule of direness. Sadly, though, it isn’t.

In fact, let me be clear right now: Return to Zork is a terrible adventure game. Under no circumstances should you play it, unless to satisfy historical curiosity or as a source of ironic amusement in the grand tradition of Ed Wood. And even in these special cases, you should take care to play it with a walkthrough in hand. To do anything else is sheer masochism; you’re almost guaranteed to lock yourself out of victory within the first ten minutes, and almost guaranteed not to realize it until many hours later. There’s really no point in mincing words here: Return to Zork is one of the absolute worst adventure-game designs I’ve ever seen — and, believe me, I’ve seen quite a few bad ones.

Its one saving grace, however, is that it’s terrible in a somewhat different way from the majority of terrible full-motion-video adventure games. Most of them are utterly bereft of ideas beyond the questionable one at their core: that of somehow making a game out of static video snippets. You can almost see the wheels turning desperately in the designers’ heads as they’re suddenly confronted with the realization that, in addition to playing videos, they have to give the player something to actually do. Return to Zork, on the other hand, is chock full of ideas for improving upon the standard graphic-adventure interface in ways that, on the surface at any rate, allow more rather than less flexibility and interactivity. Likewise, even the trendy use of full-motion video, which dates it so indelibly to the mid-1990s, is much more calculated than the norm among its contemporaries.

Unfortunately, all of its ideas are undone by a complete disinterest in the fundamentals of game design on the part of the novelty-seeking technologists who created it. And so here we are, stuck with a terrible game in spite of it all. If I can’t quite call Return to Zork a noble failure — as we’ll see, one of its creators’ stated reasons for making it so callously unfair is anything but noble — I can at least convince myself to call it an interesting one.


When Activision decided to make their follow-up to the quickie cash-in Leather Goddesses of Phobos 2 a more earnest, better funded stab at a sequel to a beloved Infocom game, it seemed logical to find themselves a real Infocom Implementor to design the thing. They thus asked Steve Meretzky, whom they had just worked with on Leather Goddesses 2, if he’d like to design a new Zork game for them as well. But Meretzky hadn’t overly enjoyed trying to corral Activision’s opinionated in-house developers from a continent away last time around; this time, he turned them down flat.

Meretzky’s rejection left Activision without a lot of options to choose from when it came to former Imps. A number of them had left the games industry upon Infocom’s shuttering three years before, while, of those that remained, Marc Blank, Mike Berlyn, Brian Moriarty, and Bob Bates were all employed by one of Activison’s direct competitors. Activision therefore turned to Doug Barnett, a freelance artist and designer who had been active in the industry for the better part of a decade; his most high-profile design gig to date had been Cinemaware’s Lords of the Rising Sun. But he had never designed a traditional puzzle-oriented adventure game, as one can perhaps see all too well in the game that would result from his partnership with Activision. He also didn’t seem to have a great deal of natural affinity for Zork. In the lengthy set of notes and correspondence relating to the game’s development which has been put online by The Zork Library, a constant early theme on Activision’s part is the design’s lack of “Zorkiness.” “As it stands, the design constitutes more of a separate and unrelated story, rather than a sequel to the Zork series,” they wrote at one point. “It was noted that ‘Zork’ is the name of a vast ancient underground empire, yet Return to Zork takes place in a mostly above-ground environment.”

In fairness to Barnett, Zork had always been more of a state of mind than a coherent place. With the notable exception of Steve Meretzky, everyone at Infocom had been wary of overthinking a milieu that had originally been plucked out of the air more or less at random. In comparison to other shared worlds — even other early computer-game worlds, such as the Britannia of Richard Garriott’s Ultima series — there was surprisingly little there there when it came Zork: no well-established geography, no well-established history which everybody knew — and, most significantly of all, no really iconic characters which simply had to be included. At bottom, Zork boiled down to little more than a modest grab bag of tropes which lived largely in the eye of the beholder: the white house with a mailbox, grues, Flood Control Dam #3, Dimwit Flathead, the Great Underground Empire itself. And even most of these had their origin stories in the practical needs of an adventure game rather than any higher world-building purpose. (The Great Underground Empire, for example, was first conceived as an abandoned place not for any literary effect but because living characters are hard to implement in an adventure game, while the detritus they leave behind is relatively easy.)

That said, there was a distinct tone to Zork, which was easier to spot than it was to describe or to capture. Barnett’s design missed this tone, even as it began with the gleefully anachronistic, seemingly thoroughly Zorkian premise of casting the player as a sweepstakes winner on an all-expenses-paid trip to the idyllic Valley of the Sparrows, only to discover it has turned into the Valley of the Vultures under the influence of some pernicious, magical evil. Barnett and Activision would continue to labor mightily to make Return to Zork feel like Zork, but would never quite get there.

By the summer of 1992, Barnett’s design document had already gone through several revisions without entirely meeting Activision’s expectations. At this point, they hired one Eddie Dombrower to take personal charge of the project in the role of producer. Like Barnett, Dombrower had been working in the industry for quite some time, but had never worked on an adventure game; he was best known for World Series Major League Baseball on the old Intellivision console and Earl Weaver Baseball on computers. Dombrower gave the events of Return to Zork an explicit place in Zorkian history — some 700 years after Infocom’s Beyond Zork — and moved a big chunk of the game underground to remedy one of his boss’ most oft-repeated objections to the existing design.

More ominously, he also made a comprehensive effort to complicate Barnett’s puzzles, based on feedback from players and reviewers of Leather Goddesses 2, who were decidedly unimpressed with that game’s simple-almost-to-the-point-of-nonexistence puzzles. The result would be the mother of all over-corrections — a topic we’ll return to later.

Unlike Leather Goddess 2, whose multimedia ambitions had led it to fill a well-nigh absurd 17 floppy disks, Return to Zork had been planned almost from its inception as a product for CD-ROM, a technology which, after years of false promises and setbacks, finally seemed to be moving toward a critical mass of consumer uptake. In 1992, full-motion video, CD-ROM, and multimedia computing in general were all but inseparable concepts in the industry’s collective mind. Activision thus became one of the first studios to hire a director and actors and rent time on a sound stage; the business of making computer games had now come to involve making movies as well. They even hired a professional Hollywood screenwriter to punch up the dialog and make it more “cinematic.”

In general, though, while the computer-games industry was eager to pursue a merger with Hollywood, the latter was proving far more skeptical. There was still little money in computer games by comparison with movies, and there was very little prestige — rather the opposite, most would say — in “starring” in a game. The actors which games could manage to attract were therefore B-listers at best. Return to Zork actually collected a more accomplished — or at least more high-profile — cast than most. Among them were Ernie Lively, a veteran supporting player from television shows such as The Dukes of Hazzard; his daughter Robyn Lively, fresh off a six-episode stint as a minor character on David Lynch’s prestigious critics’ darling Twin Peaks; Jason Hervey, who was still playing older brother Wayne on the long-running coming-of-age sitcom The Wonder Years; and Sam Jones, whose big shot at leading-man status had come with the film Flash Gordon back in 1980 and gone with its mixed reception.

If the end result would prove less than Oscar-worthy, it’s for the most part not cringe-worthy either. After all, the cast did consist entirely of acting professionals, which is more than one can say for many productions of this ilk — and certainly more than one can say for the truly dreadful voice acting in Leather Goddess of Phobos 2, Activision’s previous attempt at a multimedia adventure game. While they were hampered by the sheer unfamiliarity of talking directly “to” the invisible player of the game — as Ernie Lively put it, “there’s no one to act off of” — they did a decent job with the slight material they had to work with.

The fact that they were talking to the player rather than acting out scenes with one another actually speaks to a degree of judiciousness in the use of full-motion video on Activision’s part. Rather than attempting to make an interactive movie in the most literal sense — by having a bunch of actors, one of them representing the protagonist, act out each of the player’s choices — Activision went for a more thoughtful mixed-media approach that could, theoretically anyway, eliminate most of the weaknesses of the typical full-motion-video adventure game. For the most part, only conversations involved the use of full-motion video; everything else was rendered by Activision’s pixel artists and 3D modelers in conventional computer graphics. The protagonist wasn’t shown at all: at a time when the third-person view that was the all but universal norm in adventure games, Activision opted for a first-person view.

The debate over whether an adventure-game protagonist ought to be a blank slate which the player can fill with her own personality or an established character which the player merely guides and empathizes with was a longstanding one even at the time when Return to Zork was being made. Certainly Infocom had held rousing internal debates on the subject, and had experimented fairly extensively with pre-established protagonists in some of their games. (These experiments sometimes led to rousing external debates among their fans, most notably in the case of the extensively characterized and tragically flawed protagonist of Infidel, who meets a nasty if richly deserved end no matter what the player does.) The Zork series, however, stemmed from an earlier, simpler time in adventure games than the rest of the Infocom catalog, and the “nameless, faceless adventurer,” functioning as a stand-in for the player herself, had always been its star. Thus Activision’s decision not to show the player’s character in Return to Zork, or indeed to characterize her in any way whatsoever, is a considered one, in keeping with everything that came before.

In fact, the protagonist of Return to Zork never actually says anything. To get around the need, Activision came up with a unique attitude-based conversation engine. As you “talk” to other characters, you choose from three stances — threatening, interested, or bored — and listen only to your interlocutors’ reactions. Not only does your own dialog go unvoiced, but you don’t even see the exact words you use; the game instead lets you imagine your own words. Specific questions you might wish to ask are cleverly turned into concrete physical interactions, something games do much better than abstract conversations. As you explore, you have a camera with which to take pictures of points of interest. During conversations, you can show the entries from your photo album to your interlocutor, perhaps prompting a reaction. You can do the same with objects in your inventory, locations on the auto-map you always carry with you, or even the tape recordings you automatically make of each interaction with each character.

So, whatever else you can say about it, Return to Zork is hardly bereft of ideas. William Volk, the technical leader of the project, was well up on the latest research into interface design being conducted inside universities like MIT and at companies like Apple. Many such studies had concluded that, in place of static onscreen menus and buttons, the interface should ideally pop into existence just where and when the user needed it. The result of such thinking in Return to Zork is a screen with no static interface at all; it instead pops up when you click on an object with which you can interact. Since it doesn’t need the onscreen menu of “verbs” typical of contemporaneous Sierra and LucasArts adventure games, Return to Zork can give over the entirety of the screen to its graphical portrayal of the world.

In addition to being a method of recapturing screen real estate, the interface was conceived as a way to recapture some of the sense of boundless freedom which is such a characteristic of parser-driven text adventures — a sense which can all too easily become lost amidst the more constrained interfaces of their graphical equivalent. William Volk liked to call Return to Zork‘s interface a “reverse parser”: clicking on a “noun” in the environment or in your inventory yields a pop-up menu of “verbs” that pertain to it. Taking an object in your “hand” and clicking it on another one yields still more options, the equivalent of commands to a parser involving indirect as well as direct objects. In the first screen of the game, for example, clicking the knife on a vulture gives options to “show knife to vulture,” “throw knife at vulture,” “stab vulture with knife,” or “hit vulture with knife.” There are limits to the sense of possibility: every action had to be anticipated and hand-coded by the development team, and most of them are the wrong approach to whatever you’re trying to accomplish. In fact, in the case of the example just mentioned as well as many others, most of the available options will get you killed; Return to Zork loves instant deaths even more than the average Sierra game. And there are many cases of that well-known adventure-game syndrome where a perfectly reasonable solution to a problem isn’t implemented, forcing you to devise some absurdly convoluted solution that is implemented in its stead. Still, in a world where adventure games were getting steadily less rather than more ambitious in their scope of interactive possibility — to a large extent due to the limitations of full-motion video — Return to Zork was a welcome departure from the norm, a graphic adventure that at least tried to recapture the sense of open-ended possibility of an Infocom game.

Indeed, there are enough good ideas in Return to Zork that one really, really wishes they all could have been tied to a better game. But sadly, I have to stop praising Return to Zork now and start condemning it.

The most obvious if perhaps most forgivable of its sins is that, as already noted, it never really manages to feel like Zork — not, at least, like the classic Zork of the original trilogy. (Steve Meretzky’s Zork Zero, Infocom’s final release to bear the name, actually does share some of the slapstick qualities of Return to Zork, but likewise rather misses the feel of the original.) The most effective homage comes at the very beginning, when the iconic opening text of Zork I appears onscreen and morphs into the new game’s splashy opening credits. It’s hard to imagine a better depiction circa 1993 of where computer gaming had been and where it was going — which was, of course, exactly the effect the designers intended.

Once the game proper gets under way, however, modernity begins to feel much less friendly to the Zorkian aesthetic of old. Most of Zork‘s limited selection of physical icons do show up here, from grues to Flood Control Dam #3, but none of it feels all that convincingly Zork-like. The dam is a particular disappointment; what was described in terms perfect for inspiring awed flights of the imagination in Zork I looks dull and underwhelming when portrayed in the cruder medium of graphics. Meanwhile the jokey, sitcom-style dialog that confronts you at every turn feels even less like the original trilogy’s slyer, subtler humor.

This isn’t to say that Return to Zork‘s humor doesn’t connect on occasion. It’s just… different from that of Dave Lebling and Marc Blank. By far the most memorable character, whose catchphrase has lived on to this day as a minor Internet meme, is the drunken miller named Boos Miller. (Again, subtlety isn’t this game’s trademark.) He plies you endlessly with whiskey, whilst repeating, “Want some rye? Course you do!” over and over and over in his cornpone accent. It’s completely stupid — but, I must admit, it’s also pretty darn funny; Boos Miller is the one thing everyone who ever played the game still seems to remember about Return to Zork. But, funny though he is, he would be unimaginable in any previous Zork.


Of course, a lack of sufficient Zorkiness need not have been the kiss of death for Return to Zork as an adventure game in the abstract. What really does it in is its thoroughly unfair puzzle design. This game plays like the fever dream of a person who hates and fears adventure games. It’s hard to know where to even start (or end) with this cornucopia of bad puzzles, but I’ll describe a few of them, ranked roughly in order of their objectionability.

The Questionable: At one point, you find yourself needing to milk a cow, but she won’t let you do so with cold hands. Do you need to do something sensible, like, say, find some gloves or wrap your hands in a blanket? Of course not! The solution is to light some of the hay that’s scattered all over the wooden barn on fire and warm your hands that way. For some reason, the whole place doesn’t go up in smoke. This solution is made still more difficult to discover by the way that the game usually kills you every time you look at it wrong. Why on earth would it not kill you for a monumentally stupid act like this one? To further complicate matters, for reasons that are obscure at best you can only light the hay on fire if you first pick it up and then drop it again. Thus even many players who are consciously attempting the correct solution will still get stuck here.

The Absurd: At another point, you find a bra. You have to throw it into an incinerator in order to get a wire out of it whose existence you were never aware of in the first place. How does the game expect you to guess that you should take such an action? Apparently some tenuous linkage with the 1960s tradition of bra burning and, as a justification after the fact, the verb “to hot-wire.” Needless to say, throwing anything else into the incinerator just destroys the object and, more likely than not, locks you out of victory.

The Incomprehensible: There’s a water wheel out back of Boos’s house with a chock holding it still. If you’ve taken the chock and thus the wheel is spinning, and you’ve solved another puzzle that involves drinking Boos under the table (see the video above), a trapdoor is revealed in the floor. But if the chock is in place, the trapdoor can’t be seen. Why? I have absolutely no idea.

Wait! Don’t do it!

The Brutal: In a way, everything you really need to know about Return to Zork can be summed up by its most infamous single puzzle. On the very first screen of the game, there’s a “bonding plant” growing. If you simply pull up the plant and take it with you, everything seems fine — until you get to the very end of the game many hours later. Here, you finally find a use for the plant you’ve been carting around all this time. Fair enough. But unfortunately, you need a living version of it. It turns out you were supposed to have used a knife to dig up the plant rather than pulling or cutting it. Guess what? You now get to play through the whole game again from the beginning.

All of the puzzles just described, and the many equally bad ones, are made still more complicated by the game’s general determination to be a right bastard to you every chance it gets. If, as Robb Sherwin once put it, the original Zork games hate their players, this game has found some existential realm beyond mere hatred. It will let you try to do many things to solve each puzzle, but, of those actions that don’t outright kill you, a fair percentage lock you out of victory in one way or another. Sometimes, as in the case of its most infamous puzzle, it lets you think you’ve solved them, only to pull the rug out from under you much later.

So, you’re perpetually on edge as you tiptoe through this minefield of instant deaths and unwinnable states; you’ll have a form of adventure-game post-traumatic-stress syndrome by the time you’re done, even if you’re largely playing from a walkthrough. The instant deaths are annoying, but nowhere near as bad as the unwinnable states; the problem there is that you never know whether you’ve already locked yourself out of victory, never know whether you can’t solve the puzzle in front of you because of something you did or didn’t do a long time ago.

It all combines to make Return to Zork one of the worst adventure games I’ve ever played. We’ve sunk to Time Zone levels of awful with this one. No human not willing to mount a methodical months-long assault on this game, trying every possibility everywhere, could possibly solve it unaided. Even the groundbreaking interface is made boring and annoying by the need to show everything to everyone and try every conversation stance on everyone, always with the lingering fear that the wrong stance could spoil your game. Adventure games are built on trust between player and designer, but you can’t trust Return to Zork any farther than you can throw it. Amidst all the hand-wringing at Activision over whether Return to Zork was or was not sufficiently Zorky, they forgot the most important single piece of the Infocom legacy: their thoroughgoing commitment to design, and the fundamental respect that commitment demonstrated to the players who spent their hard-earned money on Infocom games.  “Looking back at the classics might be a good idea for today’s game designers,” wrote Computer Gaming World‘s Scorpia at the conclusion of her mixed review of Return to Zork. “Good puzzle construction, logical development, and creative inspiration are in rich supply on those dusty disks.” None of these, alas, is in correspondingly good supply in Return to Zork.

The next logical question, then, is just how Return to Zork‘s puzzles wound up being so awful. After all, this game wasn’t the quickie cash grab that Leather Goddesses of Phobos 2 had been. The development team put serious thought and effort into the interface, and there were clearly a lot of people involved with this game who cared about it a great deal — among them Activision’s CEO Bobby Kotick, who was willing to invest almost $1 million to bring the whole project to fruition at a time when cash was desperately short and his creditors had him on a short leash indeed.

The answer to our question apparently comes down to the poor reception of Leather Goddesses 2, which had stung Activision badly. In an interview given shortly before Return to Zork‘s release, Eddie Dombrower said that, “based on feedback that the puzzles in Leather Goddesses of Phobos [2] were too simple,” the development team had “made the puzzles increasingly difficult just by reworking what Doug had already laid out for us.” That sounds innocent enough on the face of it. But, speaking to me recently, William Volk delivered a considerably darker variation on the same theme. “People hated Leather Goddesses of Phobos 2 — panned it,” he told me. “So, we decided to wreak revenge on the entire industry by making Return to Zork completely unfair. Everyone bitches about that title. There’s 4000 videos devoted to Return to Zork on YouTube, most of which are complaining because the title is so blatantly unfair. But, there you go. Something to pin my hat on. I made the most unfair game in history.”

For all that I appreciate Volk sharing his memories with me, I must confess that my initial reaction to this boast was shock, soon to be followed by genuine anger at the lack of empathy it demonstrates. Return to Zork didn’t “wreak revenge” on its industry, which really couldn’t have cared less. It rather wreaked “revenge,” if that’s the appropriate word, on the ordinary gamers who bought it in good faith at a substantial price, most of whom had neither bought nor commented on Leather Goddesses 2. I sincerely hope that Volk’s justification is merely a case of hyperbole after the fact. If not… well, I really don’t know what else to say about such juvenile pettiness, so symptomatic of the entitled tunnel vision of so many who are fortunate enough to work in technology, other than that it managed to leave me disliking Return to Zork even more. Some games are made out of an openhearted desire to bring people enjoyment. Others, like this one, are not.

I’d like to be able to say that Activision got their comeuppance for making Return to Zork such a bad game, demonstrating such contempt for their paying customers, and so soiling the storied Infocom name in the process. But exactly the opposite is the case. Released in late 1993, Return to Zork became one of the breakthrough titles that finally made the CD-ROM revolution a reality, whilst also carrying Activision a few more steps back from the abyss into which they’d been staring for the last few years. It reportedly sold 1 million copies in its first year — albeit the majority of them as a bundled title, included with CD-ROM drives and multimedia upgrade kits, rather than as a boxed standalone product. “Zork on a brick would sell 100,000 copies,” crowed Bobby Kotick in the aftermath.

Perhaps. But more likely not. Even within the established journals of computer gaming, whose readership probably didn’t constitute the majority of Return to Zork‘s purchasers, reviews of the game were driven more by enthusiasm for its graphics and sound, which really were impressive in their day, than by Zork nostalgia. Discussed in the euphoria following its release as the beginning of a full-blown Infocom revival, Return to Zork would instead go down in history as a vaguely embarrassing anticlimax to the real Infocom story. A sequel to Planetfall, planned as the next stage in the revival, would linger in Development Hell for years and ultimately never get finished. By the end of the 1990s, Zork as well would be a dead property in commercial terms.

Rather than having all that much to do with its Infocom heritage, Return to Zork‘s enormous commercial success came down to its catching the technological zeitgeist at just the right instant, joining Sherlock Holmes Consulting Detective, The 7th Guest, and Myst as the perfect flashy showpieces for CD-ROM. Its success conveyed all the wrong messages to game publishers like Activision: that multimedia glitz was everything, and that design really didn’t matter at all.

If it stings a bit that this of all games, arguably the worst one ever to bear the Infocom logo, should have sold better than any of the rest of them, we can comfort ourselves with the knowledge that Quality does have a way of winning out in the end. Today, Return to Zork is a musty relic of its time, remembered if at all only for that “want some rye?” guy. The classic Infocom text adventures, on the other hand, remain just that — widely recognized as timeless classics, their clean text-only presentations ironically much less dated than all of Return to Zork‘s oh-so-1993 multimedia flash. Justice does have a way of being served in the long run.

(Sources: the book Return to Zork Adventurer’s Guide by Steve Schwartz; Computer Gaming World of February 1993, July 1993, November 1993, and January 1994; Questbusters of December 1993; Sierra News Magazine of Spring 1990; Electronic Games of January 1994; New Media of June 24 1994. Online sources include The Zork Library‘s archive of Return to Zork design documents and correspondence, Retro Games Master‘s interview with Doug Barnett, and Matt Barton’s interview with William Volk. Some of this article is drawn from the full Get Lamp interview archives which Jason Scott so kindly shared with me. Finally, my huge thanks to William Volk for sharing his memories and impressions with me in a personal interview.)

 

Tags: , , ,

Zork Zero

Zork Zero

Zork Zero the idea was kicking around Infocom for quite a long time before Zork Zero the game was finally realized. Steve Meretzky first proposed making a prequel to the original Zork trilogy as far back as 1985, when he included it on a list of possible next games that he might write after finishing his personal passion project of A Mind Forever Voyaging. The Zork Zero he described at that time not only already had the name but the vast majority of the concept of the eventual finished game as well.

As the name implies, a prequel to the Zork trilogy. It would be set in the Great Underground Empire, and covering a long period of time, from the end of the reign of Dimwit Flathead in 789 through the fall of the GUE in 883, and possibly through 948 (the year of the Zork trilogy). It would almost certainly end “west of a white house.” There would be some story, probably about as much as Enchanter or Sorcerer. For the most part, though, it would be an intensely puzzle-oriented game with a huge geography.

The fact that Meretzky knew in what years Dimwit Flathead died, the Great Underground Empire fell, and Zork I began says much about his role as the unofficial keeper of Zorkian lore at Infocom. He had already filled a huge notebook with similarly nitpicky legends and lore. This endeavor was viewed by most of the other Imps, who thought of the likes of Dimwit Flathead as no more than spur-of-the-moment jokes, with bemused and gently mocking disinterest. Still, if Infocom was going to do a big, at least semi-earnest Zork game, his obsessiveness about the milieu made Meretzky the obvious candidate for the job.

But that big Zork game didn’t get made in 1985, partly because the other Imps remained very reluctant to sacrifice any real or perceived artistic credibility by trading on the old name and partly because the same list of possible next projects included a little something called Leather Goddesses of Phobos that everyone, from the Imps to the marketers to the businesspeople, absolutely loved. Brian Moriarty’s reaction was typical: “If you don’t do this, I will. But not as well as you could.”

After Meretzky completed Leather Goddesses the following year, Zork Zero turned up again on his next list of possible next projects. This time it was granted more serious consideration; Infocom’s clear and pressing need for hits by that point had done much to diminish the Imps’ artistic fickleness. At the same time, though, Brian Moriarty also was shopping a pretty good proposal for a Zork game, one that would include elements of the CRPGs that seemed to be replacing adventure games in some players’ hearts. Meanwhile Meretzky’s own list included something called Stationfall, the long-awaited sequel to one of the most beloved games in Infocom’s back catalog. While Moriarty seemed perfectly capable of pulling off a perfectly acceptable Zork, the universe of Planetfall, and particularly the lovable little robot Floyd, were obviously Meretzky’s babies and Meretzky’s alone. Given Infocom’s commercial plight, management’s choice between reviving two classic titles or just one was really no choice at all. Meretzky did Stationfall, and Moriarty did Beyond Zork — with, it should be noted, the invaluable assistance of Meretzky’s oft-mocked book of Zorkian lore.

And then it was 1987, Stationfall too was finished, and there was Zork Zero on yet another list of possible next projects. I’ll be honest in stating that plenty of the other project possibilities found on the 1987 list, some of which had been appearing on these lists as long as Zork Zero, sound much more interesting to this writer. There was, for instance, Superhero League of America, an idea for a comedic superhero game with “possible RPG elements” that would years later be dusted off by Meretzky to become the delightful Legend Entertainment release Superhero League of Hoboken. There was a serious historical epic taking place on the Titanic that begs to be described as Meretzky’s Trinity. And there was something with the working title of The Best of Stevo, a collection of interactive vignettes in the form if not the style of Nord and Bert Couldn’t Make Head or Tail of It.

Mind you, not all of the other projects were winners. A heavy-handed satire to be called The Interactive Bible, described by Meretzky as “part of my ongoing attempt to offend every person in the universe,” was eloquently and justifiably lacerated by Moriarty.

As you noted, this game is likely to offend many people, and not just frothing nutcakes either. A surprising number of reasonable people regard the Book with reverence. They are likely to regard your send-up as superficial and juvenile. They will wonder what qualifies you to poke fun at their (or anybody’s) faith. Why do you want to write this? Do you really think it will sell?

If Zork Zero wasn’t at the bottom of anyone’s list like The Interactive Bible, no one was exactly burning with passion to make it either. Few found the idea of going back to the well of Zork yet again all that interesting in creative terms, especially as Beyond Zork was itself still very much an ongoing project some weeks from release. The idea’s trump card, however, was the unique commercial appeal most still believed the Zork trademark to possess. Jon Palace’s faint praise was typical: “I’m sure this would sell very well. It’s certainly ‘safe.'” By 1987, the commercially safe route was increasingly being seen as the only viable route within Infocom, at least until they could manage to scare up a few hits. A final tally revealed that Zork Zero had scored an average of 7.2 among “next Meretzky project” voters on a scale of 1 to 10, edging out Superhero League of America by one tenth of a point, Titanic by two tenths, and The Best of Stevo by one full point; the last was very well-liked in the abstract, but its standing was damaged by the fact that, unusually for Meretzky, the exact form the vignettes would take wasn’t very well specified.

On August 7, 1987, it was decided provisionally to have Meretzky do Zork Zero next. In a demonstration of how tepid everyone’s enthusiasm remained for such a safe, unchallenging game, an addendum was included with the announcement: “I think it is fair to add that if Steve happens to have a flash of creativity in the next few days and thinks of some more ideas for his experimental story project (Best of Stevo), nearly everyone in this group would prefer that he do that product.” That flash apparently didn’t come; The Best of Stevo was never heard of again. Also forgotten in the rush to do Zork Zero was the idea, mooted in Beyond Zork, of Zork becoming a series of CRPG/text-adventure hybrids, with the player able to import the same character into each successive game. Zork Zero would instead be a simple standalone text adventure again.

While it’s doubtful whether many at Infocom ever warmed all that much to Zork Zero as a creative exercise, the cavalcade of commercial disappointments that was 1987 tempted many to see it as the latest and greatest of their Great White Hopes for a return to the bestseller charts. It was thus decided that it should become the first game to use Infocom’s new version 6 Z-Machine, usually called “YZIP” internally. Running on Macintosh II microcomputers rather than the faithful old DEC, the YZIP system would at last support proper bitmap illustrations and other graphics, along with support for mice, sound and music, far more flexible screen layouts, and yet bigger stories over even what the EZIP system (known publicly as Interactive Fiction Plus) had offered. With YZIP still in the early stages of development, Meretzky would first write Zork Zero the old way, on the DEC. Then, when YZIP was ready, the source code could be moved over and the new graphical bells and whistles added; the new version of ZIL was designed to be source-compatible with the old. In the meantime, Stu Galley was working on a ground-up rewrite of the parser, which was itself written in ZIL. At some magic moment, the three pieces would all come together, and just like that Infocom would be reborn with pictures and a friendlier parser and lots of other goodies, all attached to the legendary Zork name and written by Infocom’s most popular and recognizable author. That, anyway, was the theory.

Being at the confluence of so much that was new and different, Zork Zero became one of the more tortured projects in Infocom’s history, almost up there with the legendarily tortured Bureaucracy project. None of the problems, however, were down to Meretzky. Working quickly and efficiently as always, his progress on the core of the game proper far outstripped the technology enabling most of the ancillary bells and whistles. While Stu Galley’s new parser went in on November 1, 1987, it wasn’t until the following May 10 that a YZIP Zork Zero was compiled for the first time.

In sourcing graphics for Zork Zero, Infocom was on completely foreign territory. Following the lead of much of the computer-game industry, all of the graphics were to be created on Amigas, whose Deluxe Paint application was so much better than anything available on any other platform that plenty of artists simply refused to use anything else. Jon Palace found Jim Shook, the artist who would do most of the illustrations for Zork Zero, at a local Amiga users-group meeting. Reading some of the memos and meeting notes from this period, it’s hard to avoid the impression that — being painfully blunt here — nobody at Infocom entirely knew what they were doing when it came to graphics. As of February of 1988, they still hadn’t even figured out what resolution Shook should be working in. “We still don’t know whether images should be drawn in low-res, medium-res, interlace, or high-res mode on the Amiga in Deluxe Paint,” wrote Palace plaintively in one memo. “Joel claims Tim should know. Tim, do you know?”

Infocom wound up turning to Magnetic Scrolls, who had been putting pictures into their own text adventures for quite some time, for information on “graphics compression techniques,” a move that couldn’t have sat very well with such a proud group of programmers. The graphics would continue to be a constant time sink and headache for many months to come. Steve Meretzky told me that he remembers the development of Zork Zero primarily as “heinous endless futzing with the graphics, mostly on an Amiga, to make them work with all the different screen resolutions, number of colors, pixel aspect ratios, etc. In my memory, it feels like I spent way more time doing that than actually designing puzzles or writing ZIL code.”

Zork Zero uses graphics more often to present the look of an illuminated manuscript than for traditional illustrations.

Zork Zero uses graphics more often to present the look of an illuminated manuscript than for traditional illustrations.

And yet in comparison to games like those of Magnetic Scrolls, the finished Zork Zero really wouldn’t have a lot of graphics. Instead of an illustration for each room, the graphics take the form of decorative borders, an illuminated onscreen map, some graphical puzzles (solvable using a mouse), and only a few illustrations for illustrations’ sake. Infocom would advertise that they wanted to use graphics in “a new way” for Zork Zero — read, more thoughtfully, giving them some actual purpose rather than just using them for atmosphere. All of which is fair enough, but one suspects that money was a factor as well; memos from the period show Infocom nickel-and-diming the whole process, fretting over artist fees of a handful of thousand dollars that a healthier developer wouldn’t have thought twice about.

The financial squeeze also spelled the end of Infocom’s hopes for a full soundtrack, to have been composed by Russell Lieblich at Mediagenic, who had earlier done the sound effects for The Lurking Horror and Sherlock: The Riddle of the Crown Jewels. But the music never happened; when Zork Zero finally shipped, it would be entirely silent apart from a warning beep here or an acknowledging bloop there.

Hemorrhaging personnel as they were by this point, Infocom found themselves in a mad scramble to get all the pieces that did wind up making it into Zork Zero together in time for Christmas 1988, months after they had originally hoped to ship the game. Bruce Davis grew ever more frustrated and irate at the delays; a contemporary memo calls him a “looming personality” and notes how he is forever “threatening a tantrum.” A desperate-sounding “Proclamation” went out to the rank-and-file around the same time: “The one who can fix the bugs of Zork Zero, and save the schedule from destruction, shall be rewarded with half the wealth of the Empire.” Signed: “Wurb Flathead, King of Quendor.”

Like a number of Zork Zero's illustrations, this one actually conveys some important information about the state of the game.

Like a number of Zork Zero‘s illustrations, this one actually conveys some important information about the state of the game rather than being only for show.

Time constraints, the fact that the beta builds ran only on the Macintosh, and Infocom’s determination to test Zork Zero primarily using new testers unfamiliar with interactive fiction meant that it didn’t receive anywhere near the quantity or quality of outside feedback that had long been customary for their games. Many of the new testers seemed bemused if not confused by the experience, and few came anywhere close to finishing the game. I fancy that one can feel the relative lack of external feedback in the end result, as one can the loss of key voices from within Infocom like longtime producer Jon Palace and senior tester Liz Cyr-Jones.

Despite the corner-cutting, Infocom largely missed even the revised target of Christmas 1988. Only the Macintosh version shipped in time for the holiday buying season, the huge job of porting the complicated new YZIP interpreter to other platforms having barely begun by that time. Zork Zero was quite well-received by the Macintosh magazines, but that platform was far from the commercial sweet spot in gaming.

The decorative borders change as you enter difference regions -- a nice touch.

A nice touch: the decorative borders change as you enter different regions.

A sort of cognitive dissonance was a thoroughgoing theme of the Zork Zero project from beginning to end. It’s right there in marketing’s core pitch: “Zork Zero is the beginning of something old (the Zork trilogy) and something new (new format with graphics).” Unable to decide whether commercial success lay in looking forward or looking back, Infocom tried to have it both ways. Zork Zero‘s “target audience,” declared marketing, would be “primarily those who are not Infocom fans; either they have never tried interactive fiction or they have lost interest in Infocom.” The game would appeal to them thanks to “a mouse interface (enabling the player to move via compass rose), onscreen hints, a new parser (to help novices), and pretty pictures that will knock your socks off!”

Yet all the gilding around the edges couldn’t obscure the fact that Zork Zero was at heart the most old-school game Infocom had made since… well, since Zork I really. That, anyway, was the last game they had made that was so blatantly a treasure hunt and nothing more. Zork Zero‘s dynamic dozen-turn introduction lays out the reasons behind the static treasure hunt that will absorb the next several thousand turns. To thwart a 94-year-old curse that threatens to bring ruin to the Great Underground Empire, you must assemble 24 heirlooms that once belonged to 12 members of the Flathead dynasty and drop them in a cauldron. Zork Zero is, it must be emphasized, a big game, far bigger than any other that Infocom ever released, its sprawling geography of more than 200 rooms — more than 2200 if you count a certain building of 400 (nearly) identical floors —  housing scores of individual puzzles. The obvious point of comparison is not so much Infocom’s Zork trilogy as the original original Zork, the one put together by a bunch of hackers at MIT in response to the original Adventure back in the late 1970s, long before Infocom was so much as a gleam in anyone’s eye.

A Tower of Hanoi puzzle, one of the hoariest of Zork Zero's tired old chestnuts.

A Tower of Hanoi puzzle, one of the hoariest of Zork Zero‘s hoary old chestnuts.

The question — the answer to which must always to some extent be idiosyncratic to each player — is whether Zork Zero works for you on those terms. In my case, it doesn’t. The PDP-10 Zork is confusing and obscure and often deeply unfair, but it carries with it a certain joyous sense of possibility, of the discovery of a whole new creative medium, that we can enjoy vicariously with its creators. Zork Zero perhaps also echos the emotional circumstances of its creation: it just feels tired, and often cranky and mean-spirited to boot. Having agreed to make a huge game full of lots of puzzles, Meretzky dutifully provides, but the old magic is conspicuously absent.

Infocom always kept a library of puzzly resources around the office to inspire the Imps: books of paradoxes and mathematical conundrums, back issues of Games magazine, physical toys and puzzles of all descriptions. But for the first time with Zork Zero, Meretzky seems not so much inspired by these resources as simply cribbing from them. Lots of the puzzles in Zork Zero are slavish re-creations of the classics: riddles, a Tower of Hanoi puzzle, a peg game. Even the old chestnut about the river, the fox, the chicken, and the sack of grain makes an appearance. And even some of the better bits, like a pair of objects that let you teleport from the location of one to that of another, are derivative of older, better Infocom games like Starcross and Spellbreaker. One other, more hidden influence on Zork Zero‘s everything-but-the-kitchen-sink approach to puzzle design — particularly on the occasional graphical puzzles — is likely Cliff Johnson’s puzzling classic The Fool’s Errand, which Meretzky was playing with some dedication at the very time he was designing his own latest game. The Fool’s Errand‘s puzzles, however, are both more compelling and more original than Zork Zero‘s. Meretzky’s later Hodj ‘n’ Podj would prove a far more worthy tribute.

Zork Zero is a difficult game, and too often difficult in ways that really aren’t that much fun. I’m a fan of big, complicated puzzlefests in the abstract, but Zork Zero‘s approach to the form doesn’t thrill me. After the brief introductory sequence, the game exposes almost the whole of its immense geography to you almost immediately; there’s nothing for it but to start wandering and trying to solve puzzles. The combinatorial explosion is enormous. And even when you begin to solve some of the puzzles, the process can be made weirdly unsatisfying by the treasure-hunt structure. Too much of the time, making what at first feels like a significant step forward only yields another object to throw into the cauldron for some more points. You know intellectually that you’re making progress, but it doesn’t really feel like it.

I much prefer the approach of later huge puzzlefests like Curses! and The Mulldoon Legacy, which start you in a constrained space and gradually expand in scope as you solve puzzles. By limiting their initial scope, these games ease you into their worlds and limit the sense of hopeless aimlessness that Zork Zero inspires, while a new set of rooms to explore provides a far more tangible and satisfying reward for solving a puzzle sequence than does another object chunked in the cauldron and another few points. The later games feel holistically designed, Zork Zero like something that was just added to until the author ran out of space. Even The Fool’s Errand restricts you to a handful of puzzles at the beginning, unfolding its mysteries and its grand interconnections only gradually as you burrow ever deeper. That Infocom of all people — Steve Meretzky of all people, whose Leather Goddesses of Phobos and Stationfall are some of the most airtight designs in Infocom’s catalog — is suddenly embracing the design aesthetic of the 1970s is downright weird for a game that was supposed to herald a bright new future of more playable and player-friendly interactive fiction.

The in-game Encyclopedia Frobozzica is a nice if somewhat underused feature. The encyclopedia could have provided nudges for some more of the more obscure puzzles and maybe even some direction as to what to be working on next. Instead that work is all shuffled off to the hint system.

The in-game Encyclopedia Frobozzica is a nice but rather underused feature. The encyclopedia could have provided more nudges for some more of the more obscure puzzles and maybe even some direction as to what to be working on next. Instead that work is all shuffled off to the hint menu, the use of which feels like giving up or even cheating.

The puzzles rely on the feelies more extensively than any other Infocom game, often requiring you to make connections with seemingly tossed-off anecdotes buried deep within “The Flathead Calendar.” I generally don’t mind this sort of thing overmuch, but, like so much else in Zork Zero, it feels overdone here. These puzzles feel like they have far more to do with copy protection than the player’s enjoyment — but then much of the time Zork Zero seems very little concerned with the player’s enjoyment.

I love the headline of the single review of Zork Zero that’s to be found as of this writing on The Interactive Fiction Database: “Enough is enough!” That’s my own feeling when trying to get through this exhausting slog of a game. As if the sheer scope and aimlessness of the thing don’t frustrate enough, Meretzky actively goes out of his way to annoy you. There is, for instance, a magic wand with barely enough charges in it; waste a few charges in experimentation, and, boom, you’re locked out of victory. There’s that aforementioned building of 400 floors, all but one of them empty, which the diligent player will nevertheless feel the need to explore floor by floor, just in case there’s something else there; this is, after all, just the type of game to hide something essential on,say, floor 383. And then there’s the most annoying character in an Infocom game this side of Zork I‘s thief, a jester who teleports in every few dozen turns to do some random thing to you, like stick a clown nose over your own (you have to take it off within a certain number of turns or you’ll suffocate) or turn you into an alligator (you have to waste a few turns getting yourself turned back, then deal with picking up all of your possessions off the ground, putting those things you were wearing back on, etc.). Some of these gags are amusing the first time they happen, but they wear out their welcome quickly when they just keep wasting your time over a game that will already require thousands of moves to finish. The jester’s worst trick of all is to teleport you somewhere else in the game’s sprawling geography; you can be hopelessly trapped, locked out of victory through absolutely no fault of your own, if you’re unlucky and don’t have the right transportation handy. Hilariously, Infocom’s marketing people, looking always for an angle, hit upon selling the jester as Meretzky’s latest lovable sidekick, “every bit as enjoyable and memorable as Floyd of Planetfall fame.” Meretzky himself walked them back from that idea.

Some of the puzzles, probably even most of them, are fine enough in themselves, but there is a sprinkling of questionable ones, and all are made immeasurably more difficult by the fact that trying out a burst of inspiration can absorb 50 moves simply transiting from one side of the world to the other. Throw in a sharply limited inventory, which means you might need to make three or four round trips just to try out all the possible solutions you can think of, and things get even more fun. Graham Nelson among others has made much of the idea that the 128 K limitation of the original Z-Machine was actually a hidden benefit, forcing authors to hone their creations down to only what needed to be there and nothing that didn’t. I’ve generally been a little skeptical of that position; there are any number of good Infocom games that feel like they might have been still a little better with just a little more room to breathe. Zork Zero, however, makes as compelling a case as one can imagine for the idea that less is often more in interactive fiction, that constraints can lead to better designs.

The in-game mapping is handy from time to time, but, split into many different regions and viewable only by typing “MAP” from the main screen as it is, is not really ideal. A serious player is likely to be back to pencil and paper (or, these days, Trizbort) pretty quickly.

Which is actually not to say that Meretzky was operating totally unfettered by space constraints. While the YZIP format theoretically allowed a story size of up to 512 K not including graphics, the limitations of Infocom’s least-common-denominator platform, the Apple II, meant that the practical limit was around 340 K, a fairly modest expansion on the old 256 K EZIP and XZIP formats used for the Interactive Fiction Plus line. But still more restrictive was the limitation on the size of what Infocom called the “pre-load,” that part of the story data that could change as the player played, and that thus needed to always be in the host machine’s memory. The pre-load had to be held under about 55 K. Undoubtedly due in part to these restrictions, Zork Zero clearly sacrifices depth for breadth in comparison to many Infocom games that preceded it. The “examine” command suffers badly, some of the responses coming off like oxymorons: “totally ordinary looking writhing mass of snakes”; “totally ordinary looking herd of unicorns.” The sketchy implementation only adds to the throwback feel of the game as a whole.

The hints are certainly nice to have given the complexity and scope of the game, but they unfortunately aren’t context-sensitive. It’s all too easy to accidentally read the wrong one when trying to sort through this jumble.

Another subtle hidden enemy of Zork Zero as a design is the online hint system. Installed with the best of intentions in this as well as a few earlier Infocom games, it could easily lead to creeping laziness on the part of a game’s Implementor. “If the player really gets stuck, she can always turn to the hints,” ran the logic — thus no need to fret to quite the same extent over issues of solubility. The problem with that logic is that no one likes to turn to hints, whether found in the game itself, in a separate InvisiClues booklet, or in an online walkthrough. People play games like Zork Zero to solve them themselves, and the presence of a single bad puzzle remains ruinous to their experience as a whole even if they can look up the answer in the game itself. Infocom’s claim that “the onscreen hints help you through the rough spots without spoiling the story” doesn’t hold much water when one considers that Zork Zero doesn’t really have any story to speak of.

More puzzling is the impact — or rather lack thereof — of Stu Galley’s much-vaunted new parser. Despite being a ground-up rewrite using “an ATN algorithm with an LALR grammar and one-token look-ahead,” whatever that means, it doesn’t feel qualitatively different from those found in earlier Infocom games. The only obvious addition is the alleged ability to notice when you’re having trouble getting your commands across, and to start offering sample commands and other suggestions. A nice idea in theory, but the parser mostly seems to decide to become helpful and start pestering you with questions when you’re typing random possible answers to one of the game’s inane riddles. Like your racist uncle who decides to help you clean up after regaling you with his anecdotes over the Thanksgiving dinner table, even when Zork Zero tries to be helpful it’s annoying. Nowhere is the cognitive dissonance of Zork Zero more plainly highlighted than in the juxtaposition of this overly helpful, newbie-friendly parser with the old-school player hostility of the actual game design. “Zork hates its player,” wrote Robb Sherwin once of the game that made Infocom. After spending years evolving interactive fiction into something more positive and interesting than that old-school player hostility, Infocom incomprehensibly decided to circle back to how it all began with Zork Zero.

The most rewarding moment comes right at the end — and no, not because you’re finally done with the thing, although that’s certainly a factor too. In the end, you wind up right where it all began for Zork and for Infocom, before the famous white house, about to assume the role of the Dungeon Master, the antagonist of the original trilogy. There’s a melancholy resonance to the ending given the history not just of the Great Underground Empire but of Infocom in our own world. Released on July 14, 1989, the MS-DOS version of Zork Zero — the version that most of its few buyers would opt for — was one of the last two Infocom games to ship. So, the very end for Infocom circles back to the very beginning in many ways. Whether getting there is worth the trouble is of course another question.

As the belated date of the MS-DOS release will attest, versions of Zork Zero for the more important game-playing platforms were very slow in coming. The Amiga version didn’t ship until March of 1989, the Apple II version in June, followed finally by that MS-DOS version — the most important of all, oddly left for last. By that time Bruce Davis had lost patience, and Infocom had ceased to exist as anything other than a Mediagenic brand. The story of Zork Zero‘s failure to save Infocom thus isn’t so much the story of its commercial failure — although, make no mistake, it was a commercial failure — as the story of Infocom’s failure to just get the thing finished in time to even give it a chance of making a difference. Already an orphaned afterthought by the time it appeared on the platform that mattered most, Zork Zero likely never managed to sell even 10,000 copies in total. So much for Infocom’s “new look, new challenge, new beginning.”

We have a few more such afterthoughts to discuss before we pull the curtain at last on the story of Infocom, that most detailed and extended of all the stories I’ve told so far on this blog. Now, however, it’s time to check in with Infocom’s counterparts on the other side of the Atlantic, with the other two of the three remaining companies in the English-speaking world still trying to make a living out of text adventures in 1988. As you have probably guessed, things weren’t working out all that much better for either of them than they were for Infocom. Yet amidst the same old commercial problems, there are still some interesting and worthy games to discuss. So, we’ll start to do just that next time.

(Sources: As usual with my Infocom articles, much of this one is drawn from the full Get Lamp interview archives which Jason Scott so kindly shared with me. Much of it is also drawn from Jason’s “Infocom Cabinet” of vintage documents. Magazine sources include Questbusters of March 1989, The Games Machine of October 1989, and the Spring 1989 issue of Infocom’s The Status Line newsletter. Huge thanks also to Tim Anderson and Steve Meretzky for corresponding with me about some of the details of this period.

If you still want to play Zork Zero after the thrashing I’ve just given it — sorry, Steve and all Zork Zero fans! — you can purchase it from GOG.com as part of The Zork Anthology.)

 
 

Tags: , ,

Beyond Zork

Beyond Zork

For a company that’s long since gone down into history as the foremost proponent of the all-text adventure game, Infocom sure spent a lot of years fretting over graphics. As early as 1982, well before starting their iconic text-only advertising campaign, they entered into discussions with Mark Pelczarski of Penguin Software about a possible partnership that would have seen Antonio Antiochia, writer and illustrator of Penguin’s hit adventure Transylvania, drawing pictures for Infocom games using Penguin’s The Graphics Magician. When that combination of Penguin’s graphics technology with Infocom’s text-only Z-Machine was judged impractical, the all-text advertising campaign went forward, but still Infocom refused to rule out graphics internally. On the contrary, they used some of the revenue from 1983, their first really big year, to start a graphics research group of their own under the stewardship of Mike Berlyn. But that attempt at a cross-platform graphics system, a sort of Z-Machine for graphical games, petered out almost as quietly as the Penguin deal, resulting only in Berlyn’s unique but ultimately underwhelming computerized board game Fooblitzky. Specifying a single set of graphics capabilities achievable by all of the diverse computers on the market meant that those graphics had to be very primitive, and, thanks to the fact that they were running through an interpreter, slow to boot. In the end Fooblitzky made it only to the Apple II, the Atari 8-bits, and the PC clones, and even that was a struggle. Certainly trying to combine this already problematic system with the sort of adventure game that was Infocom’s bread and butter seemed hopeless. Thus the graphics group was quietly dispersed amid all the other downsizing of 1985. Strike two.

After completing Trinity in 1986, Brian Moriarty decided to see if he could make the third time the charm. The fundamental problem which had dogged Infocom’s efforts to date had been the big DEC PDP-10 that remained at the heart of their development system — the same big mainframe that, once lauded as the key to Infocom’s success, was now beginning to seem more and more like a millstone around their necks. Bitmap graphics on the DEC, while possible through the likes of a VT-125 terminal, were slow, awkward, and very limited, as Fooblitzky had demonstrated all too well. Worse, mixing conventional scrolling text, of the sort needed by an Infocom adventure game, with those graphics on the same screen was all but impossible. Moriarty therefore decided to approach the question from the other side. What graphic or graphic-like things could they accomplish on the PDP-10 without losing the ability to easily display text as well?

That turned out to be, if far from the state of the art, nevertheless more than one might expect, and also much more than was possible at the time that they’d first installed their PDP-10. DEC’s very popular new VT-220 line of text-oriented terminals couldn’t display bitmap graphics, but they could change the color of the screen background and individual characters at will, selecting from a modest palette of a dozen or so possibilities. Even better, they could download up to 96 graphics primitives into an alternate character set, allowing the drawing of simple lines, boxes, and frames, in color if one wished. By duplicating these primitives in the microcomputer interpreters, Infocom could duplicate what they saw on their DEC-connected dumb terminals on the computer monitors of their customers. Like much of the game that would gradually evolve from Moriarty’s thought experiment, this approach marked as much a glance backward as a step forward. Character graphics had been a feature of microcomputers from the beginning — in fact, they had been the only way to get graphics out of two of the Trinity of 1977, the Radio Shack TRS-80 and the Commodore PET — but had long since become passé on the micros in light of ever-improving bitmap-graphic capabilities. Once, at the height of their success and the arrogance it engendered, Infocom had declared publicly that they wouldn’t do graphics until they were confident that they could do them better than anyone else. But maybe such thinking was misguided. Given the commercial pressure they were now under, maybe primitive graphics were better than no graphics at all.

Thus color and character graphics became the centerpieces of a new version of the Z-Machine that Dave Lebling, Chris Reeve, and Tim Anderson, Infocom’s chief technical architects, began putting together for Moriarty’s “experimental” project. This version 5 of the Z-Machine, the last to be designed for Infocom’s PDP-10-based development system, gradually came to also sport a host of other new features, including limited mouse support, real-time support, and the ability, previously hacked rather rudely into the old version 3 Z-Machine for The Lurking Horror, to play sampled sound files. The most welcome feature of all was one of the least flashy: an undo command that could take back your last turn, even after dying. Game size was still capped at the 256 K of the version 4 Z-Machine, a concession to two 8-bitters that still made up a big chunk of Infocom’s sales, the Commodore 128 and the Apple II. For the first time, however, this version of the Z-Machine was designed to query the hardware on which it ran about its capabilities, degrading as gracefully as possible on platforms that couldn’t manage to provide its more advanced features. The graphically primitive Apple II, for instance, didn’t offer color and replaced the unique character-graphic glyphs with rough approximations in simple ASCII text, while both 8-bitters lacked the undo feature. Mouse input and sound were similarly only made available on machines that were up to it. Infocom took to calling the version 5 games, of which Moriarty’s nascent project would be the first, “XZIPs,” the “X” standing for “experimental.” (Even after Moriarty’s game was released and the format was obviously no longer so experimental, the name would stick.)

Moriarty's new interface running on an Amiga. Note the non-scrolling status window at top left that currently displays the room description, and the auto-map at top right. You can move around by clicking on the map.

Moriarty’s new interface running on an Amiga. Note the non-scrolling status window at top left that currently displays the room description, and the auto-map at top right. You can move around by clicking on the map as well as by typing the usual compass directions.

The interface gracefully (?) degraded on the Apple II.

The same interface gracefully (?) degraded on the Apple II.

Of course, it was still up to Moriarty to decide how to use the new toolkit. He asked himself, “What can I do within the constraints of our technology to make the adventuring experience a little easier?” He considered trying to do away with the parser entirely, but decided that that still wasn’t practical. Instead he designed an interface for what he liked to call “an illuminated text adventure.” Once again, in looking forward he found himself to a surprising extent looking back.

In watching myself play, I found the command I typed most was “look.” So I said this is silly, why can’t the room description always be visible? After all, that’s what Scott Adams did in his original twelve adventures, with a split-screen showing the room description, exits, and your inventory at the top, while you type in the bottom. So I said, let’s take a giant step backward. The screen is split in half in most versions. On the left side of the top half is a programmable window. It can contain either the room description or your inventory. So as you walk from room to room, instead of the description coming in-line with your commands, as it does now in our games, this window is updated. If you say, “inventory,” the window changes to show your possessions.

Another thing I liked about the old Scott Adams games was the list of room exits at the top of the screen. That’s a part of writing room descriptions that has always bugged me: we have to have at least one sentence telling where the exits are. That takes up a lot of space, and there are only so many interesting ways to do that. So I said, let’s have a list of exits at the top. Now, that’s not such a revolutionary idea. I thought of putting in a compass rose and all this other stuff, but finally I came up with an onscreen map that draws a typical Infocom map — little boxes with lines and arrows connecting them. The right side of the upper screen has a little graphics map that draws itself and updates as you walk around. It shows rooms as boxes and lines as their connections. If you open a door, a line appears to the next room. Dark rooms have question marks in them. The one you’re in is highlighted, while the others are outlined.

This onscreen map won’t replace the one you draw yourself, but it does make it much easier to draw. It won’t show rooms you haven’t been to yet, but shows exits of your current room and all the exits of the adjoining rooms that you’ve visited.

There’s even more to this re-imagining of the text adventure. On machines equipped with mice, it’s possible to move about the world by simply clicking on the auto-map; function keys are now programmable to become command shortcuts; colors can be customized to your liking; even objects in the game can be renamed if you don’t like the name they came with. And, aware that plenty of customers had a strong traditionalist streak, Moriarty also made it possible to cut the whole thing off at the knees and go back to a bog-standard text-adventure interface at any time by typing “mode” — although, with exits now absent from room descriptions, it might be a bit more confusing than usual to play that way.

Remapping the function keys.

Remapping the function keys, just one of many useful bells and whistles.

But really, it’s hard to imagine any but the most hardcore Luddites choosing to do so. The new interface is smart and playable and hugely convenient, enough to make you miss it as soon as you return to a more typical text adventure, to wish that it had made it into more than the single Infocom game that would ultimately feature it, and to wonder why more designers haven’t elected to build on it in the many years since Infocom’s demise. Infocom wrote about the new interface in their Status Line newsletter that, “never a company to jump into the marketplace with gaudy or ill-conceived bells and whistles, we have always sought to develop an intelligently measured style, like any evolving author would.” It does indeed feel like a natural, elegant evolution, and one designed by an experienced player rather than a marketeer.

With the new interface design and the technology that enabled it now well underway, Moriarty still needed an actual game to use it all. Here he made another bold step of the sort that was rapidly turning his erstwhile thought experiment into the most ambitious game in Infocom’s history. It began with another open-ended question: “What kind of game would go well with this interface?” He settled on another first for Infocom: still a text adventure, but a text adventure “with very strong role-playing elements,” in which you would have to create a character and then build up her stats whilst collecting equipment and fighting monsters.

I spent a lot of time playing fantasy games like Ultima and Wizardry and, one that I particularly liked a lot, Xyphus for the Macintosh. And I realized it was fun to be able to name your character and have all these attributes instead of having just one number — a score — that says how well you’re doing. I thought it would be nice to adopt some of the conventions from this kind of game, so you don’t have one score, you have six or seven: endurance, strength, compassion, armor class, and so on. Your job in the game is, very much like in role-playing games, to raise these statistics. And your character grows as you progress through the puzzles, some of which cannot be solved unless you’ve achieved certain statistics. You can somewhat control the types of statistics you “grow” in order to control the type of character you have.

Viewing the state of your character

Viewing the state of your character.

In conflating the CRPG with the text adventure, Moriarty was, yet again, looking backward as much as forward. In the earliest days of the entertainment-software industry, no distinction was made between adventure games and CRPGs — small wonder, as text adventures in the beginning were almost as intimately connected with the budding tabletop RPG scene as were CRPGs themselves. Will Crowther, creator of the original Adventure, was inspired to do so as much by his experience of playing Dungeons and Dragons as he was by that of spelunking in Kentucky’s Mammoth Cave. Dave Lebling was a similarly avid Dungeons and Dragons player at the time that he, along with three partners, created the original mainframe Zork, the progenitor of everything that would follow for Infocom. It was Lebling who inserted Dungeons and Dragons-style randomized combat into that game, which survived as the battles with a certain troll and thief that served as many players’ introductions to the perils of life in the Great Underground Empire in Zork I on microcomputers. About the same time, Donald Brown was creating Eamon, the world’s first publicly available text-adventure creation system that also happened to be the first publicly available CRPG-creation system. When Byte magazine made the theme of its December 1980 issue “Adventure,” no editorial distinction was made between games where you wandered around solving puzzles and those where you wandered around killing monsters. Years before Infocom would adopt the term “interactive fiction” as their preferred name for their creations, Lebling described Zork for that issue as a “computerized fantasy simulation,” a term which today smacks much more of the CRPG than the text adventure. In the same issue Jon Freeman, creator of Temple of Apshai and many of its sequels, spent quite some pages laboriously describing his approach to adventuring, which offered “character variation” that “affects the game in many ways.” He struggles, with mixed results, to clarify how this is markedly different from the approach of Zork and Scott Adams. In retrospect, it’s obvious: he’s simply describing the difference between a CRPG and a text adventure. Over time this difference became clearer, even intuitive, but for years to come the two forms would remain linked in gamers’ minds as representing separate sub-genres more so than categories onto themselves. “Adventure-game columnists” like Computer Gaming World‘s Scorpia, for instance, continued to cover both into the 1990s and beyond.

That’s by no means inexplicable. The two forms shared plenty in common, like a story, a love of fantasy settings, and the need to explore a computer-simulated world and (usually) to map it. The forms were also connected in being one-shot games, long experiences that you played through once and then put aside rather than shorter experiences that you might play again and again, as with most action and strategy games of the period. And then there was the simple fact that neither would likely have existed in anything like the form we know them if it hadn’t been for Dungeons and Dragons. Yet such similarities can blind us, as it did so many contemporary players, to some fairly glaring differences. We’ll soon be seeing some of the consequences of those differences play out in Moriarty’s hybrid.

By the time that Moriarty’s plans had reached this stage, it was the fall of 1986, and Infocom was busily lining up their biggest slate of new games ever for the following year. It had long since been decided that one of those games should bear the Zork name, the artistic fickleness that had led the Imps to reject Infocom’s most recognizable brand for years now seeming silly in the face of the company’s pressing need for hits. Steve Meretzky, who loved worldbuilding in general, also loved the lore of Zork to a degree not matched even by the series’s original creators. While working on Sorcerer, he had assembled a bible containing every scrap of information then “known” — more often than not in the form of off-hand asides delivered strictly for comedic effect — about the Flathead dynasty, the Great Underground Empire, and all the rest of it, attracting in the process a fair amount of ridicule from other Imps who thought he was taking it all far, far too seriously. (One shudders to think what they would say about The Zork Compendium.) Now Meretzky was more than eager to do the next Zork game. Whirling dervish of creativity that he was, he even had a plot outline to hand for a prequel, which would explain just how the Great Underground Empire got into the sorry state in which you first find it in Zork I. But there was a feeling among management that the long-awaited next Zork game ought to really pull out all the stops, ought to push Infocom’s technology just as far as it would go. That, of course, was exactly what Moriarty was already planning to do. His plan to add CRPG elements would make a fine fit with the fantasy milieu of Zork as well. Moriarty agreed to make his game a Zork, and Meretzky, always the good sport, gave Moriarty his bible and proceeded to take up Stationfall, another long-awaited sequel, instead.

It was decided to call the new Zork game Beyond Zork, a reference more to its new interface and many technical advancements than to the content of the game proper. Having agreed to make his game a Zork, Moriarty really made it a Zork, stuffing it with every piece of trivia he could find in Meretzky’s bible or anywhere else, whilst recreating many of the settings from the original Zork trilogy as well as the Enchanter trilogy. Beyond Zork thus proclaimed to the world something that had always been understood internally by Infocom: that the Enchanter trilogy in a different reality — a better reality according to marketing director Mike Dornbrook’s lights — could have just as easily shipped as Zork IV through VI. But Moriarty didn’t stop there. He also stuffed Beyond Zork with subtler callbacks to almost the entire Infocom catalog to date, like the platypus, horseshoe, and whistle from Wishbringer and the magical umbrella from Trinity. Mr. Prosser, Arthur Dent’s hapless nemesis from The Hitchhiker’s Guide to the Galaxy, shows up as the name of a spell, and Buck Palace, the statuesque B-movie star from Hollywood Hijinx, is the default name for your character if you don’t give him another. There are so many references that upon the game’s release Infocom sponsored a contest to see who could spot the most of them. Especially in retrospect, knowing as we do that we are now coming close to the end of the line for Infocom, all of the backward glances can take on an elegiac quality, can make Beyond Zork feel like something of a victory lap for an entire era of adventure gaming.

I’m less pleased with the actual plot premise of Beyond Zork, which, as unplanned sequels so often tend to do, rather clumsily undermines the message of its predecessor. Moriarty’s game builds on Spellbreaker, which saw you destroying magic in the name of saving the world. The ending of Spellbreaker:

You find yourself back in Belwit Square, all the Guildmasters and even Belboz crowding around you. "A new age begins today," says Belboz after hearing your story. "The age of magic is ended, as it must, for as magic can confer absolute power, so it can also produce absolute evil. We may defeat this evil when it appears, but if wizardry builds it anew, we can never ultimately win. The new world will be strange, but in time it will serve us better."

Strange as it may sound, I judge that jarring last sentence to be nothing less than Dave Lebling’s finest moment as a writer. It’s an ending that can be read to mean many things, from an allegory of growing up and leaving childish things behind to a narrative of human progress as a whole — the replacing of a “God of the gaps” with real knowledge, simultaneously empowering and depressing in the way it can leach the glorious mystery out of life.

But still more depressing is the way that Beyond Zork now comes along to muck it up. You play a novice enchanter (where have we heard that before?) who, even as the wise and powerful hero of Spellbreaker sets about destroying magic, is dispatched by the Guild to retrieve the “Coconut of Quendor” that can safeguard it for a return at some point in the future. Why couldn’t Moriarty just leave well enough alone, find some other premise for his game of generic fantasy adventure? About the best thing I can say about the plot is that you hardly know it’s there when you’re actually playing the game; until the last few turns Beyond Zork is largely a plot-free exercise in puzzles, self-improvement (in the form of easily quantifiable statistics and equipment), and monster bashing.

On those terms, Beyond Zork seems to acquit itself quite well in the early going. Brian Moriarty remains the most gifted prose stylist among the Imps, crafting elegant sentences that must make this game, if nothing else, among the best written generic fantasies ever. Zork always had a bipolar personality, sometimes indulging in unabashed comedy and at others evoking majestic, windy desolation. Perhaps surprisingly in that it comes from the author of the doom-laden Trinity, Beyond Zork leans more toward the former than the latter. But then anyone who’s played Wishbringer knows that Moriarty can do comedy — and the juxtaposition of comedy with darker elements — very well indeed. I particularly like the group of Implementors you meet; this sort of meta-comedy was rapidly becoming another Zork tradition, dating back to the appearance of the adventurer from the original trilogy in Enchanter or, depending on how you interpret the ending of Zork III, possibly even earlier. Moriarty lampoons the reputation the Imps had within Infocom, usually expressed jokingly but not always without an edgy undercurrent, of being the privileged kids who always got all the free lunches and other perks, not to mention the public recognition, while the rest of the company, who rightfully considered themselves also very important to Infocom’s success, toiled away neglected and anonymous.

Ethereal Plane Of Atrii, Above Fields
The thunderclouds are compressed into a flat optical plane, stretching away below your feet in every direction.

A group of Implementors is seated around a food-laden table, playing catch with a coconut.

One of the Implementors notices your arrival. "Company," he remarks with his mouth full.

A few of the others glance down at you.

>x coconut
It's hard to see what all the fuss is about.

A tall, bearded Implementor pitches the coconut across the table. "Isn't this the feeb who used the word 'child' a few moves ago?" he mutters, apparently referring to you. "Gimme another thunderbolt."

>get coconut
The Implementors won't let you near.

A cheerful-looking Implementor catches the coconut and glares down at you with silent contempt.

"Catch!" cries the cheerful-looking Implementor, lobbing the coconut high into the air.

"Got it." A loud-mouthed Implementor jumps out of his seat, steps backwards to grab the falling coconut... and plows directly into you.

Plop. The coconut skitters across the plane.

>get coconut
As you reach towards the coconut, a vortex of laughing darkness boils up from underfoot!

"More company," sighs the cheerful-looking Implementor.

You back away from the zone of darkness as it spreads across the Plane, reaching out with long black fingers, searching, searching...

Slurp! The coconut falls into the eye of the vortex and disappears, along with a stack of lunch meat and bits of cutlery from the Implementors' table. Then, with a final chortle, the vortex draws itself together, turns sideways and flickers out of existence.

"Ur-grue?" asks the only woman Implementor.

"Ur-grue," nods another.

>ask implementors about ur-grue
"I think I just heard something insignificant," remarks an Implementor.

"How dull," replies another, stifling a yawn.

"This is awkward," remarks a loudmouthed Implementor. "No telling what the ur-grue might do with the Coconut. He could crumble the foundations of reality. Plunge the world into a thousand years of darkness. We might even have to buy our own lunch!" The other Implementors gasp. "And it's all her fault," he adds, pointing at you with a drumstick.

"So," sighs another Implementor, toying with his sunglasses. "The Coconut is gone. Stolen. Any volunteers to get it back?"

One by one, the Implementors turn to look at you.

"I'd say it's unanimous," smiles the cheerful-looking Implementor.

A mild-mannered Implementor empties his goblet of nectar with a gulp. "Here," he says, holding it out for you. "Carry this. It'll keep the thunderbolts off your back."

>get goblet
The Implementor smiles kindly as you take the goblet. "And now you will excuse us. My fellow Implementors and I must prepare for something too awesome to reveal to one as insignificant as you."

At first, the CRPG elements seem to work better than one might expect. Randomized combat in particular has for many, many years had a checkered reputation among interactive-fiction fans. It’s very difficult to devise a model for combat in text adventures that makes the player feel involved and empowered rather than just being at the mercy of the game’s random-number generator. There weren’t many fans of the combat in Zork I even back in the day, prompting Infocom to abandon it in all of their future games; only the usually pointless “diagnose” command remained as a phantom limb to remind one of Zork‘s heritage in Dungeons and Dragons. Even Eamon as time went by moved further and further away from its original incarnation as a sort of text-only simulation of Dungeons and Dragons, its scenarios beginning to focus more on story, setting, and puzzles than killing monsters.

Battling a rat-ant (say, is that a Starcross reference?) in Beyond Zork.

Battling a rat-ant (say, is that a Starcross reference?) in Beyond Zork.

Beyond Zork doesn’t entirely solve any of the problems that led to those developments, but it does smartly make the process of preparing for the combats more compelling than the combats themselves can possibly be. As you explore the world and solve puzzles, you level up and improve your ability scores. Among other things, this lets you fight better. You can also sell treasure for money, a nice twist on the treasure-hunt model of old-school text adventuring, and use it to buy better weapons, armor, and magic items. As you improve yourself through these means and others, you find that you can challenge tougher monsters. Thus, while the combat is still not all that interesting in itself — it still comes down to the same old “monster hits you for X points of damage, you hit monster for Y points of damage,”  rinse and repeat until somebody is done for — the sudden ability to win a battle in which you previously didn’t have a chance can feel surprisingly rewarding. Many of the monsters take the place of locked doors in more conventional text adventures, keeping you out of places you aren’t yet ready for. It feels about as satisfying to defeat one of these as it does to finally come across the key for a particularly stubborn lock in another game. Indeed, I wish that Moriarty had placed the monsters more carefully in order to guide you through the game in the right order and keep you from locking yourself out of victory by doing the right things in the wrong order, as I’ll describe in more detail shortly.

It also helps that the combats are usually quite low-stakes. While undo is disabled in combat — how ironic that Infocom introduced undo in their first game ever with a good reason not to allow it! — it’s always obvious very quickly when you’re over-matched, and usually fairly trivial to back away and go explore somewhere else before you get killed. One other subtle touch, much appreciated by a big old softie like me who can even start to feel bad for the monsters he kills in Wizardry, is that you rarely actually kill anything in Beyond Zork. Monsters usually “retreat into the darkness” or something similar rather than expiring — or at least before expiring. While I suspect Moriarty did this more to avoid implementing dead monster bodies than to make a statement, I’ll take my instances of mercy wherever I can find them.

Beyond Zork blessedly doesn’t take itself all that seriously, whether as a CRPG or as anything else. Instead of the painfully earnest orcs and dragons that populate most CRPGs, Beyond Zork‘s monsters are almost uniformly ridiculous: Christmas tree monsters singing dreadful carols (anyone who’s ever visited a shopping mall on Black Friday can probably relate), cruel puppets who attack by mocking you (they “recite your nightly personal habits in excruciating detail” among other attacks), dust bunnies (got any lemony-fresh Pledge handy?). In a sense Beyond Zork is just another genre exercise for Infocom, albeit in a ludic rather than a literary genre this time, and as usual for them it can sometimes feel as much parody as homage. There are alternate, puzzlely solutions that can be used to defeat many monsters in lieu of brute strength, and usually in more entertaining fashion at that. In fact, in many cases cleverness is the only way forward. Vital clues to the various monsters’ weaknesses are found in the game’s version of Dungeons and Dragons‘s Monster Manual, the accompanying feelie “The Lore and Legends of Quendor”; if only other companies offered such entertaining and clever copy protection! And if fighting fair and square or puzzling your way around the monsters aren’t enough alternatives for you, you can just use your Wand of Annihilation or other magic. Not just monsters but also many other problems can similarly be defeated in various ways. Beyond Zork is surprisingly flexible.

But, sadly, it’s often dangerously flexible. Which brings us to the part of this article where I have to tell you why, despite its many innovations, despite Moriarty’s usual fine writing, and despite some fine puzzles, Beyond Zork in my opinion just doesn’t add up to all that great of a game.  I think I can best explain how things go wrong by dissecting one particularly dismaying sequence in some detail.

Fair warning: heavy spoilers begin here!

So, I come upon a mother and baby hungus — “part sheep, part hippopotamus,” as “The Legends and Lore of Quendor” helpfully tells us — caught in a bad situation.

Quicksand
A strip of dry path winds alongside a pool of quicksand. You see a baby hungus stuck into the wet, gritty surface.

A mother hungus is standing nearby, gazing anxiously at her baby.

Luckily, I happen to have found a Stave of Levitation. What I need to do seems pretty obvious.

>point stave at baby
The baby hungus bellows with surprise as he rises out of the quicksand! Sweat breaks out on your forehead as you guide the heavy burden over the mud and safely down to the ground.

[Your strength just went down.]

The ungainly creature nuzzles you with his muddy snout, and bats his eyelashes with joy and gratitude. Then he ambles away into the jungle to find his mother, pausing for a final bellow of farewell.

[Your compassion just went up.]

Heartwarming, isn’t it? But you know what’s less heartwarming? The fact that I’ve just locked myself out of victory, that’s what. I needed to do something else, which we’ll get to in a moment, before rescuing the little fellow. It will, needless to say, likely be a long, long time before I realize that, especially given that the game actually rewards me by increasing my Compassion score. (As for the Strength loss, never fear, I’ll recover that automatically in a few turns.) This is terrible design, of the sort I expect from early Sierra, not late Infocom. I wrote in an earlier article that “we should reserve a special layer of Hell for those designs whose dead ends feel not just like byproducts of their puzzles and other interactive possibilities but rather intentional traps.” See you down below, Beyond Zork.

Beyond Zork is absolutely riddled with these sorts of traps, forcing you to restart many, many times to get through it, wondering all the while whether you didn’t render this latest puzzle you’re wrestling with insoluble a long time ago by some innocent, apparently correct action like the one above. What makes this even more baffling is that in other ways Beyond Zork is presented as an emergent, replayable experience, the sort of game where you take your lumps and move on until you either win or lose rather than constantly restoring and/or restarting to optimize your play. Taking a cue from the roguelike genre, large chunks of Beyond Zork‘s geography are randomly generated anew every time you play, as are the placement of most magic items and their descriptions; that Stave of Levitation I just used may be a Stave of Annihilation in another playthrough, forcing me to make sure I make good use of the various shops’ ability to identify stuff for me. As the manual tells you, “No two games of Beyond Zork are exactly alike!” But what’s the point of that approach when most of those divergent games leave you fruitlessly wandering about, blocked at every turn, wondering where you went wrong? Even an infamously difficult roguelike like NetHack at least puts you out of your misery when you screw up. By the time you do figure out how to tiptoe through this minefield of dead ends, you’ve internalized the whole game to such an extent that the randomness is just a huge annoyance.

The deterministic text adventure and the emergent CRPG end up rubbing each other raw almost every time they touch. When you discover a cool new magic wand or spell, you’re afraid to use it to vanquish that pesky monster you’ve been struggling with for fear that you’ll need to use it to solve some deterministic puzzle somewhere else. Yes, resource management was always a huge part of old-school dungeon crawls like Wizardry and The Bard’s Tale — arguably the hugest part, given that their actual combat engines were often little more sophisticated than that of Beyond Zork — but there was always a clear distinction between things you might need to solve puzzles and things for managing combat. The lack of same here is devastating; the CRPG aspects are really hard to enjoy when you’re constantly terrified to actually use any of the neat equipment you collect.

And just as the text adventure undermines the CRPG, the CRPG also undermines the text adventure. Near the hungi — never fear, we’ll return to that problem shortly! — I find this, yet another callback to Spellbreaker:

>w
Idol
A stone idol, carved in the likeness of a giant crocodile, stands in a clearing.

You see a tear-shaped jewel in its gaping maw.

You see a tear-shaped jewel on the idol's maw.

>x idol
This monstrous idol is approximately the size and shape of a subway train, not counting the limbs and tail. The maw hangs wide open, its lower jaw touching the ground to form an inclined walkway lined with rows of stone teeth. A tear-shaped jewel adorns the idol's face, just below one eye.


>enter idol
You climb up into the idol's maw.

The stone jaw lurches underfoot, and you struggle to keep your balance. It's like standing on a seesaw.

>get jewel
The idol's maw tilts dangerously as you reach upward!
Slowly, slowly, you draw your hand away from the tear-shaped jewel, and the jaw settles back to the ground.

>u
You edge a bit further into the open maw.

Creak! The bottom of the jaw tilts backward, pitching you helplessly forward...

This sequence begins a veritable perfect storm of problems, starting with the text itself, which contradicts itself and thus makes it hard to figure out what the situation really is. Is the jewel in the idol’s maw, on the idol’s maw (the weird doubled text is present in the original), or stuck to the idol’s face? Or are there two jewels, one stuck on his face and one in (on?) his maw? I’m still not sure. But let’s continue a bit further.

>turn on lantern
Click. The lantern emits a brilliant glow.

Inside Idol
This long, low chamber is shaped much like the gizzard of a crocodile. Trickles of fetid moisture feed the moss crusting the walls and ceiling.

>squeeze moss
The moss seems soft and pliant.

The moss is “Moss of Mareilon.” As described in “The Lore and Legends of Quendor,” squeezing it as I’ve just done will lead to a dexterity increase a few turns from now. And so we come to one of the most subtly nasty bits in Beyond Zork. To fully explain, I need to back up just a little.

Earlier in the game, in a cellar, I needed to climb a “stairlike spiral” of crates to get something on top of them. Alas, I wasn’t up to it thanks to a low dexterity: “You teeter uncertainly on the lowest crates, lose your balance, and sprawl to the ground. Not very coordinated, are you?” Luckily, some Moss of Mareilon was growing right nearby; I could squeeze it to raise my dexterity enough to get the job done.

So, when I come to the idol, and find more Moss of Mareilon growing conveniently just inside it, that combined with the description of my somewhat clumsy effort to grab the jewel leads to a very natural thought: that I need to squeeze the moss inside the idol to increase my dexterity enough to grab the jewel. I do so, then use a handy magic item to teleport out, then do indeed try again to grab the jewel. But it doesn’t work; I still can’t retrieve the jewel, still get pitched down the idol’s throat every time. After struggling fruitlessly with this poorly described and poorly implemented puzzle — more on that in a moment — I finally start thinking that maybe it can’t be solved because I didn’t give my character enough dexterity at the very beginning of the game. This can actually happen; Beyond Zork is quite possibly the only text adventure ever written in which you can lock yourself out of victory before you even enter your first command. (“The attributes of the ‘default’ [pre-created] characters are all sufficient to complete the story,” says the manual, which at least lets you know some of what you’re in for if you decide to create your own.) So, I create a brand new character with very good dexterity, and spend an hour or so, cursing all of the randomizations all the while, to get back to the idol puzzle. But I still can’t fetch the diamond, not even after squeezing the moss.

It does seem that the game has, once again, actively chosen to mislead me and generally screw me over here. But, intentionality aside, a more subtle but more fundamental problem is the constant confusion between player skill, the focus of a text adventure, and character skill, the focus of a CRPG. Let me explain.

A text adventure is a much more embodied experience than a CRPG. There’s a real sense that it’s you — or, increasingly in later games, a character that you are inhabiting — whom you are guiding through the simulated world. Despite the name, meanwhile, a CRPG is a more removed experience. You play a sort of life coach guiding the development of one or more others. Put another way, one genre emphasizes player skill, the other character skill. Think about what you spend the most time doing in these games. The most common activity in a text adventure is puzzle-solving, the most common in a CRPG combat. The former relies entirely on the wit of the player; the latter, if it’s done well, will involve plenty of player strategy, but it’s also heavily dependent on the abilities of the characters you guide. After all, no amount of strategy is going to let you win Wizardry or The Bard’s Tale with a level 1 party. CRPGs are process-intense simulations to a greater degree than text adventures, which rely heavily on hand-crafted content, often — usually in these early years — in the form of puzzles of one stripe or another. This led Jon Freeman in his Byte article to call text adventures, admittedly rather reductively, not simulations or even games at all but elaborate puzzle boxes built out of smaller puzzles: “It can be quite challenging to find the right key, the right moment, and the right command to insert it in the right lock; but once you do, the door will open — always.” In a CRPG, on the other hand, opening that lock might depend on a random number and some combination of a character’s lock-picking ability, the availability of a Knock spell, and/or the quality of the lock picks in her pack. More likely, the locked door won’t be there at all, replaced with some monster to fight. Sure things aren’t quite so common.

Of course, these distinctions are hardly absolute. CRPGs, for example, contained occasional player-skill-reliant puzzles to break up their combat almost from the very beginning. Muddying up the player-skill/character-skill dichotomy too much or too thoughtlessly can, however, be very dangerous, as Beyond Zork has just so amply demonstrated. While one hardly need demand an absolutely pure approach, one does need to know where the boundaries lie, which problems you can solve by solving a puzzle for yourself and to which you need to apply some character ability or other. Those boundaries are never entirely clear in Beyond Zork, and the results can be pretty ugly.

All that said, I still haven’t actually solved the idol puzzle. I guessed quite quickly that the description of standing on the idol’s maw as “like standing on a seesaw” was a vital clue. The solution, then, might be to place a counterweight at the front of the maw to keep it from pitching up and pitching me in. Yet the whole thing is so sketchily described and implemented that I remain unsure what’s actually happening. If I start piling things up inside the maw, they do fall down into the idol’s stomach along with me when I reach for the jewel — apparently, anyway; they’re not described as falling with me, but they do show up in the stomach with me once I get there. But if those things fall through, why not the jewel? If it’s sitting on the idol’s tongue, it’s hard to imagine why it wouldn’t. Or is it actually stuck to the idol’s face, and the stuff about it being in the maw is all a big mistake? Yes, we’re back to that question again. It’s worth noting that the quicksand area and the hungi therein are similarly subtly bugged. If I rescue the baby before doing something with the mother, he’s described as “ambling away into the jungle to find his mother” even though she’s right there, and, since I’ve just hopelessly screwed up, will now remain there forevermore.

The solution to the idol puzzle is clued in “The Lore and Legends of Quendor.” A hungus, it says, “will instantly charge at anything that dares to threaten its kin.” I have to attack the baby hungus and make use of the mother’s rage before rescuing her son.

>attack baby
[with the battleaxe]
Your battleaxe misses the baby hungus. It's just beyond your reach.

A sound like a snorting bull turns your attention to the mother hungus. It looks as if she's about to attack!

The mother hungus charges you. Ooof!

[Your endurance just went down.]

The baby hungus bellows helplessly, and its mother responds.

>ne
The baby hungus bellows mournfully as you walk away.


Birdcries
The unnerving cries of exotic birds echo in the treetops.

>z
Time passes.

The mother hungus storms into view!

>s
Idol
A stone idol, carved in the likeness of a giant crocodile, stands in a clearing.

You see a tear-shaped jewel in its gaping maw.

You see a tear-shaped jewel on the idol's maw.

>z

Time passes.

The mother hungus storms into view!

[Your endurance is back to normal.]

>enter maw
You climb up into the idol's maw.

The stone jaw lurches underfoot, and you struggle to keep your balance. It's like standing on a seesaw.

The mother hungus clambers onto the bottom of the idol's maw, snorting with rage!

>get jewel
The idol's maw tilts dangerously as you reach upward, standing on tiptoe to grasp the sparkling treasure...

Got it! The jewel pops off the idol's face, slips from your grasp and rolls down to the mother hungus's feet, where she promptly eats it, turns and lumbers off the jaw.

Creak! The bottom of the jaw tilts backward, pitching you helplessly forward...

Getting the jewel out of the hungus is another puzzle, but a much better one, so I won’t spoil it here. (No, it doesn’t involve a laxative…)

Spoilers end.

The sequence I’ve just described is probably the ugliest in the game, but other parts suffer to a greater or lesser degree from many of the same problems. Many of the bugs and textual confusions can doubtless be laid at the feet of an overambitious release schedule, while some of the sketchy implementation is likely down to the space limitations of even the 256 K Z-Machine. In recent correspondence with another Infocom aficionado, we talked about how Trinity, another 256 K game, really doesn’t feel all that huge, how much of the extra space was used to offer depth in the form of a larger vocabulary, richer text, and more player possibility rather than breadth in the form of more rooms and puzzles. Beyond Zork, by contrast, does feel quite huge, marks the most overstuffed game that Infocom had yet released. Considering that it must also support a full-fledged, if simplistic, CRPG engine, depth was quite obviously sacrificed in places.

But Beyond Zork‘s most fundamental failing is that of just not knowing what it wants to be. In an effort to make a game that would be all things to all people, the guaranteed hit that Infocom so desperately needed, Moriarty forgot that sometimes a game designer needs to say, no, let’s save that idea for the next project. Cognitive dissonance besets Beyond Zork from every angle. The deterministic, puzzle-oriented text adventure cuts against the dynamic, emergent CRPG. The cavalcade of in-jokes and references to earlier games, catnip for the Infocom hardcore, cuts against appealing to a newer, possibly slightly younger demographic who are fonder of CRPGs than traditional text adventures. The friendly, approachable interface cuts against a design that’s brutally cruel — sometimes apparently deliberately so, sometimes one senses (and this is in its way more damning to Brian Moriarty as a designer) accidentally so. Beyond Zork stands as an object lesson in the perils of mixing ludic genres willy-nilly without carefully analyzing the consequences. I want it to work, love many of the ideas it tries to implement. But sadly, it just doesn’t. It’s a bit of a mess really, not just difficult, which is fine, but difficult in all the wrong, unfun ways. Spellbreaker is a perfect example of how to do a nails-hard text adventure right. Beyond Zork, its parallel in the Zorkian chronology, shows how to do it all wrong. I expect more from Infocom, as, based on Wishbringer and Trinity, I do from Brian Moriarty as well.

Despite receiving plenty of favorable reviews on the basis of its considerable surface appeal, Beyond Zork didn’t turn into the hit that Infocom needed it to become. Released in October of 1987, it sold a little over 45,000 copies. Those numbers were far better than those of any of the other Infocom games of 1987, proof that the Zork name did indeed still have some commercial pull, but paled beside the best sellers of previous years, which had routinely topped 100,000 copies. Enough, combined with the uptick in sales of their other games for the Christmas season, to nudge Infocom into the black for the last quarter of 1987, Beyond Zork wasn’t enough to reverse the long-term trends that were slowly strangling them. Moriarty himself left Infocom soon after finishing Beyond Zork, tempted away by Lucasfilm Games, whose own adventure-gaming star was rising as Infocom’s fell, and where he would at last be able to fulfill his ambition to dump the parser entirely. We’ll be catching up with him again over there in due time.

(Sources: As usual with my Infocom articles, much of this one is drawn from the full Get Lamp interview archives which Jason Scott so kindly shared with me. Other sources include the book Game Design Theory and Practice by Richard Rouse III; Byte of December 1980; Questbusters of August 1987 and January 1988; Commodore Magazine of March 1988.

Beyond Zork is available for purchase as part of The Zork Anthology on GOG.com.)

 
 

Tags: , ,

Sorcerer

Sorcerer

Steve Meretzky was a boundless fount of creative energy which couldn’t be contained by even his official projects for Infocom, many and varied as they were, and spilled over into daily life around the office in the form of elaborate themed parties, games that ranged from a multiplayer networked version of Boggle played over the DEC minicomputer to intense Diplomacy campaigns, and endless practical jokes. (“Memo hacking” became a particular favorite as Business Products ramped up and more and more buttoned-down business types started to appear in the office.) The lore and legends of daily life at Infocom, eagerly devoured by the faithful via the New Zork Times newsletter, is largely the lore and legends of Steve Meretzky, instigator and ringleader behind so much of the inspired lunacy.

Yet there was also another, oddly left-brained side to Meretzky. He was a compulsive organizer and even a bit of a neat freak; his meticulous and breathtakingly thorough archives informed much of Jason Scott’s Get Lamp project and, by extension, much of the Infocom history on this site. Mike Dornbrook, Infocom’s marketing director, calls Meretzky the most productive creative person he has ever met, one who evinced not a trace of the existential angst that normally accompanies the artistic temperament. Writer’s block was absolutely unknown to him; he could just “turn it on” and pour out work, regardless of what was happening around him or how things stood in his personal life.

But there was still another trait that made Meretzky the dream employee of any manager of creative types: he was literally just happy to be at Infocom, thrilled to be out of a career in construction management and happy to work on whatever project needed him. And so when Dave Lebling decided he’d like to write a mystery game and Marc Blank wanted to work on technology development, leaving the critical second game in the Enchanter trilogy without an author, Meretzky cheerfully agreed to take it on. When a certain famous but mercurial and intimidating author of science-fiction comedies came calling and everyone else shied away from collaborating with him, Meretzky said sure, sounds like fun. And when Tor Books offered Infocom the chance to make a series of Zork books in the mold of the absurdly successful Choose Your Own Adventure line, and everyone on the creative staff turned up their noses at such a lowbrow project even as management rubbed their hands in glee at the dollar figures involved, Meretzky took the whole series on as his moonlighting gig, cranking out four books that were hardly great literature but were better than they needed to be. Most gratifyingly of all, Meretzky ripped through all of these projects in a bare fifteen months whilst offering advice and ideas for other projects and, yes, getting up to all that craziness that New Zork Times readers came to know and cherish. Meretzky was truly a dream employee — and a dream colleague. One senses that if management had asked him to go back to testing after finishing Planetfall he would have just smiled and kicked ass at it.

Sorcerer, his sequel to Enchanter and Infocom’s first game of 1984, was, like so much of Meretzky’s work in this period, a bit of a thankless task. He neither got to devise the overarching plot and mechanics for the trilogy nor to bring things to a real conclusion, merely to write the bridge between fresh beginning and grand climax. Middle works in trilogies have always tended to be problematic for this very reason, and, indeed, Sorcerer is generally the most lightly regarded of the Enchanter games. I won’t really argue with that opinion, but I will say that Sorcerer is a very solid, entertaining work in its own right. It’s just that it gets a bit overshadowed by its towering companions, together arguably the best purely traditional adventure games ever to come out of Infocom, while also lacking the literary and thematic innovations that make games like Planetfall and Infidel — to neither of which it’s actually markedly inferior in overall quality — so interesting for people like me to write about.

Sorcerer casts you as the same budding enchanter you played in the game of that name. Having vanquished Krill, however, your star has risen considerably; you are now a member of magic’s innermost circle, the Circle of Enchanters, and protege of the Leader of the Circle, Belboz. Sorcerer opens with one of its most indelibly Meretzkian sequences. You are snug in your bed inside the Guild of Enchanters — but you don’t actually realize that for a few turns.

You are in a strange location, but you cannot remember how you got here. Everything is hazy, as though viewed through a gauze...

Twisted Forest
You are on a path through a blighted forest. The trees are sickly, and there is no undergrowth at all. One tree here looks climbable. The path, which ends here, continues to the northeast.
A hellhound is racing straight toward you, its open jaws displaying rows of razor-sharp teeth.

>climb tree
Tree Branch
You are on a large gnarled branch of an old and twisted tree.
A giant boa constrictor is slithering along the branch toward you!
The hellhound leaps madly about the base of the tree, gnashing its jaws.

>i
You are empty-handed.
The snake begins wrapping itself around your torso, squeezing the life out of you...

...and a moment later you wake up in a cold sweat and realize you've been dreaming.

SORCERER: INTERLOGIC Fantasy
Copyright (c) 1984 by Infocom, Inc. All rights reserved.
SORCERER and INTERLOGIC are trademarks of Infocom, Inc.
Release 4 / Serial number 840131

Your frotz spell seems to have worn off during the night, and it is now pitch black.

Like the similarly dynamic openings of Starcross and Planetfall, albeit on a more modest scale, Sorcerer‘s dream sequence can be a bit of a misnomer. The rest of the game is much more open-ended and much less plot-driven than this sequence might imply. As you explore the conveniently deserted Guild — everyone except you and Belboz have gone into town to shop for the Guild picnic — you soon realize that Belboz has mysteriously disappeared. And so the game is on, fueled by the same sort of magic-based puzzles that served Enchanter so well. Indeed, Meretzky copied the code for the Enchanter magic system wholesale into Sorcerer, along with some of the same spells, which had to be a great help for someone working on as tight a timetable as he was. Sorcerer‘s one big magical innovation is a set of potions to accompany its spell scrolls, something notably absent not only from Enchanter but also from Lebling’s Spellbreaker, the final game of the trilogy.

Like all of the Enchanter trilogy a very traditional game, Sorcerer is divided into two open-ended areas of exploration, the Guild of Enchanters and a sprawling wilderness and underground map which ultimately proves to house Belboz’s abductor, the demon Jeearr (another thoroughly Meretzkian name, and a character who also turns up in the last of the Zork gamebooks he was writing at the same time). The overall feel is looser than Enchanter, with the first game’s understated humor replaced with a more gonzo sensibility that can rub some players the wrong way. This player, who felt that Planetfall often seemed to be trying just a bit too hard, doesn’t exactly find Sorcerer hilarious but never really found it irritating in the way that Meretzky’s earlier game could occasionally be either. Perhaps the fact that Sorcerer wasn’t explicitly billed as a comedy left Meretzky feeling freer not to force the issue at every possible juncture.

Another Planetfall trait, that of lots of Easter eggs and red herrings, is also notable in Sorcerer, but again to a lesser extent. The useless bits, such as a functioning log flume and roller coaster inside the amusement park inexplicably located almost next door to Jeearr’s infernal lair, are mostly good fun. The sadomasochistic “potion of exquisite torture” is a standout that is just a bit risque for the prudish world of adventure gaming:

>drink indigo potion
The potion tastes like a combination of anchovies, prune juice, and garlic powder. As you finish swallowing the potion, a well-muscled troll saunters in. He whacks your head with a wooden two-by-four, grunting "You are playing Sorcerer. It was written by S. Eric Meretzky. You will have fun and enjoy yourself." He repeats this action 999 more times, then vanishes without a trace.

Another great bit comes if you use the aimfiz spell — “transport caster to someone else’s location” — to try to find Meretzky himself:

>cast aimfiz on meretzky
As you cast the spell, the moldy scroll vanishes!
You appear on a road in a far-off province called Cambridge. As you begin choking on the polluted air, a mugger stabs you in the back with a knife. A moment later, a wild-eyed motorist plows over you.

**** You have died ****

Like any old-school adventure game, Sorcerer is full of goofy and often random ways to die, from wandering into a room that’s missing a floor to getting buried under coins by an overenthusiastic slot machine. Still, Meretzky manages to skirt the letter if not quite the spirit of Andrew Plotkin’s Cruelty Scale through the gaspar spell: “provide for your own resurrection.” Gaspar returns you upon your death alive and well to the place where you last cast it, a handy substitute for the technological rather than arcane solution of restoring a saved game. If nothing else, its presence proves that Infocom was thinking about the arbitrary cruelty of most adventure games and wondering if a friendlier approach might be possible. (Space limitations would, however, always limit how far they could travel down this path. It would always be easier to simply kill the player than try to implement the full consequences of a bad — or simply unplanned for — decision.) Another sign of evolving thought on design comes in the form of the berzio potion (“obviate need for food or drink”), which slyly lets you bid adieu to the hunger and thirst timers of Enchanter and Planetfall. A year later, Spellbreaker would not even bother you with the whole tedious concept at all.

As the presence of amusement parks and casinos next to abducted enchanters and demons would imply, Sorcerer doesn’t concern itself at all with the fictional consistency that marked Planetfall or even, for that matter, Enchanter. Plot also takes a back seat for most of the game. You simply explore and solve puzzles until you suddenly bump into Jeearr and remember why you’re here. Likewise, some of the writing is a bit perfunctory if we insist on viewing Sorcerer as a literary experience. That, however, is not its real strength.

I find Meretzky slightly overrated as a writer but considerably underrated as a master of interlocking puzzle design. Sorcerer is full of clever puzzles, one of which, a relatively small part of the brilliant time-travel sequence in the coal mine, represents the last little bit of content which Infocom salvaged from the remaining scraps of the original MIT Zork. Yet it isn’t even one of the most memorable puzzles in Sorcerer; those are all Meretzky originals. In addition to that superb time-travel puzzle, there’s a fascinating thing that seems to be a maze but isn’t — quite. Both time-travel puzzles and pseudo-mazes were already burgeoning traditions at Infocom; both would remain obsessions of the Imps for years to come. Meretzky does both traditions proud here. I won’t say too much more about Sorcerer‘s puzzles simply because you really should enjoy them for yourself if you haven’t already. They’re always entertaining, clever, and (sudden deaths and one tricky sequence involving a timed mail delivery early in the game aside) fair, and don’t deserve to be spoiled by the likes of me.

Sorcerer shipped in March of 1984 in a box that was fairly plebeian for this era of Infocom. The crown jewel was contained inside the box this time, in the form of the infotater, an elaborately illustrated code wheel that was both one of Infocom’s most blatant uses yet of a feely as unabashed copy protection and so cool that it didn’t really matter. The infotater is today among the rarer pieces of Infocom ephemera. It remained in production for just a few months before Infocom switched to a standardized box format that was too small to accommodate it, and were thus forced to replace it with a less interesting table of information on plain paper.

Sorcerer sold decently, although not quite as well as Enchanter or the Zork games. (The steady downward trend in sales of Infocom’s flagship line of fantasy games would soon become a matter of increasing concern — but more on that in future articles.) Lifetime sales would end up in the vicinity of 45,000, with more than two-thirds of those coming in 1984 alone. It’s not one of the more ambitious games of Infocom nor, truth be told, one of the absolute best, but it is a solid, occasionally charming, playable game. If you find yourself in the mood for an enjoyable traditional text adventure that plays relatively fair with you, you could certainly do a lot worse.

(As always, thanks to Jason Scott for sharing his materials from the Get Lamp project.)

 
 

Tags: , , ,

Enchanter

Enchanter

In his seminal book Hackers, Steven Levy compares the differing cultures of the East Coast hackers at MIT and the West Coast hackers at Stanford during the glory days of 1970s institutional computing by riffing on their literary preferences. The MIT folks, he claims, preferred “the battle-strewn imagery of shoot-’em-up science fiction,” while those at Stanford went in for “the gentle lore of elves, hobbits, and wizards.” He then goes on to describe how these preferences show up the differing cultures inside the institutions. MIT is competitive, practical, a bit traditionalist and perhaps even prudish, a microcosm of the high-strung East-Coast establishment; while Stanford, having imbibed from the remnants of the hippie dream that persisted in northern California into the 1980s, is more laid-back, more willing to dream about the social potential for computers outside the lab. Like most such clever but broad comparisons, it’s ridiculously reductive.

Yet it also may contain more than a few grains of truth. For all that they enjoyed riffing on the Zork milieu with its grues and its Flathead dynasty, amongst the early Implementors only Dave Lebling read much fantasy literature — and that was because Lebling, an omnivorous and voracious reader then as he remains now, read a lot of everything. If there was a consensus literary genre of choice amongst this group, it was science fiction. You can see this clearly by looking at the string of games Infocom released between the fall of 1982 and the summer of 1983. At this stage, with the company ramping up quickly but with a structured marketing department not yet in place to tell the Imps what kind of games they needed to make to fill in empty spaces in a matrix of genres, everyone just wrote the game he wanted to write. The result was that out of five games by five different authors three were science fiction.

Still, Infocom remained the house that Zork had built. To not continue that series, to ignore the fantasy genre that still remained (as it still does today) the preferred genre of the gaming public at large, would have been crazy. Having lived with the idea of an “original Zork trilogy” for so long, it can be surprising and even a bit counterintuitive for us to recognize that neither Infocom nor their customers saw Zork in that way when the original three games were being written and released. As far as they were concerned Zork was an open-ended series of numbered games of the sort that Ultima and Wizardry would become. Nowhere is that made clearer than in Zork III itself. Here Marc Blank, having incorporated bits of Zork I and Zork II into what stands today as the first of an eventual several brilliant Infocom time-travel puzzles, added an additional little Easter egg: a preview of the as-yet unwritten Zork IV in the form of a grisly episode in which the player gets sacrificed by an evil priest of some sort.

Thus, for all their high-brow write-ups in the New York Times Book Review and the pushes they had made into new literary genres and new styles of play, Infocom needed during 1983 to deliver another good old traditional Zork game — and one that incorporated, Mad Libs-style, Blank’s ugly sacrifice scene — even if it felt like something of a step back. Problem was, it wasn’t clear where to go next with Zork. It may not have been consciously designed as the climax of a trilogy, but Zork III did nevertheless have an air of finality about it. At its end the player had completed her existential journey by becoming the being she had spent all three games struggling against, the Dungeon Master. What could follow that?

The game that they eventually created is a testament to Infocom’s skill at balancing artistic credibility with commercial considerations. It began when Lebling, looking for a reason to get excited about a Zork IV, started thinking back to the ending of his previous Zork game, Zork II. There the player, after vanquishing her irritating nemesis the Wizard of Frobozz, could claim his magic wand and try a few spells for herself. It made a relatively tiny part of the game, and not a terribly deeply-implemented part at that, but it was just such an intrinsically cool idea; you just knew Lebling was onto something here that deserved further pursuit. Lebling, the only Implementor with any grounding in Dungeons and Dragons, now worked up an almost D&D-like magic system for Zork IV. Such adaptations from the world of tabletop RPGs were one of Lebling’s tics as a designer; he was, you may remember, also responsible for the little-loved randomized combat in Zork I.

Fortunately, the magic system he now created is much more fondly remembered. You carry a spell book containing a few beginning spells. Over the course of the game you can collect more spells on scrolls, most of which you can inscribe into your spell book, thus becoming an ever more flexible and formidable magic user. Prior to casting a spell you have to “memorize” it (or load it into your head like a piece of ammunition), just like in D&D. Once cast, a given spell is gone from memory until memorized again. And there is, of course, a limitation to the number of spells you can have in your memory at once.

All told, the magic system was an absolutely brilliant addition to an otherwise standard text-adventure template. Collecting spells and using them proved to just be fun as all get-out. Removing so many puzzles from the realm of the mechanical to that of the arcane even hid many of the implementational seams that usually showed through; when stuck, the player tended to spend her time casting her spells at various objects, a more manageable set of possibilities to deal with than having her try all sorts of crazy physical manipulations. Indeed, Lebling and his co-author, the indefatigable Marc Blank, quickly realized that seeing their spells fail was almost as much fun to players as using them to solve puzzles. Lebling and Blank therefore spent a lot of effort to make sure that, say, casting Nitfol (“converse with beasts in their own tongue”) on any creature in the game got you something appropriate — and usually entertaining — back in return.

At some point fairly early in the new game’s development Lebling and Blank decided that the addition of magic made it feel so qualitatively different from what had come before that releasing it as Zork IV just didn’t feel right. Further, in these heady days when they were being touted as pioneers of a new interactive literature, they were eager to live up to their billing, to demonstrate a certain eclecticism and literary integrity rather than just continuing to crank out the Zork games. They therefore made the brave decision to rename the game Enchanter, first of a new, open-ended series of fantasy games with an emphasis on spellcraft. (As with Zork, Infocom wouldn’t definitively decide this series should be a trilogy until much later.) Having declared their artistic independence, Infocom could then temper things a bit by declaring the new series to be “in the Zork tradition” and by including plenty of callbacks within the game to make it clear that, while this may have been a new series, it took place in the same beloved fantasy world. Thus they thought they could have their cake and eat it too — and in this they were partially if (as we shall see) perhaps not entirely correct.

As Enchanter begins an evil warlock by the name of Krill has been growing in power, and now threatens to conquer the entire world. The Circle of Enchanters was not initially sure how to respond. To send one of their own number to fight Krill would be “ill-omened,” for Krill would sense the intruder’s magical aura as soon as he entered his stronghold and send his minions to destroy him. Therefore, borrowing a plot element from The Lord of the Rings that would subsequently be used by a thousand CRPGs to explain just why your party of first-level nobodies are entrusted with saving the world, they have decided to send you, a “novice Enchanter with but a few simple spells in your book,” instead. They teleport you onto a deserted road close to Krill’s stronghold, and the game begins.

Enchanter‘s structure feels very old school when contrasted with the handful of Infocom games that preceded it. Not only is it a very traditional game, lacking the radical formal experimentation of the mysteries and Suspended, but it lacks even the initial narrative thrust of Starcross and Planetfall. Both of those games opened with a dynamic scene to get the plot wheels cranking and set up the non-linear exploration of the long middle. Enchanter, however, simply plops you down in an expansive world and tells you to get started with mapping, collecting objects and spells, and solving puzzles, just like Zork I.

Some of the first puzzles you encounter, before you even get into the castle, involve collecting food and drink. Like Planetfall, Enchanter is the product of a very brief era when Infocom was suddenly enamored with the idea of requiring the player to deal with these necessities. In fact, it’s even more stringent than Planetfall in this respect, implementing eating and drinking as two separate necessities in addition to the need for sleep. Hunger and sleep timers would soon become passé at Infocom (not to mention since Infocom’s era) as pointless annoyances that add little to the games into which they’re shoehorned. Yet, as in Planetfall, they don’t bother me greatly here, and even manage to feel somehow organic to the experience. When you sleep your dreams even deliver vital clues.

Once you get inside Krill’s stronghold you find a brilliant collection of interlocking puzzles that are challenging but solvable. Even better are little touches of wit and whimsy that abound everywhere, a sign of Dave Lebling really coming into his own as an author. Although Enchanter is credited as a joint production of Blank and Lebling, it feels like there is a lot more of the loquacious, playful Lebling than the terser, more stoic Blank here. Indeed, for being yet another struggle of Good vs. Ultimate Evil Enchanter has a remarkably light tone, with only a few discordant touches — most notably the sacrifice scene previously advertised in Zork III, which seems dropped in from another game entirely for the very good reason that it was — to remind you of the stakes. Let me tell you about a few bits that particularly delight me.

On the beach just outside the castle we meet the most prominent of a few animals in the game, a turtle, “his enamelled shell shining with all the colors of the rainbow.” When we dutifully cast Nitfol on him we learn how his shell got that way:

"How do you like my shell? A wizard did that to me about 75 years ago. It's nice to find a human who talks turtle. Not many do, you know. Most people think turtles are boring, just because we talk slowly."

Our new friend turns out to be a droll but helpful old fellow whom I find just about as charming as Planetfall‘s Floyd in yet vastly less space:

"Are you a magician? Are you going to do something about that annoying Warlock, then?"

The turtle is the centerpiece of a puzzle that is superficially similar to the one that required us to order a robot about in Zork II, the first Infocom game that allowed us to talk and give orders to others. This time it’s much more fun, however, because, well, it’s our turtle friend who’s helping us rather than a personality-deprived robot. We just need to speed him up before we get started, which we can accomplish with a touch of magic. When his task is finished:

The turtle drops a brittle scroll at your feet. "Not bad, huh?"

I’ve always loved this little guy, as has Lebling; he lists him as one of his favorite creations. The turtle and a few other creatures, all accessible to us thanks to the Nitfol spell, bring life to Enchanter, pulling it a million miles from the windy solitude of Zork III.

But the most remembered character of all in Enchanter is actually you — not the you who is playing the game now, but the you who dutifully marched through the three Zork games to get here. In one area of the castle we find a “Hall of Mirrors,” behind which lies a dim underground labyrinth. In it we occasionally catch a glimpse of “a bedraggled adventurer, carrying a brass lantern and an elvish sword, which is glowing dimly.” He is, of course, our old avatar from Zork. We can use our magic to summon him to the castle.

All at once, the bedraggled adventurer appears before you, brightly glowing sword in hand. His jaw has dropped and his eyes are bulging. His eyes dart this way and that, as if looking for a way to escape.

The game then proceeds to mercilessly but affectionately lampoon this rather dim fellow, along with the old-school design tropes he represents. By far his biggest interest is in collecting valuable objects to put in the trophy case he presumably has back in his white house:

The adventurer offers to relieve you of some of your possessions.

The adventurer asks what you would be needing treasures for.

The adventurer, not overly tactful, asks what you're holding.

In effect we’re seeing the adventurer as the troll, the thief, and their buddies in Zork I must have seen him (us?). He wanders about snarfing every object that isn’t nailed down, fiddling constantly with a weird map (“a convoluted collection of lines, arrows, and boxes”), and serving as an extended in-joke to anyone who spent any time with the Zork games.

The adventurer tries to make some small talk, but only mumbles. He'll have to speak up if he expects you to hear him.

The adventurer waves his sword menacingly in your direction.

The adventurer stares at his possessions as if expecting a revelation.

The adventurer seems to have dropped out of existence. In a voice that seems to recede into the void, you hear his final word: "Restore...." You muse about how a mere adventurer might come to possess a spell of such power.

The adventurer smiles at you like an idiot.

The adventurer asks for directions to Flood Control Dam #3.

The adventurer stops and stares at the portraits. "I've met him!" he gasps, pointing at the Wizard of Frobozz. He doesn't appear eager to meet him again, though. "And there's old Flathead! What a sight!" He glances at the other portraits briefly and then re-checks his map.

The adventurer waves at you and asks "Hello, Sailor?" Strange, you've never even been to sea.

In the spirit of shoe-on-the-other-foot, he also proves annoying in the way many of the non-player characters within the Zork games were, scattering objects hither and yon so you never know just where anything is.

At the risk of ruining a great joke by making of it grist for some theoretical mill, it’s remarkable that Infocom is already playing with the clichés and expectations of the adventure-game form so early, just six years after Adventure itself. This sort of knowing self-referentiality is a very modern phenomenon, one that appeared only after decades or centuries in other art forms. It’s the sort of thing I want to point to when I say that Infocom was more knowing, more sophisticated — just a little bit smarter — about what they were doing than their peers. And yet Infocom is doing it from within what is ultimately a very old-school design of its own, a perfect example of their talent for giving the people what they want, but doing it with a grace and style that eluded most of their competitors.

Enchanter would make an ideal case study in gated puzzle design. Its wide-open map conceals several intricate chains of puzzle dependencies that give the game a structure that Zork, with its mostly unrelated puzzles strewn randomly about its geography, lacked. The adventurer, annoying as he can be, is also a critical link in one of these chains. He gives us our key for solving the “maze.”

A certain fascination with pseudo-mazes is another of Lebling’s design tics, one which he also passed to Steve Meretzky. He claims to have lost interest in the standard approach to mazes even before his friends at MIT added a couple of monstrously cruel examples of the form to the original PDP-10 Zork. What he delighted in instead was to give us areas that seem to be mazes, but which have some trick — other than the tried-and-true dropping of objects and plotting connections, that is — to solving them. His first pseudo-maze, the baseball puzzle in Zork II, misfired horribly. His second attempt in Starcross was much more reasonable, a labyrinth that could be solved only by convincing someone else to guide you. His third attempt is here in Enchanter in the form of the “Translucent Rooms,” and it’s even more clever. I’m going to spoil here its concept, although not the mechanics of its solution, as an illustration of the marvelous and varied puzzle design inside Enchanter.

So, with the adventurer’s aid we come upon a map which we quickly realize shows the Translucent Rooms.

The map consists of a drawing with nine points, each represented by a strange character, with interconnecting thin pencil lines. Using your native alphabet, it looks like this:

B       J
!      / \
!     /   \
!    /     \
!   K       V
!          / \
!         /   \
!        /     \
R-------M       F
 \     /        
  \   /        
   \ /        
    H       P

 

We also find a magic pencil, using which we can draw in new connections between rooms and also erase them. When we do so, the connections appear not only on the (paper) map but also within the real-life maze. The catch, however — there’s always a catch — is that we have enough lead left to draw just two lines, and enough eraser left to erase just two. That shouldn’t be any problem, right? As you’ve probably guessed, the currently inaccessible room at P contains the item — a powerful spell we can use to banish Krill to “another plane of existence” — that is the point of this whole exercise. Unfortunately, it also contains a powerful entity of eternal Evil who makes old Krill look like a pussycat in comparison. We glean from a book found elsewhere in the game that he was banished there many centuries ago by our magic-using ancestors to save the world (evidently this world of ours tends to need a lot of saving). As soon as we give the entity an escape route to the exit, room B on the map, he’ll start moving toward it. When he’s in a room with us, meanwhile, we’re too terrified to do anything at all. So, the puzzle is to lure the entity out of room P, but to shut off his escape route before he gets all the way out while ourselves getting into room P and then out of the maze — all without using more than two pencil strokes and two erases.

Even in 1983, when adventure-game engines from other companies were beginning to make technological strides, Infocom was the only company who could have made such an intricate, dynamic puzzle with the associated necessity for a parser capable of understanding the likes of “draw line from H to P.” I’ve made this point before, but it’s worth stating again that Infocom’s parser was not just a wonderful luxury; it enabled better puzzles, better game design. This puzzle is a good example of the sort found throughout the game, being fair, challenging but not exasperating, and built with some intricate programming that, like all the best intricate programming, is likely to go completely unremarked by the player; it just works.

Lest I be accused of overpraising, let me also note here that Enchanter is a product of 1983, and does show some signs of its age. In addition to hunger, thirst, and sleep timers (the first of which gives a hard limit to the time you can spend in the game, since there is only so much food to eat), there is an inventory limit. And there’s a fair amount of learning by death. Whatever you do, don’t get the bright (ha!) idea of casting the Frotz spell on yourself so as to have a constant source of light; since there is no way to extinguish this spell and since one puzzle is dependent on darkness, you’ll lock yourself out of victory thereby. Worse, you’ll probably have no idea why you can’t proceed, and when you finally break down and turn to the hints will throw the game against the (metaphorical) wall and hate it forever. The big climax is another offender in this department, although one less likely to force you to replay large swathes of the game. You have only seconds to defeat Krill and the minions he throws at you, and no idea which spells you need to have memorized to do so without dying a few times to gather that information. But other than its past-lives issues in this and a few other places, Enchanter plays very fair. Just remember, as a wise man once said, to save early and often.

It’s probably safe to say that Infocom’s decision to make Enchanter its own thing had commercial consequences. It sold reasonably well, but lagged behind the older Zork games. Released in September of 1983, it sold just over 19,000 copies before the end of that year, followed by a little over 31,000 copies the following year. Enchanter did prove to have longer legs than many older Infocom titles in the company’s later years. All told, it sold over 75,000 copies as a standalone game or as a part of the Enchanter Trilogy bundle. Today it stands as one of the more fondly remembered of Infocom’s games, with more than its fair share of appearances on favorites lists, and has served as the template for some well-regarded games of more modern vintage. Its individual spells, meanwhile, have taken on a life of their own within modern IF circles, being used as the names of interpreters and various other programs and bits of technology — not to mention the name of the domain on which you’re reading this. As my choice of domains may indicate, Enchanter is in my personal top five or so of Infocom games, the first I’ve come to on this blog about which I can say that. Unlike my other favorites, which tend to push the envelope of what a text adventure can be in one way or another, Enchanter stands for me almost as a platonic ideal of an old-school, traditional adventure game, executed with thoroughgoing charm and craftsmanship. I love it dearly.

 

Tags: , , , ,