China's Computer Revolution: The Role of Modding

The Challenges of Chinese Character Output in Early Computing
In a prior discussion published on TechCrunch, the significant obstacles faced by computer engineers attempting to accommodate tens of thousands of Chinese characters within memory systems initially designed for smaller alphanumeric sets were investigated.
This analysis now shifts focus to the complexities of Chinese character output – encompassing monitors, printers, and associated peripherals – where further hurdles arose as engineers sought to ensure compatibility between Western-made personal computers and Chinese text.
The Central Role of Peripherals
Despite often being termed “peripherals,” implying a secondary function, these components were actually central to the development of Chinese computing.
This was true from the considerable limitations experienced in the 1970s and 80s, through to the substantial advancements achieved from the 1990s onwards.
Initial Incompatibility with Chinese Characters
During the early expansion of consumer PCs in the 1980s, no personal computer, printer, monitor, operating system, or other peripheral manufactured in the West natively supported Chinese character input or output.
These devices exhibited a similar English-language and Latin alphabetic bias seen in earlier technologies like telegraphic codes and mechanical typewriters, as previously explored.
A Period of Hacking and Innovation
The 1980s witnessed a period of intensive modification and reverse engineering within China and the broader Chinese-speaking world.
Engineers systematically adapted Western computing hardware and software to accommodate the Chinese language, resulting in a decentralized, yet remarkably innovative, era of experimentation.
Broader Implications for Computing Ecology
Examining the wider computing ecosystem – including printers, monitors, and essential supporting hardware – this second part of the series on Chinese computing highlights two key conclusions.
The Pervasiveness of Alphabet-Based Computing
The dominance of alphabet-based computing – or “alphabetic order” – extended beyond just keyboard layouts and computer memory.
Similar to the typewriter, computing devices, programming languages, and communication protocols were initially developed in English-speaking environments and subsequently adapted for other languages and non-Latin writing systems.
Chinese engineers consistently had to overcome the limitations of existing computing peripherals, hardware, and software to achieve even fundamental functionality.
Re-evaluating Perceptions of "Copycatting"
This analysis will challenge the commonly held, and often oversimplified, notion of Chinese “copycatting” and “piracy” that has characterized Western perspectives on Chinese computing during this crucial period of the late 1970s and 1980s.
Programs like “Chinese DOS” are often dismissed as mere “knock-offs,” but this perspective overlooks a critical point: without these “forgeries,” Western-designed software suites would have been entirely unusable within a Chinese character computing context.
Essentially, these adaptations were not simply imitations, but necessary modifications for functionality.
Dot-matrix printing and the underlying metallurgical constraints of alphabetic character sets
Our initial focus lies with the printer, specifically dot-matrix models. The story of dot-matrix printing in relation to Chinese computing commenced with the prevailing configurations of standard industry printer heads – the nine-pin heads present in nearly all mass-produced dot-matrix printers throughout the 1970s.
These commercially available dot-matrix printers could generate low-resolution Latin alphabet bitmaps with a single pass of the print head. This outcome wasn’t accidental; the selection of nine pins was specifically optimized for low-resolution Latin alphabetic character printing.
However, the same printer heads required a minimum of two complete passes to produce low-resolution Chinese character bitmaps. This two-pass process significantly increased printing time for Chinese compared to English and also introduced potential graphical imperfections, stemming from platen inconsistencies, uneven ink distribution, or paper malfunctions.
Visually, two-pass printing could lead to characters exhibiting varying ink densities between their upper and lower sections. Furthermore, without modification, all Chinese characters would be at least double the height of English text, regardless of the selected font size. This resulted in disproportionate printouts where English appeared concise, while Chinese characters seemed excessively large, and also led to considerable paper waste, resembling large-print materials.
A visual demonstration of the operational principles of these printer heads can be found in this video, provided by the author:
The prioritization of the Latin alphabet extended beyond initial expectations, as demonstrated by the work of pioneering Chinese computing expert Chan Yeh. Yeh, aiming to digitize Chinese characters, initially based his system on a bitmap grid of 18 by 22, and his first thought was to decrease the pin diameter to accommodate more pins on the print head. His research, however, revealed a more complex issue.
He discovered that the bias towards the Latin alphabet in impact printing was embedded within the physical properties of the printer components. Specifically, the metal alloys used for the printer pins were calibrated for nine-pin Latin alphabetic printing, meaning reducing their diameter to the size needed for Chinese characters would cause deformation or breakage.Engineers devised a solution to enable Western-built printers to accommodate up to 18 dots within the same vertical space normally occupied by nine dots.
This technique was both clever and straightforward. After a standard two-pass printing process, the first pass laid down an initial array of dots. Instead of positioning the second array directly beneath the first, they manipulated the printer to register them between the initial nine dots, similar to interlocking zipper teeth.
To accomplish this, engineers reprogrammed printer drivers to alter the printer’s paper advance mechanism, adjusting it to rotate by a very small increment – as little as 1/216th of an inch.
Pin configurations weren’t the sole obstacle. Standard dot-matrix printers were also optimized for the ASCII character encoding system, preventing them from processing Chinese text as actual text. In English-language word processing, printing involved transmitting a raster image, but English text could be sent directly as ASCII-encoded text, resulting in faster printing speeds.
To print Chinese characters on Western-built dot-matrix printers, the printers had to be utilized in graphics mode, typically reserved for raster images, as their “text” mode was incompatible.
This presented an irony for students of the Chinese language: to function on early Western dot-matrix printers, Chinese characters were forced to be treated as images or pictographs. This perception of Chinese characters as pictographs was a long-held Western assumption, though largely inaccurate, but within the context of dot-matrix printing, it became a necessity.
Ultimately, a new generation of impact printers emerged: 24-pin dot-matrix printers, with pin diameters of 0.2 mm (compared to 0.34 mm for nine-pin printers). Notably, the primary manufacturers of these printers were predominantly Japanese companies, including Panasonic, NEC, Toshiba, Okidata, and others. Japanese engineers faced comparable challenges due to the character requirements of their own language.
The Rise of Chinese Character Monitors
The manufacturing of computer monitors represented another significant area within the evolving landscape of Chinese computing. Similar to the challenges faced with printers, the issue of character distortion proved central. Chinese character bitmaps, even at low resolutions, inherently required significantly more space – both vertically and horizontally – than Latin letters.
This disparity resulted in Chinese text appearing disproportionately large when displayed alongside Latin text, a visual effect readily observable. Standard monitors produced in Western countries were also limited in the number of Chinese characters they could display simultaneously, impacting both the length of lines and the total number of lines visible on the screen.
A unique challenge specific to Chinese character display emerged: the necessity of the pop-up menu. Due to the iterative nature of Chinese input methods, where users are continually presented with character options based on their keystrokes, a mechanism for reviewing these candidates became essential.
Interestingly, the concept behind the pop-up menu isn't new. Its origins can be traced back to the 1940s with Lin Yutang’s experimental Chinese typewriter, which featured a “Magic Eye” – effectively the first mechanical pop-up menu.
The arrival of personal computers saw this mechanical window integrated directly into the computer’s display. Instead of a separate physical device, the pop-up menu transitioned into a software-controlled “window” or bar on the screen.
However, this integration further reduced the usable space on the monitor. Consequently, “pop-up menu design” became a crucial focus for research and innovation within Chinese personal computing from its earliest stages. Companies actively explored various menu styles, formats, and behaviors to optimize the user experience.
Each design choice involved trade-offs. Menus displaying more character candidates increased the speed of selection but consumed more screen space. Conversely, smaller windows were less intrusive but necessitated scrolling through multiple “pages” of options.
These limitations drove Chinese engineers and companies to continually pursue monitors with improved capabilities. While the demand for higher resolution monitors existed globally, the underlying motivations were particularly strong within the Chinese-language market.
The Importance of Resolution
- Higher resolution monitors allowed for more characters to be displayed.
- This addressed the space constraints inherent in Chinese character representation.
- Improved user experience by reducing the need for scrolling.
Innovation in monitor technology was therefore not simply about improving visual quality, but about enabling effective Chinese language computing.
The Limitations of Modification
Despite the ingenuity demonstrated in each of these modifications, they ultimately remained alterations to existing systems. True power resided in the ability to design and build original systems – the very ones that subsequently required modification.
While modding fostered a diverse range of systems, it often compromised interoperability. Continuous attention was essential, as no permanent, hands-off solution existed.
The Challenge of Program Compatibility
Each new computer program, and each updated version, demanded scrutiny from Chinese programmers. They had to meticulously examine the code to identify and adjust parameters affecting components like the computer monitor.
For instance, many English-language word processing programs defaulted to a 25-by-80 character display format (zifu fangshi xianshi). This was incompatible with Chinese character display, necessitating manual adjustments within the program’s code. Remarkably, this was often accomplished using the standard “DEBUG” software.
Through practical experience, engineers gained a deep understanding of the assembly code underlying prominent programs.
Ongoing System Changes and Adaptations
Even after systems like CCDOS were developed, changes continued to emerge. IBM’s announcement of the OS/2 operating system, for example, created significant challenges. An article from 1987 noted the disruption to Chinese-language systems, both in Taiwan and mainland China, as adaptation to the new OS was required.
This sparked a competitive effort to develop solutions compatible with IBM’s MS/DOS platform.
Historical Misunderstanding of Modding
Historically, the contributions of modders have often been overlooked or misrepresented. Their work was frequently dismissed as theft or piracy, rather than recognized as essential reengineering to enable compatibility with the Chinese language.
A 1987 cartoon in PC Magazine, for example, satirized Sinicized operating systems with the caption “It Runs on MSG-DOS.”
The Incorporation of Chinese Modifications
As Western manufacturers integrated many of these Chinese modifications into their core system architectures – alongside adaptations for Japanese and other non-Western languages – the origins of these changes are often forgotten.
It’s easy to assume that Western computers have always been language-agnostic and universally accessible, overlooking the influence of engineers in China and the non-Western world.
A Neglected History of Innovation
This crucial period in computing history remains largely undocumented due to a prevailing Western perspective. These modifications were not viewed as “experimentation” or “innovation,” but rather as “copycatting,” “mimickry,” or “piracy.”
When Chinese engineers reverse-engineered Western dot-matrix printers to enable Chinese character printing, or adapted Western operating systems for Chinese input method editors, the dominant Western perception was one of theft.





