Tech Innovations to make the Tibetan Language a First-class Citizen in the Digital World

In general, word processing software programs were originally designed with English text in mind, meaning that they support relatively short paragraphs (possibly up to a few pages), with words separated by spaces that can have a flexible width. These assumptions don’t work for written Tibetan, however, where paragraphs are virtually limitless and contain very few spaces. The result is that such a word processor performs very poorly, if it works at all, when opening a long Tibetan text, and it will therefore be impossible to use for any serious publication project. 

LibreOffice is one of the only mature and stable open source word processors (i.e. inspired by MS Word) that is available for free on all possible platforms, including Linux. It is relevant to Tibetan Studies and the Tibetan community because it is free and open source, while popular commercial software programs such as Word or InDesign which can handle long Tibetan texts, are expensive and in many parts of Asia often used in pirated versions. So long as LibreOffice could not handle long paragraphs there was essentially no free tool to publish Tibetan. 

In 2015, BDRC’s CTO, Elie Roux, reported the issue to LibreOffice. Unfortunately an intervention in the code of LibreOffice is a big project that would have required weeks of research and development, and therefore there was no evolution on that front. That is until a few weeks ago, when a developer named Jonathan Clark tackled the issue and fixed it! A long text like Longchenpa’s Yishindzö (view on the BDRC archive at: yid bzhin mdzod), which consists of one “paragraph” of 153 letter size pages, now opens quickly and can be edited in LibreOffice. The conversion of the RDF file of this particular text into a PDF used to stall after 45 minutes and it is now completed in 13 seconds! This is a remarkable improvement of several orders of magnitude, something quite rare in software optimization.