Repository navigation
Language Authority Research Registry
github-actions[bot] edited this page Oct 11, 2026
·
1 revision
Purpose: establish a traceable, locale-specific research workflow for the 40 regional ARB catalogs. This registry is a starting inventory, not a claim that every listed institution has approved HWcontrol2.0 wording. Verify availability, scope, current URLs, access conditions, and reuse rights before depending on a source or automating access.
-
Start from the actual locale, not just its base language. Regional catalogs must use regional vocabulary and conventions where these differ (notably
en-US/en-GB/en-CA,es-ES/es-MX,pt-PT/pt-BR, andzh-CN/zh-TW). - Use a source hierarchy: (a) laws, standards, government language bodies and official terminology databases; (b) national academies and academic language institutes; (c) recognized specialist dictionaries, standards organizations and technical glossaries; (d) real product UI and established technical usage. Record when a source is not governmental or not prescriptive.
- Do not infer authority from a domain name. If a language has no single official regulator, record that fact and use multiple reputable references. English, for example, has no single worldwide official language academy.
- Research meaning in context. Search terms for hardware monitoring, cooling, temperature, firmware, permissions, diagnostics, updates and cybersecurity. Record definition, grammatical form, regional scope, source URL, access date, licensing/reuse notes, decision and rejected alternatives.
- Do not blindly scrape. Respect each site’s terms, robots/access policy, rate limits, API terms, copyright and database rights. Prefer documented APIs, bulk downloads and open-licensed datasets. Never bypass login, CAPTCHA, access controls or rate limits.
- Sources guide translation; they do not replace linguistic review. An official general dictionary may not contain specialist hardware vocabulary. A term can be common but still misleading in a safety-critical message.
- No silent source-of-truth split. ARB remains the canonical runtime translation catalog. This registry is research metadata only; it must not become a second translation catalog.
-
No unverified claims. A translation is not “officially approved” unless the relevant body has explicitly reviewed or approved it. Mark unresolved terms as
needs-review. - Automated quality gates: exact key coverage, ICU/placeholder parity, non-empty values, locale-tag validity, bidirectional-text checks, critical-message review, UI overflow and accessibility checks. Passing structural tests is not proof of native-quality language.
- Use current official spelling and terminology guidance. Record the version/date of any rule or standard that influences a decision.
- Unicode CLDR / LDML — locale identifiers, inheritance, plural categories, date/time and number formatting, collation, units and message-format conventions. Use it for locale behavior, not as the sole authority for prose or technical terminology.
- Unicode CLDR data — locale data and change history; check the applicable release and license.
- W3C Internationalization — internationalization and text-direction guidance.
- RFC 5646 / BCP 47 — language tags.
- Unicode Bidirectional Algorithm, UAX #9 — bidirectional text behavior.
- UNTERM — UN-system terminology in Arabic, Chinese, English, French, Russian and Spanish; useful for general institutional terminology, not a substitute for consumer hardware terminology.
- TermAB — Türkiye Ministry of Foreign Affairs EU Affairs terminology database, including Turkish, English and French.
- IATE — EU inter-institutional terminology database; useful for supported EU language pairs and institutional terminology, not a universal language regulator.
- Microsoft Language Portal — software UI terminology and style references where available; treat as product terminology guidance, not a national language authority.
- Google developer documentation style guide — technical writing guidance; not a language regulator.
“Primary” means the first place to investigate, not an endorsement of every entry or a claim of exclusive authority. For each locale, consult all applicable spelling, dictionary, terminology, grammar, technical and public-sector plain-language resources; log which source supports each term.
| Locale (ARB) | Regional language scope | Primary sources to investigate | Scope / caution |
|---|---|---|---|
ar_SA |
Arabic, Saudi Arabia | King Salman Global Academy for Arabic Language; UNTERM | Check Modern Standard Arabic and Saudi usage; verify the academy’s specific terminology products. |
bg_BG |
Bulgarian, Bulgaria | Institute for Bulgarian Language, Bulgarian Academy of Sciences | Normative dictionary, spelling and linguistic resources; verify specialist technical coverage. |
bn_BD |
Bengali, Bangladesh | Bangla Academy | Bengali lexicography and usage; verify current spelling/terminology publications. |
ca_ES |
Catalan, Spain | Institut d’Estudis Catalans; DIEC | Normative Catalan lexicon and terminology; use Catalan-specific sources rather than Spanish equivalents. |
cs_CZ |
Czech, Czechia | Institute of the Czech Language, Czech Academy of Sciences | Czech spelling, dictionary and language advice; supplement with technical sources. |
da_DK |
Danish, Denmark | Danish Language Council; Ordnet | Danish spelling and dictionaries; verify technical terms in context. |
de_DE |
German, Germany | Council for German Orthography; DWDS | Spelling rules and lexical evidence; distinguish normative spelling from usage corpora. |
el_GR |
Greek, Greece | Centre for the Greek Language | Greek dictionaries and language resources; check contemporary technical usage. |
en_CA |
English, Canada | TERMIUM Plus, Government of Canada; Government of Canada Language Portal | Canadian spelling and bilingual terminology; English has no single global regulator. |
en_GB |
English, United Kingdom | GOV.UK style guide; Oxford English Dictionary | Government writing conventions plus lexicographic evidence; OED is not a government regulator. |
en_US |
English, United States | PlainLanguage.gov; NIST for technical terminology | Plain-language and specialist standards; no single official US English academy. |
es_ES |
Spanish, Spain | RAE; FundéuRAE | Spanish spelling and usage; check whether recommendations are pan-Hispanic or Spain-specific. |
es_MX |
Spanish, Mexico | RAE / ASALE; Academia Mexicana de la Lengua | Review Mexican vocabulary and usage; do not copy Spain-only terms automatically. |
fi_FI |
Finnish, Finland | Institute for the Languages of Finland (Kotus) | Finnish language recommendations, dictionaries and terminology. |
fil_PH |
Filipino, Philippines | Komisyon sa Wikang Filipino | Filipino language policy and vocabulary; verify whether a technical term is natural and current. |
fr_FR |
French, France | FranceTerme; Académie française | Government terminology and dictionary resources; distinguish official terminology recommendations from general usage. |
he_IL |
Hebrew, Israel | Academy of the Hebrew Language | Hebrew spelling, grammar and terminology; check current technical coinages. |
hi_IN |
Hindi, India | Commission for Scientific and Technical Terminology; SHABD terminology portal | Government scientific/technical terminology; verify actual UI readability and regional register. |
hr_HR |
Croatian, Croatia | Institute of Croatian Language | Croatian dictionaries, orthography and terminology; distinguish Croatian forms from closely related standards. |
hu_HU |
Hungarian, Hungary | Hungarian Research Centre for Linguistics; Hungarian electronic dictionaries | Hungarian linguistic and lexicographic resources; verify whether each source is normative or descriptive. |
id_ID |
Indonesian, Indonesia | Badan Pengembangan dan Pembinaan Bahasa; KBBI | Indonesian dictionary and language guidance; verify the current agency domain and dictionary access. |
it_IT |
Italian, Italy | Accademia della Crusca; Treccani | Italian usage, dictionary and language advice; specialist computing terminology may need additional sources. |
ja_JP |
Japanese, Japan | Agency for Cultural Affairs; National Institute for Japanese Language and Linguistics | Government language policy, writing conventions and linguistic research; validate UI conventions and terminology. |
ko_KR |
Korean, South Korea | National Institute of Korean Language; Standard Korean Dictionary | Official Korean dictionary, spelling, terminology and public-language resources. |
ms_MY |
Malay, Malaysia | Dewan Bahasa dan Pustaka; PRPM | Malay dictionary and terminology; verify Malaysian standard and contemporary technical wording. |
nb_NO |
Norwegian Bokmål, Norway | Language Council of Norway; Bokmålsordboka | Bokmål spelling and lexicon; do not mix Nynorsk and Bokmål without a deliberate locale decision. |
nl_NL |
Dutch, Netherlands | Dutch Language Union; Woordenlijst Nederlandse Taal | Dutch spelling and terminology; check Netherlands/Belgium regional differences where relevant. |
pl_PL |
Polish, Poland | Polish Language Council; PWN Dictionary | Polish orthography, usage and dictionary evidence; confirm current recommendations. |
pt_BR |
Portuguese, Brazil | Academia Brasileira de Letras; VOLP | Brazilian spelling and vocabulary; do not treat Portugal usage as interchangeable. |
pt_PT |
Portuguese, Portugal | Academia das Ciências de Lisboa; Priberam | Portugal-specific vocabulary and spelling; distinguish institutional from commercial dictionary sources. |
ro_RO |
Romanian, Romania | Romanian Academy; DEX online | Romanian spelling and lexicon; DEX online is an aggregator, so check the underlying dictionary source. |
ru_RU |
Russian, Russia | Vinogradov Institute of the Russian Language; Gramota.ru | Russian language advice and lexicographic resources; distinguish official/academic guidance from reference portals. |
sv_SE |
Swedish, Sweden | Institute for Language and Folklore (ISOF); Svenska Akademiens ordböcker | Swedish language recommendations and dictionaries; verify contemporary technical usage. |
th_TH |
Thai, Thailand | Office of the Royal Society | Thai dictionary and terminology resources; confirm current official portal paths and technical vocabulary. |
tr_TR |
Turkish, Türkiye | Turkish Language Association (TDK); Güncel Türkçe Sözlük; TDK spelling guide; TermAB | Spelling, established Turkish equivalents and bilingual institutional terminology. Record decisions and alternatives; avoid forced translations that obscure technical meaning. |
uk_UA |
Ukrainian, Ukraine | National Commission on State Language Standards; Ukrainian Language and Information Fund | State-language standards and dictionary resources; verify current government URLs and terminology. |
ur_PK |
Urdu, Pakistan | National Language Promotion Department | Urdu dictionary and terminology resources; confirm script, spelling and current technical usage. |
vi_VN |
Vietnamese, Vietnam | Vietnamese Institute of Linguistics | Vietnamese lexicography and linguistic research; supplement with current Vietnamese technical usage. |
zh_CN |
Simplified Chinese, mainland China | Ministry of Education of China; National Language Commission resources | Mainland standards and simplified-character usage; locate specific current terminology standards for each technical term. |
zh_TW |
Traditional Chinese, Taiwan | Ministry of Education Dictionary; National Academy for Educational Research | Taiwan-specific terminology and Traditional Chinese; do not merely convert Simplified characters mechanically. |
Create one record per disputed or high-impact term, not one vague source list per language.
| Field | Required content |
|---|---|
| Source key | Stable ARB key, e.g. fanSpeed
|
| Concept | Meaning in HWcontrol2.0, including what it does not mean |
| Locale | Exact BCP 47 locale |
| Candidate terms | Candidate wording, grammatical form and regional variants |
| Evidence | Direct source URL, source title, quoted minimal definition or relevant excerpt, lookup date |
| Source status | Government / statutory standard / academy / academic / commercial / community |
| Reuse rights | License, terms or manual-review-only status |
| Decision | Selected wording and short rationale |
| Safety impact | Whether ambiguity could affect security, hardware control or recovery |
| Review status |
unreviewed, researched, native-reviewed, approved
|
| Reviewer / date | Reviewer identity or role and date, when available |
- Extract the actual 235+ canonical source keys and group them by concept rather than translating isolated strings.
- Prioritize high-risk concepts first: fan-control state, temperature limits, permissions, bridge authentication, security blocks, update verification, diagnostic consent and recovery instructions.
- For each of the 40 locale catalogs, verify the locale-specific source links above and record broken links, missing official resources, access restrictions and the best replacement source.
- Build a controlled terminology list with definitions, allowed variants, forbidden/ambiguous variants, grammatical notes, source citations and locale tags.
- Apply terminology consistently to ARB translations and metadata; do not rewrite unrelated configuration JSON or INI files.
- Add automated checks for terminology drift and source-registry schema. Automated checks should flag for review rather than blindly replacing natural-language strings.
- Run Flutter localization generation, analyzer, tests and visual overflow/accessibility checks. Have native or demonstrably proficient reviewers review every locale before claiming linguistic quality.
- Revisit this registry periodically. Record source last-checked dates and investigate broken URLs rather than silently treating old references as current.
- Registry created as a research starting point: yes.
- All 40 locale rows represented: yes.
- Every source independently verified for current availability, exact jurisdiction and reuse rights: not yet.
- All ARB translations reviewed against sources: not yet.
- Native-speaker review completed: not yet.
- This document must not be presented as proof that all translations are officially endorsed or production-quality.