Khmer script in computing

Khmer script in computing is the history of how the Khmer script came to be typed, stored and read on computers and phones.

The Khmer Unicode keyboard layout (NiDA). Photo: Kiensvay, CC BY-SA 3.0

History details
History type	Movement
Present-day countries	Cambodia
Subjects	Language & Literacy; History & Remembrance

Before a shared standard, Khmer text was written in fonts that each used their own encoding, and text could be read only with the same font. A Khmer encoding for Unicode was settled in 1999, and the Cambodian government accepted its code table in 2002. Volunteers and the KhmerOS project then brought the script to free software on Linux and Windows, and developers later brought it to phones.

Legacy fonts

Javier Solá, who coordinated the KhmerOS project, told The Cambodia Daily that Khmer script had up to 30 different encodings before Cambodia had a standard, a situation he called a "mess". A 2005 United Nations Development Programme primer on software localization counted up to 26 font encodings, which worked well enough in Microsoft Word.

Fonts in wide use included Khek, Limon and Zero-Space. Most were not built to the Unicode Standard, and web browsers did not support them, so many could be read only with special software drivers.

A reader needed the same font as the writer to read an e-mail, a document or a web page. In 2006 Tharum Bun wrote that a Cambodian student in Japan turned his Khmer text into an image before he posted it on his weblog. He also quoted a 2002 Cambodia Daily article: a government minister who wanted to write to his wife, who did not speak English, had been told that e-mail could not carry Khmer.

Khmer in Unicode

Work on encoding Khmer in Unicode started in 1996 in a committee outside the government, and the standard was settled in 1999. At a meeting of the International Organization for Standardization (ISO) working group in Beijing in 2000, the Cambodian delegation pointed out serious glyph errors in the Khmer chart of Unicode 3.0. The characters U+17BE to U+17C5 had been printed wrongly, and the same error was in the ISO standard. Cambodia's Committee for Standardization of Khmer Characters in Computers objected to the code table in letters of May 2001 and asked for it to be replaced. It said that the table did not match how Cambodians understand their language, and that it had been adopted without the Cambodian national standards body. At the working group's 41st meeting, in Singapore, the Cambodian government formally objected to the Khmer block and proposed a new encoding to replace it. In a reply of December 2001, Paul Nelson of Microsoft argued that the existing characters had to stay as encoded, while granting that Cambodian officials may never have signed off on the standard. He proposed adding the lunar date symbols put forward by the Cambodian delegation in a new block near U+19E0 to U+19FF.

Unicode kept the table. On May 7, 2002, the committee decided "with regret" to adopt it as published in Unicode 3.2, because Cambodia could not afford further delay or an incompatible national standard.

In January 2010 the Council of Ministers spokesman Phay Siphan said that a memorandum of December 24 gave ministries 60 days to switch their computers to Unicode. Sok Tha of the Ministry of Education said Khmer Unicode had been used in Cambodian schools since 2007.

Khmer on Linux

When the KhmerOS project was first planned, Khmer worked in some programs under Windows through Microsoft's text engine, but no free software ran in Khmer on GNU/Linux. The typographer Danh Hong put one of his Khmer OpenType fonts into the public domain, which helped draw open-source developers to the language.

Lin Chear, a Canadian engineer of Khmer origin, wrote the programs that added Khmer to Pango, the text layout library of the GNOME desktop, in a few days. With his help the maintainer of Qt added Khmer to KDE, and he got Mozilla working in Khmer on GNU/Linux with a Pango patch. Maurice Bauhahn, who kept a reference site on Khmer in Unicode, recorded his Pango work in December 2003 and his Linux implementation in January 2004.

Lin Chear also wrote a tool that converted text typed in the ABC fonts' encoding to Unicode. His original converter is among the files of the KhmerOS project's ABC and Limon font to Unicode converters, dated 2005.

KhmerOS

Javier Solá, a Spanish computer scientist living in Cambodia, planned an operating system, office programs, fonts, a keyboard and training, all in Khmer. The project found a home at the Open Forum of Cambodia, a nonprofit in Phnom Penh. With its first $1,500 in donations it hired two computer scientists to build a glossary of Khmer computer terms.

On November 21, 2003, Solá set out the plan on a GNOME translators' mailing list as "KhmerOS 1.0". He counted about 53,000 program messages and 1,300 pages of documentation to translate, and planned to put around 50 fonts in the public domain. Writing in 2006, Solá dated the start of the work to early 2004. That year the team grew from three to eight full-time members, translated basic office and internet programs, and prepared fonts and a keyboard definition. At the start of 2005 the National ICT Development Authority (NiDA) joined, which made KhmerOS a joint project of the government and a civil society group. The team wrote 65 hours of training materials in Khmer for OpenOffice.org, Firefox and Thunderbird.

In 2005 The Cambodia Daily described it as a joint project of the government and the Open Forum. Khmer versions of OpenOffice.org, Firefox and Thunderbird were in testing on over 1,000 computers around the country, and 1,000 Khmer keyboards were being made in China. Solá said, "You cannot run a government on a foreign language." The KhmerOS fonts, among them Khmer OS System, a font hinted for reading at small sizes, are the work of Danh Hong. Their copyright is held by Danh Hong and the Open Forum of Cambodia, and they are released under the GNU Lesser General Public License. By 2010 the KhmerOS project was run by the Open Institute.

Windows and Mac

Microsoft lists Khmer UI, a font from Cambodia, among the fonts supplied with Windows 7 and later versions. In Windows 8.1 the interface font for Khmer changed from Khmer UI to Leelawadee UI. The Cambodian Information Center notes that Mac OS X can read and write Khmer Unicode with no added font, and that Windows Vista includes Khmer Unicode but has to be set up to use it.

Phones

A developer's account published by DAI says that by 2013 many young smartphone users in Cambodia could not type Khmer, because their phones did not readily support the script, and many wrote Khmer in Roman letters. He released the Khmer Smart Keyboard for iOS in September 2014 and for Android in March 2015. Within a year more than 300,000 people had downloaded it, and it won the 2015 Cambodia ICT Award from the Ministry of Post and Telecommunications.

In 2015 Marc Durdin, who was learning Khmer, adapted a desktop keyboard by Andrew Cunningham, based on the NiDA Khmer layout, into a Khmer keyboard for Keyman on iPhone, iPad and Android. He noted that other Khmer keyboards already existed for both systems.

References