Retro Japanese Computers: Gaming's Final Frontier

Thanks to the Internet, nearly any game of the past can be downloaded and emulated, and almost every piece of information has been documented somewhere... except perhaps the world of Japanese home computers, arguably the last uncharted frontier for English-speaking games enthusiasts.

Japan has long been viewed by the West as a console-centric country, ever since Nintendo and the NES. But there is another, mostly forgotten world of Japanese gaming history, in which thousands of games were developed for various Japanese computers over an 18 year period that stretches from the late 1970s to the mid-1990s. For all that Nintendo started, it was the open hardware of NEC and other companies that allowed small groups to form and become giants. In fact, some of Japan's most recognizable franchises, such as Metal Gear and Ys, actually began as computer games. The early Japanese computing scene was an intense flurry of creativity that launched the careers of many prominent figures in the video game industry, while also establishing some of the most famous video game companies, such as Square, Enix, Falcom, and Koei.

Japanese computer games were also exempt from any of the licensing and content restrictions that all console makers have imposed in various forms. These early games give us a rare glimpse into a world of Japanese creativity unfettered by censorship and outside pressures, which has never since been replicated. The content ranges from rampant drug use and presidential assassination (XZR), to tender explorations of love, sex, and relationships (Dokyusei), to mature and suspenseful horror (Onryo Senki), and even to one of the first rape simulators (177), predating the infamous RapeLay by 20 years. The content is not always tasteful, but the lawless atmosphere resulted in some of the most unique titles in video game history.

Unfortunately, this important part of gaming history has been largely obscured by time and language barriers. If you want to play these games you have to work at it, since they're not easy to find, and rarely described in languages other than Japanese. Until now, most English-language coverage has focused only on the "eroge" (erotic games, or hentai). Erotic games constitute a large and important fraction of the Japanese computer game library, and some are definitely worth your time if you are comfortable with the content, but there are hundreds of other underappreciated titles in almost every genre.

But finding these by games by searching the internet with romanised titles often results in nothing, while using the original names will only bring up Japanese websites which Babelfish renders into lunatic gibberish. Even downloading complete file archives, which are always missing titles due to a lack of definitive listings, will often present you with folders in Japanese characters (assuming your computer can even display them). Finding a good game can become an annoying case of trial and error.

But nothing compares to discovering the holy grail of hearing music and seeing sights few others have, and clicking in that Saturn USB pad for some of the best gaming of your life (and anyone who doesn't own a Saturn USB pad for emulation should be ashamed of themselves). Some narrow-minded purists will frown on the idea of emulating these titles, insisting you should either buy the original hardware or go without. But don't let these sad, misguided fools put you off - with Windows now having a respectable range of highly compatible emulators for just about every Japanese computer that existed, there's little need to struggle with archaic and difficult-to-acquire hardware. Furthermore, emulators benefit from a range of swanky features, such as save states and frameskipping, which are invaluable given how flaky some old computers can be.

A forgotten era

The personal computer industry in Japan began much like everywhere else: as a response to Intel's creation of the world's first microprocessor, the 4004, in 1971. Both NEC and Toshiba successfully developed their own microprocessors in 1973, and over the next few years a number of personal computer kits and homebew packages were released by companies such as Hitachi, Fujitsu, NEC, Toshiba, and Sharp. Much like in the West, these early computers were primarily for electronics tinkerers and enthusiasts, and had to be programmed by the users themselves.

The early 80s saw the release of the first fully-fledged 8-bit computers designed with average users in mind, rather than amateur programmers (although there were still plenty of those). Three companies eventually shared the 8-bit crown: NEC, with its PC-8800 series; Fujitsu, with the popular FM-7; and Sharp, with the X1. NEC would later come to dominate the Japanese computing scene for over 10 years with another computer, the 16-bit PC-9801, but Fujitsu and Sharp were able to maintain a small but loyal following by staying competitive and eventually releasing two incredible 16-bit machines of their own: the Fujitsu FM Towns, and the Sharp X68000.

One important thing to note is that much like the early computers produced by Western companies such as Apple, Commodore, Atari, and IBM, almost all Japanese computers were incompatible with each other. This led to intense competition among the computer makers, with each vying to establish its own architecture as the dominant standard. Once NEC, Fujitsu, and Sharp took the majority of the market, the remaining computer makers banded together around the MSX, a shared computing standard developed by Microsoft Japan and ASCII. Despite being fourth place in the Japanese computer race, the MSX and its successor eventually achieved popularity in South America and Europe, while the Big Three never found success outside of Japan.

Alone at the edge of the world

But why didn't Japan use Western computers, and why does the West know so little about Japanese computers? It's not an easy question to answer, and the various reasons are rooted in economic, technical, and linguistic problems. One problem was simply the fierceness of the market on both sides of the Pacific, making it hard for any company to compete on foreign soil. America in particular became a doomsday market once the CEO of Commodore declared a price war against rival Texas Instruments. In this respect, the early Japanese computing scene resembles the British scene, which enjoyed a similar life of its own.

Nevertheless, most Japanese computer makers did attempt to sell their hardware overseas. NEC had a subsidiary company based in Boston, and presented various computers at American electronics shows in the early 1980s, such as the PC-6001 (rebranded the NEC Trek), the PC-8001, and the PC-9801 (rebranded the NEC APC). In Europe, computers from Panasonic, Sharp, Casio, and Fujitsu were debuted at the 1983 Hanover Messe, and Sharp's MZ series attracted a loyal following. The NEC APC (PC-9801) was even voted the Australian personal computer of the year for 1983. In general, Japanese computers received glowing reviews, as seen in this review of the NEC APC (1, 2, 3, 4), and this review of the NEC APC III (1, 2, 3), both published in a New Zealand computing magazine. Unfortunately, third-party support was a problem, and only a few Western software companies (such as Magicsoft) actively supported Japanese systems. Sales were sluggish, and these early computers were eventually forgotten.

From the Western perspective, Japan was probably more trouble than it was worth. The big Japanese electronics firms already had a tight grip on their domestic market. Furthermore, Japanese demand for personal computers was smaller compared to other countries, due to their perceived difficulty of use, a lack of good software, and the continued popularity of wapuro (i.e., word processors, highly specialized electronic typewriters designed to handle Japanese text). Japanese localization of software and documentation was also difficult in a less international era. As a result, most Western computer makers didn't bother establishing Japanese subsidiaries, and Japan was basically left alone until Compaq burst onto the scene in 1992 with its ultra-cheap IBM PC compatibles. One notable exception was Commodore, who released a Japanese version of the Commodore 64 in late 1982. Unfortunately, the C64 was dead on arrival, thanks to its high price, lack of accompanying software, and compatibility problems with imported titles. Otherwise, Japanese computer enthusiasts dying to get their hands on an Atari 800 or a ZX Spectrum were forced to import one themselves or rely on an independent distributor (such as the early Nihon Falcom). In the late 70s and early 80s, saavy Japanese geeks could also find cloned Apple II boards and other pirate hardware in the small electronics shops dotting Akihabara and Nihonbashi.

The most formidable barrier by far was the problem of electronically reproducing Japanese text. The Japanese written language has three major components: hiragana, katakana, and kanji. Hiragana and katakana are relatively simple phonetic alphabets with 46 letters each, somewhat similar to upper and lower case letters in English. In contrast, kanji is a collection of thousands of complex glyphs originally from China. The sheer number and visual complexity of these characters were beyond the memory and display abilities of most early computers (this is why the specialized wapuro were favored in Japan for so long). For Western languages, a single byte was sufficient to encode most letters and numbers (the US ASCII scheme uses even less memory). But a single byte can only express a maximum of 256 characters, around a tenth of what is needed to adequately display Japanese text. Furthermore, whereas 8x8 pixel blocks are sufficient to clearly display individual characters in Western alphabets, kanji become unreadable blobs without higher resolution. For this reason, any Western computer would require serious hardware modifications to enable full Japanese text support. This remained a problem throughout the decade, until a software-only solution (DOS/V) was developed in 1990. (Apple had created its own software solution in 1986, but it was too sluggish on contemporary hardware.) Meanwhile, Japanese computer makers specifically designed their hardware around higher-resolution display modes to accomodate Japanese text. Remember this, because display resolution became a key difference between Japanese and Western computers, and had a significant impact on game design.

Unsurprisingly, the computer that came to dominate the Japanese market also possessed an exceptionally good Japanese rendering architecture. First released in October 1982, NEC's PC-9801 was a true 16-bit computer with dedicated text VRAM for displaying kanji. A special font ROM was also created to store the thousands of kanji commonly used in Japanese writing. This kanji font ROM was initially sold separately, but came built-in with later models. Two custom graphic display controllers (GDCs) were implemented, one for text and one for graphics, offering a maximum resolution of 640x400 with 8 simultaneous colors (in 1982!). As a result, the PC-9801 could render Japanese text faster than any other personal computer on the market, making it perfect for business use. By 1987, the PC-9801 series had captured 90% of the Japanese market. Unfortunately, marketing strategy and internal politics at NEC caused the PC-9801 to develop a stodgy reputation as a business machine, and games for the system didn't really flourish until the late 1980s. In the meantime, the 8-bit computers were at the center of the burgeoning game industry.

Given this background, it makes sense to divide Japanese computer game history into 8-bit and 16-bit eras, each with its own trinity of hardware.

SPECIAL THANKS

Many thanks to Ben, Danjuro, Peter and everyone else at the Tokugawa forums for providing so much expert help, photos and more information than we could ever print. Also, thanks to www.NFGgames.com and www.pc98.org and for use of images. Thanks also to those on the forums at HG101, SelectButton, NTSC-uk, and anywhere else I've forgotten.