At the end of last year, there were a couple of articles about Japan’s failure to be a giant in the new digital age (Newsweek on Why Apple Isn’t Japanese, and there were some interesting comments in a blog response Japan is no longer a leader in Electronics). Unfortunately, the Newsweek article completely ignores the technical background which forms the basis for Japan’s current position in the digital age. In this post, I want to explore this technical background and show why business-types such as CEO’s and Newsweek readers really do need to understand the underlying technical issues of a problem.
The real start of the digital gulf between Japan and the western world started back in the late 1970s and early 1980s when the first 8-bit home computers began to be released. One of the most important factors at this time was the complexity of the Japanese language. Put simply, an 8-bit computer with only 64k of memory simply does not have the capacity to edit Japanese. As an example, the first Japanese word processor to use the modern kana-kanji text entry system was the Toshiba JW-10. The JW-10 was a dedicated word processor with no other functionality. Released in February 1979, the JW-10 weighed 220kg and had a price tag of 6,300,000 yen (around $30,000). Here in the west, we could get similar capabilities with a $300 Commodore Vic-20 connected to a cheap 8-pin dot matrix printer. (In fact, you could argue that the Vic-20 offered better functionality). Before we continue with the history, let’s look at the technical details. (Skip this section if you don’t care).
There are two separate technical barriers for the Japanese language. The first is displaying the characters. For anyone who hasn’t used an Apple II, Commodore-64 or other 8-bit system, here is a screen-shot showing the cutting-edge 8×8 pixel fonts of 1982. For comparison, let’s look at a few common Japanese characters (these are ranked as the 35th, 64th, and 104th most commonly used characters, respectively):
議 選 調
It should be pretty clear that the Japanese are not getting away with 8×8 pixel characters on a 320×200 pixel screen. The NEC PC-9801 was one of the leading Japanese personal computers of the time, and offered a 640×400 display with 16×20 pixel characters. For all 6802 characters defined in the 1978 Japanese Industrial Standards (JIS C 6226-1978), we get 6802×20x16 bits = 265 kBytes of font data. (Compared to a mere 2 kBytes for 256 8×8 ASCII characters).
The second problem is the text entry process – whereby the 100 or so keys on the keyboard are used to select from the 6000 or so characters. In the days of telegraph, this was done using a 94×94 cell table written on a sheet of paper with all of the characters listed on it. The “kuten” codes were the coordinates of the characters in this table. The modern kana-kanji conversion approach involves entering characters phonetically, and converting based on dictionaries. The size of the dictionaries is directly related to how easy it is to enter text. For comparison, the conversion dictionaries on Windows XP are 36 MB. An 8-bit system is pretty much limited to the size of a disk.
Back to 1983, and we can see the choices available to the Japanese consumer.
1) The Nintendo Entertainment System released 1983. 14400 yen (around $70). No keyboard. Japanese text is displayed as graphics stored in the game (only the font data for displayed characters need be included).
2) 8-bit computer with “katakana” Japanese characters. (Examples include the NEC PC-8001 series. The Mark II was released in 1983 was 123,000 yen (around $600)). Katakana is a Japanese phonetic alphabet only requiring 50 characters. Before the advent of computers, katakana was never used to write entire sentences. Computers that only offered katakana had no practical value to home users over the NES, except for computer enthusiasts. Although these computers are similar to the Commodore 64/Apple II in terms of technical capabilities, they are crippled in terms of functionality from the viewpoint of home consumers.
3) 16-bit computers with special hardware for displaying kanji. The NEC PC-9801 was released in 1982 and became the number one selling home computer in Japan. The first model retailed for around 300,000 yen ($1500). The ROM chips containing the kanji font data were extra (50,000 yen – $250), and only support Level 1 characters (around half of the number in the Japanese standards). Kanji input was by kuten code (i.e. you had to type in the hexadecimal number of the characters you wanted to display). The PC-9801 featured a standard resolution of 640×400, which would not be surpassed as a standard in the west until the release of VGA and Macintosh II in 1987.
4) Specialist word processing systems (i.e. the Japanese equivalent of an electronic typewriter). The JW-1 (grandchild of the JW-10 discussed in the intro) was released in 1983 for 500,000 yen (around $2,500).
So, 6 years after the Apple II first brought computers into homes in America, the Japanese still had no reason to buy a home computer unless they wanted to type by entering hexadecimal codes. Whereas computer manufacturers in the west could advertise under the premise of Why Buy Just a Video Game?, Japanese households were faced with a choice between $1500 for a world of hexadecimal nightmare compared to $70 for the world of Mario Bros.
Appliances vs General Computers
Throughout the 1980s, western countries saw explosive growth in general-purpose computing platforms, with even games manufacturer Atari switching from a console to a general-purpose computer (Atari ST). Japan, on the other hand, was deeply immersed in an appliance mentality of computers. Companies like Nintendo, Sega, and Sony dominated the games console market. And while American developers worked on desktop operating systems such as Windows and MacOS for general purpose computers, Japanese engineers were working on embedded operating systems such as TRON and T-Kernel for running embedded devices such as fax machines and car engine control systems. (As an aside, although the name TRON is not well-known among consumers, the number of devices running TRON is on par with the install-base of Windows).
By the time the iPod was released in 2001, Japanese mobile phones were already e-mail and internet capable. Although personal computer numbers had grown, more Japanese were accessing the Internet through their mobile phone than through a computer, and Japanese manufacturers were locked into the appliance mindset. As an example, consider the Sharp J-SH51 mobile phone released in 2002 which also offered a built-in MP3 player and digital camera. Despite being one of the most advanced mobile phones in the world at the time, the J-SH51 could not be connected to a computer. So how did you get music onto your phone? Well, you took an analog audio cable and plugged it into the aux. out plug on your CD player.
In the west, the home computer was already being viewed as the central hub of the digital age. It was obvious that devices such as digital cameras and MP3 players would need connectivity with the home computer, and that people would transfer pictures from the digital camera to their computer, or would use their computer as the central storage for music files to upload to their iPod or other music player as needed. The iPod, for example, requires a home computer. Without one, there is no way to get music on or off the device.
In Japan, however, things were different. Perhaps the easiest way to understand the Japanese market at the time is to imagine that home computers did not exist. From this perspective, the direction that the Japanese electronics industry took makes perfect sense. Everything needed to be designed as stand-alone appliance. The basis for much of this was the digital memory card, particularly the SD card. Digital cameras and camera-phones stored everything on a memory stick, and offered DPOF configuration options for configuring printing options. Color printers went on sale offering SD card slots so that these photos could be printed without a computer in the middle. MP3 players took a similar turn, offering either analog cable connectivity or SD card slots for music transfers. New stereo systems also offered an additional SD card slot. The SD card was like the new cassette. Record stores even began offering machines that sold digital music directly stored on your SD card. 3G phone handsets were released in 2001, and Japanese telecoms envisioned a world where consumers would buy and download music directly onto their mobile phones. It all makes sense if nobody owns a home computer, and when the mobile phone is the dominant form of Internet connectivity.
The flip-side of all this support for stand-alone appliances that do not require a home computer is that the Japanese electronics manufacturers offered virtually no support at all for home computers. Many devices simply could not be connected to a PC. For those that could be connected, the support software was unfriendly and extremely primitive. Let’s take the SD card as an example. SD cards offered a ‘feature’ called SD Audio whereby music was stored protected by a DRM system. However, only one manufacturer ever produced USB card readers that actually supported this scheme. Even if you did manage to track down the lone card reader that supported SD Audio, you still can’t transfer music to your SD card. In fact, you now had to purchase a special version of RealPlayer (that’s right, you had to pay for free software).
Of course, this kind of situation wasn’t going to fly in the west, where everyone had a home computer. Even in the Japanese market, this wasn’t going to fly. By the year 2000, most of the technical difficulties facing computers in Japan in the 80s and 90s had been resolved, and home computers were becoming mainstream. Japanese consumers wanted PC connectivity from their appliances, and the iPod offered a well-designed, highly functional package. So Apple created the iPod, and Japanese electronics manufacturers were left to re-evaluate a new world where the home computer is the hub for digital media.
Japanese manufacturers are finally catching on and have started to offer PC connectivity. Most mobile phones now come with USB cables and software for transferring photos and music to and from a PC. While the latest music-capable phones still come with an SD card, support has been added for Apple’s MP4 format, and files can be transferred using a generic card reader. It’s difficult to say whether the misjudgment of the Japanese electronics companies was a one-off falter, or if it is the start of a trend. The Nintendo Wii is certainly a strong indicator that all is not lost, even if it is back on the Japanese strong ground of an appliance-oriented device.
- Japanese Typewriters: Mechanical typewriters capable of typing thousands of different characters
- Using an IME: Japanese text entry in modern computers
- A history of character encodings up to ASCII in terms of the mechanical systems that limited them
- How 5-bit serial communication was decoded without transistors all the way back in the 1880s – Baudot’s telegraph