Monthly Archives: July 2012

One of the computers installed at Monthly ASCII‘s offices was a DEC PDP-11 minicomputer. Amongst other programs, this machine housed many of the games popular with Western institutional hackers of the early 1980s, such as Adventure and Zork. Many staffers played these games obsessively before and after working hours — and occasionally during them — in spite of the challenges their English language presented. ASCII also owned some imported Western PCs, along with early adventure games from Scott Adams, On-Line, and Infocom. With such games still unknown on homegrown Japanese machines, some staffers naturally began thinking about remedying that by writing a text adventure of their own. By this time various Western magazines had examined the technology used to develop professional text adventures. In addition to the many type-in BASIC adventures, there were articles like the one that appeared in Practical Computing in August of 1980, describing a machine-language text-adventure engine similar to the one used by Scott Adams.

With all this information at hand, two ASCII employees, Hideki Akiyama and Suguho Takahashi, proposed that the magazine develop and publish a text adventure for the 1982 Yearly Ah-SKI! parody issue. With permission granted, Akiyama developed an adventuring engine, and Takahashi wrote the first scenario for the system: Omotesando Adventure, after the street in Tokyo where ASCII‘s offices were located. It was a difficult project, particularly as both men could only work on it during down-time from their regular jobs on the magazine’s editorial staff. Still, as the first adventure made in Japan, both were determined that it should not be “shoddily made.” Omotesando was, in the words of Takahashi, a “pretty outrageous idea” for one reason in particular: despite being published in a Japanese magazine for a Japanese audience and running on Japanese computers, it was actually written in English.

Amongst the languages of the world, English, with its simple verb-object imperative construction, is rather unusually well-suited to a parser-driven adventure game. Even a related language like German makes coding a parser more complex through its many two-part verbs and its fondness for reflexive pronouns. Japanese, meanwhile, is still vastly less amenable to traditional parsing algorithms. As my translation and research partner for this article, Oren Ronen, told me, “Japanese doesn’t lend itself easily to the verb-noun pattern without sounding completely broken.” I suppose that ASCII could have had Omotesando Adventure output text in Japanese and accept input in English, but that would have created huge problems of its own, as legions of imperfect English speakers tried to figure out how to reference this or that Japanese word in the text in English. So, pure English it was — and luckily so for those of us who don’t know a lick of Japanese, as it gives us a chance to peek inside this important artifact.

Omotesando was developed on a Japanese Oki IF-800 computer running the English operating system CP/M. This machine was chosen because it wasn’t a particularly popular one with the other staffers, and thus normally spent its time gathering dust in a corner of ASCII‘s offices. When development was complete, Akiyama needed only write some bridge code for the various other Japanese Z80-based PCs, such as the very popular PC-8001 and PC-8801. With a little help from some other staffers, he soon had Omotesando running on a substantial percentage of Japanese PCs.

ASCII introduced Omotesando excitedly as a whole new experience for Japanese computer users:

Among the various things we can use a personal computer for, games are the most constant and popular. Games are in high demand, and software houses are pushing them into the market continuously. This issue we are introducing the Adventure Game. It’s an entirely new genre, the like of which was never seen on a computer. We may even call it a “New Type” of computer games.

The premise has you an employee of an unnamed rival magazine to ASCII. You’ve been ordered to sneak into ASCII‘s offices and find some way to sabotage the operation. It’s a nice change of pace from the dragons and spaceships that still dominated American text adventures, and there’s a certain postmodern sort of cleverness to all of the self-referentiality, a precursor to the generations of amateur bedroom text-adventure implementers who would get their start by implementing, well, their own bedrooms and/or apartments.

As one might expect, Omotesando is proudly old school in its design sensibilities.

Let’s state this clearly: we don’t believe you, who are experiencing an adventure for the first time, will be able to easily achieve the game’s goals. For several dozen tries you probably won’t even manage to sneak into ASCII Publishing’s offices. You will be caught in traps and die frustrating deaths. By trying again and again dozens or hundreds of times, you’ll learn things like “if I do this here I can go through” or “before I do this I must not do that” by trial and error and will be able to proceed. Solving the riddle that is the entire game can take several months and try your endurance.

In any case, since this is a game of endurance, we added the ability to save your position to a cassette and reload the game to continue from the same point. We’re very kind.

That said, to stay in the spirit of the adventure game, you must look for the way to save the game in the course of playing it (we’re not always kind). Once you find the way, you will be able to solve the game’s puzzles much more quickly.

As the illustration above shows, the English of the game is far, far from perfect. The same quirky sense of humor that marks the Yearly Ah-SKI! editorial content is as present as the sharp limitations on text length will allow. There is also the unusual (and annoying) “feature” of having to LOOK every time you enter a new room to determine what is really there. Other than that, and of course its setting, there’s oddly little about Omotesando to separate it from its inspirations. Perhaps the most surprising thing about it is that it is quite technically and even artistically competent within its modest aspirations. If you’d like to experience it for yourself, I’ve prepared a little care package for you. It includes Omotesando packaged up with a Japanese PC-8001 emulator as a Windows executable, along with translations and originals of the magazine articles that accompanied it. Sorry to be so platform-specific this time. Given the trials and tribulations of Japanese, this was about the best I could do. I do believe it should also run fine under WINE.

Omotesando Adventure, then, marks the beginning of the text adventure’s brief, precarious existence in Japan. The following year’s Yearly Ah-SKI! brought another, more complex game, once again implemented in English. By that time a Japanese company called Starcraft had begun a project to translate some of the more popular American illustrated text adventures for the domestic market, including most of On-Line’s Hi-Res Adventure line (with the expected exception of the white elephant Time Zone). Still, the difficulty of parsing Japanese made these games an awkward fit for the country; some actually resorted to making the player type a Japanese verb, hit enter, and then input a noun as the least painful and grammatically ugly approach.

As an alternative to such kludges, amateurs and professionals alike soon began experimenting with menu-driven interfaces in place of parsers. Thus was born the tradition of the Japanese visual novel, which became extraordinarily popular in spite of or because of the fact that many were partly or entirely exercises in pornography. Whatever objections some might have to their subject matter, or frustrations with their limited scope for player choice, these games did at least depart definitively from text adventures’ obsessions with puzzles and low-level object interactions to focus on telling more complex stories at a less granular level than most of their peers in the West. In fact, visual novels and the related genre of dating simulations remain important to Japanese computing culture even today, decades after text adventures faded from store shelves in the West.

A much more successful Western genre in Japan in comparison to the text adventure was the CRPG. Wizardry and its sequels in particular were massively popular when translated into Japanese. Just as the Beatles and other British groups of the 1960s mirrored back to America the music that America had originally created, Japanese CRPGs eventually made their way back to the West, where their unique, heavily story-oriented sensibility made them more popular with many players than the more hardcore, stats-oriented Western games.

Indeed, I’m afraid this blog will necessarily have to view Japan mostly through the lens of the West, through the games that made it back to these shores in English translations. I don’t know any Japanese, you see, and, while Oren did amazing work to help me with this article, I can’t expect him to keep doing my research and translating huge chunks of material for me. So, take this article as primer only on what was going on inside a country that would soon become hugely important to Western gaming culture. But before I move on, I’ll leave you with one last historical curiosity from inside Japan that Oren dug up for us.

During the early 1990s, after parser-driven games were considered commercially dead not only in Japan but also in the West, four of Infocom’s classic titles were ported to Japanese computers, and translated into Japanese in the process. Let me just let Oren finish this story:

There were 4 translations: Zork 1 is the only one that’s mentioned anywhere on the western web, but that’s only because it got a strange console release on the PlayStation and Sega Saturn. Most references get the facts wrong by saying those were the only releases it got. In fact, there was also a PC release with a proper text-entry based parser. In addition, Enchanter, Planetfall and (of all things) Moonmist got translations. All were wrapped in a fancy, Zork Zero-like interface with a changing graphical background depending on your location, a clickable compass-rose, and menus for saving and loading the game. They were released in the early ’90s by a software company called SystemSoft that mostly did game localizations for the Japanese market. They all mention Activision in the documentation, so the deal was made after Infocom’s buyout. The translations themselves are pretty good, as far as my non-native literary appreciation skills can tell.

The most remarkable thing about them is that they are the best implementation I can find anywhere of an IF parser in Japanese. Because of the complexity of parsing Japanese text, adventure games moved to menu-based systems much faster than in the West, and the few parser-based games that were released either required input in English or did a no-frills two word Japanese parser that was completely ungrammatical. The Infocom games parse complete, grammatical Japanese sentences (which is, again, a much more difficult task than in English) and even do all the fancy Infocom tricks like “take all except the stone” and remembering context for pronouns. I’m quite amazed they went through all the effort several years after parser-based games were completely forgotten about in the country.

The translations even reproduced the feelies from the original American releases. Alas, it appears they were, unsurprisingly, not successful, and they mark the last gasp of the text adventure’s short and fitful life in Japan.

(My thanks again to Oren Ronen for all his help researching and translating for this article and the preceding one, which included translating a short interview with the designers of Omotesando Adventure.)

There were two possibilities that kept those who built the PC industry in the United States and Britain awake at night, both all the more terrifying in that they seemed all but inevitable. One was the arrival of the sleeping giant that was IBM. The pioneers could easily imagine Big Blue bludgeoning its way into the industry with billions of dollars behind them and a whole slate of “standards” of its own, destroying everything they had built. This was, after all, exactly what IBM had done years before to the big mainframe market which it now all but owned for itself. When Big Blue finally came to the PC, however, it showed unprecedented eagerness to partner with already extant PC firms like Microsoft. IBM’s entrance in the fall of 1981 did eventually result in one, near universal computing standard, but the process of sweeping out what the pioneers had built would end up taking more than ten years to come to completion — and, most shockingly, neither the hardware nor the software standard that resulted would remain under IBM’s control. This gave companies, at least those able to see which way the winds were blowing, plenty of time to adapt to the new computing order. (Or, in the case of Apple, to stake out their own territory as the anti-IBM.) Yet even as IBM’s entrance proved nowhere near as bad as it might have been, everyone still waited for the other shoe to drop: Japan.

Many people who know much more about such things than I have analyzed Japan’s postwar economic miracle. Suffice to say that the devastated remains of the country after World War II rose again, and more quickly than the most optimistic could have predicted. Already in 1968 Japan became the world’s second largest economy, its explosive growth initially fueled largely by exports of steel and textiles and other heavy industry. It was at about this point that a new form of Japanese expansionism made the country again a source of concern rather than pride for the United States: a new wave of Japanese consumer products began reaching American shores, most notably consumer electronics and cars. Today we remember mostly the latter. The products of Nissan, Toyota, and Honda devastated the U.S. auto industry, and for good reason. They were cheaper, better built, better engineered, safer, and, very important in the wake of the oil crisis, much more fuel efficient than American cars of the era. Yet the American auto industry did survive in a damaged state, and in time even learned to compete again. Huge swathes of the domestic electronics industry had no such luck. Japan didn’t just damage American consumer-electronics companies, they effectively destroyed them. Already by 1982 it was getting hard to find a television or stereo made in America, and a decade later it would be, a few high-end boutique brands aside, impossible. And even as they took over the old, Japan also pioneered new technologies, such as the iPod of the early 1980s, the Sony Walkman.

In one of those fun paradoxes of capitalism, Americans fueled this expansion by buying Japanese products even as they also lived in increasing fear of this new economic menace. Television advertisements from the Detroit auto manufacturers were filled with slurs against Japan that veered uncomfortably close to unabashed racism, while newspapers and magazines were filled with awed exposés of Japanese workplace culture, which came across a bit like the Borg of Star Trek fame. The average Japanese, people were told, placed loyalty to his employer over loyalty to his family or his country; forewent hobbies, friends, vacations, and family life in favor of work; began each 12-hour-plus working day by singing company songs and doing calisthenics with his coworkers; demanded only the most modest of salaries; considered an employment contract a lifetime through-good-or-ill commitment akin to a marriage vow. Add to this the complete, threatening otherness of Japanese society, with its insular artistic culture and its famously difficult language. The long and painful recession of the early 1980s only increased the sense of dread and paranoia. An earlier generation of Japanese had been willing to fly their planes deliberately into enemy ships in the service of an obviously lost cause, for God’s sake. How could American employees compete with that kind of commitment? What industry would the Japanese be coming for next time — and was there even any point in trying to stop them when they did?

The burgeoning Western PC industry seemed like it should be right in Japan’s wheelhouse. After all, this was a country that was cranking out well-trained electrical engineers at the rate of 40,000 per year by 1982, and that increasingly enjoyed global domination of most other types of electronics. To the paranoid and the conspiracy-minded, the PC industry seemed ripe to follow the pattern set in those other sectors. Japan would let Americans do the hard work of development and innovation, and then, when the technology was mature enough to be practical for the average home and business, sweep in with an avalanche of cheap PCs to make the industry their own. This vision of Japan is of course a confused one, especially in insisting on seeing the entire Japanese economy as a single, united front; in reality it was built, like the American economy, from many individual companies who competed with one another at least as hard as they competed with the West. Still, that’s how it looked to some on the outside, especially after the arrival of Space Invaders, followed by a steady deluge of other standup arcade games, showed that Japan was no slouch when it came to computer technology.

A couple of people in the PC industry worried about a Japanese-driven apocalypse even more than most, for the very good reason that it had already happened to them once: Jack Tramiel of Commodore and, in Britain, Clive Sinclair. Both had seen their companies driven to the brink of bankruptcy in the previous decade, when low-priced Japanese models drove them out of previously profitable gigs as calculator manufacturers in a matter of months while they could only stand around wondering what had just hit them. Now, both were determined to make sure it didn’t happen to them again in their new roles as PC purveyors.

Commodore caused increasing chaos in the American PC industry of the early 1980s by selling first the VIC-20 and then the Commodore 64 at ever lower prices. Most attributed this to Tramiel’s maniacal determination to dominate at any cost, as summed up by his famous motto, “Business is war.” Yet an at least equally important source of his price-cutting mania was his fear of the Japanese, his fear that they would come in and undercut everyone else if he didn’t do it first. As he said in another of those little aphorisms he was so good at coining, “We will become the Japanese.” His competitors who cursed him as the ruthless SOB he was likely also had him to thank that the doomsday scenario of a Japanese takeover never came to pass. Japan would certainly play its role as a source for chips and other components, and companies like Epson and Okidata became huge in the printer market, but the core of the American computer industry would remain largely American for many years. With Commodore already dominating the home market with its low-priced but reasonably solid and usable machines, Japanese firms didn’t have the opportunities to upstage and undercut the competition that they did in other industries.

In a turnabout play that he must have greatly enjoyed, Tramiel actually opened his war with Japan on the enemy’s home front. Six months before introducing the VIC-20 to North America and Europe in the spring of 1981, Commodore debuted it in Japan as the VIC-1001. It became a big success there, although its limitations ultimately ensured it was a fairly short-lived one, just like in North America and Europe. Commodore would try to bring its later machines to Japan as well, but never got the same traction that it had with the VIC. Instead, Japan’s PC market remained peculiarly, stubbornly insular throughout the 1980s. If Japanese computers never invaded the West the way so many had feared, Western manufacturers also never had much luck in Japan. They were partly kept at a distance, as had been so many Westerners before them, by the challenges of the Japanese language itself.

Japanese is generally acknowledged to have the most complicated system of writing in the world. It’s comprised of three different traditional scripts, often interleaved in modern times with the occasional foreign phrase written using Latin characters. One script is kanji, a script borrowed from Chinese with thousands of characters, each of which represents a concept rather than a phonetic sound. In addition, there are two syllabaries that can be used to write the language phonetically: hiragana and katakana. Including about 50 characters each, they are simply different scripts for the same phonetic values, similar, but not quite equivalent, to upper-case and lower-case letters in Western languages. Normal Japanese text is written using all three intermixed: kanji is used for the stems of content words; hiragana for suffixes, prepositions and function words; and katakana for a variety of other tasks, most notably to write loan-words from other languages and to add emphasis to a certain point. To complete this confused picture, and as anyone who has browsed Japanese videogame boxes can attest, many Western proper names are simply written in their original form, using Latin characters.

All of this amounts to a nightmare for anyone trying to make a computer input and output Japanese. Computers have traditionally stored text using a neat arrangement of one byte per character, which gives the possibility of having up to 256 separate glyphs. That’s more than enough not only for every English character but also for the various accents, umlauts, ligatures, and the like found in many non-English Western European languages. For Japanese, however, it’s sadly inadequate. Then there’s the problem of designing a keyboard to input all those characters. And there’s an additional, more subtle problem: with so many characters, each Japanese glyph must be quite intricate in comparison to the simpler abstracts that are Latin letters, and the differences between glyphs are often quite small. The blocky displays of the early PC era were incapable of rendering such complexity in a readable way.

Faced with such challenges, which made localizing their existing designs for Japan all but impossible, most Western companies largely chose to leave the Japanese market to Japanese companies. Although the first PCs available for purchase in Japan came from Apple and Radio Shack, it didn’t take Japanese companies long to jump in and take over. Hitachi and NEC introduced the first pre-assembled PCs to be built in Japan in 1979. By 1982, Japanese manufacturers controlled 75 percent of the Japanese market, which was dominated by three machines: the NEC PC-8801, the Sharp X1, and the Fujitsu FM-7. None completely solved the fundamental problems of displaying Japanese. In addition to the standard English glyphs, each could display only katakana, chosen because its characters are square and easy to make out even on a low-resolution display, and perhaps a handful of kanji pictographs, for extremely common concepts like “date” and “time.” Katakana, on its own, is fine for short sentences or simple messages, but reading and writing longer passages in it alone is a difficult, wearisome task. The machines had to be commanded, meanwhile, in good old English-derived BASIC (usually sourced from Microsoft), a particular challenge in a country large and insular enough that a good knowledge of English, especially at this time, was a fairly rare trait.

In light of these tribulations, Japanese companies worked feverishly to develop hardware capable of storing and displaying proper Japanese writing. In 1982 NEC introduced the first machine that could do that reasonably painlessly in the form of the PC-9801, an advanced 16-bit computer built around the Intel 8086 processor. It soon ruled the Japanese business market despite being incompatible with the IBM PC that dominated business computing elsewhere. By the time it was finally discontinued in 1997, the PC-9801 line had sold well over 15 million units, virtually all within Japan. The need to display kanji characters and compete with NEC drove other manufacturers, with the happy byproduct that most Japanese computers soon had very advanced display capabilities in comparison to most Western machines.

The PC-9801 and its competitors clearly pointed the way forward for computing in Japan, but they were initially expensive machines targeted very much at business. Home users would remain stuck for at least another couple of years with their more primitive 8-bit machines built around the Zilog Z-80 or Motorola 6809. Just like in the United States, these early adopters built a distribution network for games written by amateurs and semi-amateurs, called “doujinsoft” in Japan. Again like in the United States, a common early means of distribution was through program listings printed in the enthusiast magazines. The first and most respected of these, sort of the Japanese Byte, began in 1977 and was called Monthly ASCII. In 1981 the magazine began to publish an annual supplement for April Fool’s Day, called Yearly Ah-SKI! It mostly consisted of fake articles and advertisements mocking the computer industry, but included one full, working program as well each year. The second Yearly Ah-SKI!, in 1982, hosted what was almost certainly the first text adventure, first adventure game, and first full-fledged digital ludic narrative to be published in Japan. We’ll look at that next time.