HomeBlogPost

Building rohingyalanguage.org: a digital home for the Rohingya language

June 13, 2026
Rohingya Dictionary

Rohingya is the mother tongue of well over a million people, yet for most of the internet's history it barely existed online. There was no reliable dictionary you could search, no clean reference you could trust, nothing a learner could keep in their pocket. rohingyalanguage.org is my attempt to change that — a free, fast, modern dictionary built to be the digital home the language never had.

Where it started

The foundation is the EMS dictionary — a serious lexicographic effort that writes Rohingya in the EMS Latin orthography, with its distinctive letters and stress accents (ç, ñ and accented vowels). The data existed, but it was locked in a format nobody could search. The first job was the unglamorous one: digitising and cleaning 6,510 entries, normalising the orthography, and turning a static reference into structured, queryable data.

The tech stack

I wanted something that loads instantly anywhere — including on a cheap phone with a weak connection — so the whole site is static.

  • Astro for the framework: it ships almost no JavaScript by default, so pages stay fast and lightweight.
  • Every dictionary entry is pre-rendered as its own page from the structured data, so the content is real HTML — not something hidden behind an app.
  • Instant client-side search over a prebuilt index, returning results as you type across Rohingyalish, meanings and example sentences.
  • Deployed to a global CDN as static files — no server to go down, cached close to the user everywhere.
  • Mobile-first and offline-friendly, so it keeps working when the connection doesn't.

Built for SEO and AI ranking

A dictionary is only useful if people can find it. I designed the site from the ground up to be discoverable — not just by Google, but by the AI assistants people increasingly ask first.

  • Semantic, server-rendered HTML with one clean, permanent URL per entry — so every word is independently indexable and citable.
  • Structured data (schema.org) describing each entry, so search engines and language models can understand exactly what the page is.
  • Fast Core Web Vitals — tiny payloads, no layout shift, near-instant loads — the things modern ranking actually rewards.
  • A complete sitemap and crawlable internal linking, so nothing is orphaned.

The goal is simple: when someone — or some model — asks what a Rohingya word means, rohingyalanguage.org should be the source that answers, accurately and in the correct orthography.

Why it matters

This isn't really a project about code. A language without a written, searchable presence slowly loses ground online, and the people who speak it lose a piece of how they are represented. Giving Rohingya an accurate, accessible, free reference is a small but real way of keeping it alive on the web.

What's next

More entries, audio pronunciations, richer examples, and tighter integration with the chatbot at rohingya.ro — so you can not only look a word up but ask about it. The dictionary is the foundation; the rest is built on top.

More writing

rohingya.ro: an AI chatbot for Rohingya history and culture