Making the Tibetan language a first-class citizen in the digital world
Efforts to make Tibetan a “first-class” digital language highlight how conventional software assumptions—like reasonably short paragraphs and space-separated words—break down for scripts that use continuous text across hundreds of pages. Contributors describe recent LibreOffice optimizations that finally allow efficient layout of extremely long Tibetan paragraphs, enabling practical editing and publishing of Tibetan and related languages. The exchange broadens into questions of how minority and endangered languages should adapt (or not) to digital norms, the role of Unicode and open-source projects in preserving diverse writing systems, and the cultural stakes of maintaining Tibetan under political and social pressure.
LibreOffice and Long Tibetan Paragraphs
- Discussion centers on a key LibreOffice fix that made layout of extremely long paragraphs scalable for Tibetan text.
- A linked patch replaces an O(n²) script-run computation with an LRU cache and adjusts layout context sizing, turning a long‑standing performance bug into a small final change.
- Some note the bug existed for nearly a decade, suggesting Tibetan text was rarely tested; others counter that the visible “5‑line fix” caps a years‑long refactor that made such a solution possible only recently.
- A LibreOffice QA volunteer describes ongoing efforts to improve RTL/CTL/CJK support, invites bug reports and donations, and notes a meta‑bug tracking many minority scripts (Tibetan, Mongolian, Uyghur, etc.).
Tibetan Writing Conventions and Paragraph Length
- The thread highlights that Tibetan manuscripts often use extremely long, unbroken text streams—sometimes spanning tens of pages or more—without Western‑style paragraphs.
- This clashes with word processor assumptions like “paragraphs are short” and “text has spaces,” joining the list of “falsehoods programmers believe about text.”
- Comparisons are made to scriptio continua in ancient Greek/Latin and to legal documents that ban paragraph breaks, producing single paragraphs across hundreds of pages.
Language Preservation vs Evolution
- Some argue software should faithfully support existing Tibetan conventions, including massive paragraphs, both for current users and historical texts.
- Others suggest Tibetan orthography might reasonably evolve—adding spaces and paragraphs in the digital era, as other scripts did when media changed.
- A counterpoint stresses that making software more flexible enables, rather than blocks, innovation in all languages.
Politics and the Stakes for Tibetan
- Several comments tie digital Tibetan support to broader concerns about cultural erasure and Sinicization, arguing that language technology is part of preserving Tibetan identity.
- Others say geopolitical realities made “Free Tibet” activism fade in the West, as confronting a powerful China is seen as unrealistic.
- There is disagreement over whether such political discussion is appropriate in a technical thread, but defenders insist the linguistic work is inseparable from the political context.
Unicode, Scripts, and Related Languages
- Tibetan script has long been in Unicode; complaints about Unicode prioritizing emoji are challenged as unfounded in this context.
- Clarifications: many Tibetic languages share a single modern Tibetan script; Dzongkha also uses it, though feature completeness for specific languages may vary.
- Bengali/Assamese and Tibetan scripts share historical roots (via Gupta) but encode unrelated language families and are not mutually intelligible.
Historical and Community Efforts
- The thread recalls early work digitizing Tibetan, such as HyperCard pronunciation tools and a 1990s project to build Tibetan bibliographies, fonts, and library systems.
- An open‑source Tibetan dictionary project is mentioned as another piece of the digital ecosystem.