Skip to content

Credits

Writing and localization guidance includes adaptations of the following works. The adaptations combine and paraphrase their advice, correct identified errors, and apply FontLab and Vexy terminology. Examples written for this guide do not imply endorsement by the original publishers.

Selected language and character checks also use the GOV.UK style guide, Cambridge Dictionary, Unicode punctuation guidance, OQLF language and typography guidance, the Translation Bureau’s language guidance, and the Académie française on accented capitals. These checks inform local examples and distinctions; they do not replace the house rules.

Spanish language checks also use the RAE and ASALE dictionaries, including regional usage and grammatical forms, and FundéuRAE’s decimal-separator guidance.

Portuguese language checks also use Ciberdúvidas on number scales and verb constructions, the Priberam dictionary, and Embrapa’s initialism guidance.

German spelling checks use the official rules published by the IDS. Polish spelling checks use the Rada Języka Polskiego’s 2026 rules and announcement.

Italian language checks also use the Accademia della Crusca’s guidance on borrowed-word plurals and accents and apostrophes. Indonesian spelling checks use the Badan Pengembangan dan Pembinaan Bahasa’s EYD V.

Dutch language checks also use Taaladvies.net and Team Taaladvies, including guidance on compounds, capitals and verb order. Afrikaans checks also use VivA and selected spelling and notation sections of Stellenbosch University’s style guide. Their institutional conventions do not replace the house rules.

Danish language checks use Dansk Sprognævn and sproget.dk, including current spelling changes, genitives, accents and punctuation. Swedish checks use Språkrådet’s Frågelådan for pronouns, compounds, punctuation and numeric spacing. These local conventions do not replace the house rules.

Norwegian checks use Språkrådet for Bokmål and Nynorsk grammar, pronouns and notation, and the University of Bergen and Språkrådet dictionaries for current word forms. House preferences among accepted variants are distinguished from spelling rules.

Finnish language checks use Kielitoimiston ohjepankki for case, compound and punctuation rules. Estonian checks use the Eesti Keele Instituut’s EKI teatmik and Sõnaveeb for grammar, word forms and notation. The examples apply these distinctions to interface writing.

Hungarian checks use A magyar helyesírás szabályai, 12th edition, the Helyesírás portal’s research on forms of address, and the Nominal Structures of Uralic Languages database for targeted grammatical distinctions. Examples apply those distinctions to interface writing.

Czech checks use the Internetová jazyková příručka for spelling, grammar and typography, and CzechEncy on formal address. The examples distinguish accepted variants from house preferences and preserve message meaning.

Slovak checks use selected Slovak spelling rules, the language institute’s Kultúra slova archive, Unicode CLDR plural rules, and NIST unit notation. These targeted checks preserve Slovak distinctions from Czech.

Slovenian checks use the Fran Ramovš Institute’s Jezikovna svetovalnica for selected grammar, spelling and notation distinctions, and Unicode CLDR for plural categories. Examples apply these checks to interface writing.

Croatian checks use the Institute of Croatian Language’s Hrvatski pravopis, Hrvatska školska gramatika and Mrežnik for selected spelling and grammar distinctions. Unicode CLDR supplies plural-category checks.

Bosnian checks use selected rules from Pravopis.ba, Hurija Imamović’s study of modal and phase-verb complements, and Unicode CLDR plural categories. These checks preserve Bosnian notation choices and distinguish grammatical alternatives from house preferences.

Serbian (Cyrillic, Bosnia and Herzegovina) guidance adapts Microsoft’s locale summary and Mozilla-derived shared review material. Selected checks use the Serbian Language Institute’s standardization decision in Naš jezik LVI/2, pages 97–99, CLDR 48 Serbian locale data, plural rules, and the OpenType localized-forms specification. RFC 1034 supplies the DNS expansion. These references support selected linguistic and technical distinctions, not every product example.

The Mozilla Serbian community guide, contributed by Marko Andrejić and Marko Kostić with community support, explicitly carries Creative Commons Attribution-ShareAlike 4.0 International. It acknowledges “Гномов правопис и језик превода (2011)” as a model. Its adapted guidance on Cyrillic script, formal lowercase address, retained names, capitals and abbreviations in Serbian for Serbia retains CC BY-SA 4.0, including our contributions to that adaptation. The adaptation condenses the advice, omits Pontoon account instructions and screenshots, and uses house context and descriptors for names. This notice applies to that attributed material, not unrelated works.

Serbian measurement checks use the Directorate of Measures and Precious Metals’ notation rules and unit definitions, plus NIST’s binary-prefix definitions. Serbian locale and glyph checks use the CLDR and OpenType references above; ISO’s account supplies its name and short-name distinction.

Serbian Latin guidance adapts Microsoft’s Latin-script summary and the Mozilla-derived shared localization-review material. Selected character, number and sorting checks use CLDR 48 Serbian Latin data, its collation imports and the imported collation rules. The Serbian standardization, plural, measurement, DNS and ISO references above also inform this page. Shared collation rules do not imply interchangeable language varieties. The separate Mozilla Serbian community guide’s Cyrillic-script instruction is not applied to the Latin-script page.

Macedonian guidance adapts Microsoft’s summary and the Mozilla Macedonian community guide. The Mozilla-derived material is distributed under CC BY-SA 4.0. That adaptation condenses its standard-language, consistency and meaning-based approach, omits contributor-tool instructions and adds house examples. Selected spelling checks use the Institute for Macedonian Language and Kultura’s 2017 orthography, rules 162–165 and 325, and the official Macedonian dictionary. CLDR 48 Macedonian data and plural rules support character, number and formatting checks. These targeted readings do not replace review of complete translations.

Bulgarian guidance adapts Microsoft’s summary and Mozilla-derived shared review material. Selected checks use the Institute for Bulgarian Language’s guidance on full and short articles, count forms, occupational nouns, polite address, уебсайт and quotation punctuation. CLDR 48 Bulgarian data, its inherited root percent pattern and plural rules support selected character and formatting checks. OpenType’s localized-forms reference above and NIST’s unit checklist support the glyph and measurement distinctions. These targeted checks do not certify complete translations.

Romanian guidance adapts Microsoft’s summary and the Mozilla Romanian community guide. The Mozilla-derived adaptation is distributed under CC BY-SA 4.0; it condenses meaning, diacritic, context, token and count guidance, distinguishes the community’s register and accelerator conventions from house choices, and updates older borrowing rules. Selected spelling and grammatical checks use the Romanian Academy’s language institute publication DOOM3, especially its graphic signs, introductory recommendations and dictionary guidance. CLDR 48 Romanian data, inherited root delimiters and plural rules support selected character and formatting checks. NIST’s unit guidance and RFC 1034 above support the technical distinctions. These targeted checks do not certify complete translations.

The brief Arabic, Bambara, Cebuano, Esperanto, Hiligaynon, Tagalog, Toki Pona and Yoruba supplements adapt Mozilla and the respective localization communities’ guides: ar, bm, ceb, eo, hil, tl, tok and yo. Their adapted material is distributed under CC BY-SA 4.0. The adaptations condense selected advice, link shared review rules, add house examples and omit incomplete templates, mismatched-language passages and unsupported blanket rules. The Toki Pona supplement retains the community’s nasin pu baseline as an editorial choice. These are limited supplements, not complete language policies.

Selected character checks use CLDR 48 Bambara, Esperanto and Yoruba exemplar data. Esperanto verb-form checks use Bertilo Wennergren’s PMEG discussion of the U-form; PIV is linked as a word-usage resource. These selected checks do not certify complete translations.

Bodo, Dogri in Devanagari, Kashmiri in Devanagari and Maithili guidance adapts the respective Microsoft localization summaries and their supplied full PDF guides, with Mozilla-derived shared review material distributed under CC BY-SA 4.0 with the source notices acknowledged above. Selected original PDF pages were visually checked to repair damaged extracted text. The adaptations preserve local address, abbreviation and script distinctions, correct technical examples, and omit corrupted or irrelevant supplier marketing passages. Named dictionaries in the source bibliographies were not all independently consulted.

Character checks use the Unicode Standard’s Devanagari discussion, particularly sections 12.1.4–12.1.5, and selected CLDR 48 data for Bodo, Dogri, Kashmiri in Devanagari and Maithili. Unicode supports the linguistic-apostrophe, avagraha and independent-vowel distinctions; CLDR exemplars are not complete orthographies. The NIST unit guidance and RFC 1034 cited above support notation and DNS corrections. These targeted checks do not certify complete translations or all locale formatter behavior.

Sanskrit, Manipuri in Bengali script, Santali in Ol Chiki and Sindhi in Devanagari guidance adapts the respective Microsoft summaries and selected pages of their supplied PDF guides. Mozilla-derived shared review material is distributed under CC BY-SA 4.0 with the source notices acknowledged above. The adaptations repair damaged extracted examples, preserve script and address distinctions, correct changed values and character identities, and omit unreliable supplier marketing and grammar examples. Source bibliographies were not all independently consulted.

Selected Sanskrit pronoun checks use Learn Sanskrit Online and A. K. Aruna's Sanskrit Grammar, section 7.6, to distinguish number, register and verb agreement. Script checks use Unicode 17's Devanagari discussion, especially section 12.1.5 on Sindhi consonants, and its Meetei Mayek and Ol Chiki discussion, especially section 13.10.1 on Ol Chiki marks and punctuation. The NIST reference above supports unit notation. These targeted checks do not certify complete translations or every locale formatter.

Malagasy, Somali and Kashmiri in Perso-Arabic guidance adapts the respective Microsoft summaries and selected supplied PDF pages, plus shared Mozilla-derived review material. Malagasy also adapts the Mozilla Malagasy community guide, under the adaptation terms and source notices above. The adaptations preserve local usage, distinguish commands from labels and states, correct changed roles and quantities, and omit unreliable templates and damaged long examples. Supplier bibliographies were not all independently consulted.

Selected Malagasy word checks use the Malagasy Dictionary and Encyclopedia, including ianao, izy, mpiasa and mpampiasa. Somali checks use Morgan Nilsson's Learner's Somali Grammar (2025), especially sections 2.1–2.4 and 8.1.2. Kashmiri character checks use Unicode 17's Arabic discussion; mixed-direction layout follows W3C's inline direction guidance. NIST's cited unit reference supports notation. These selected checks do not certify complete translations or every formatter.

Kashmiri examples use Noto Naskh Arabic, distributed under the SIL Open Font License 1.1. The bundled Arabic subset is unchanged from Google Fonts.

Product and service names identify the products of their respective owners.

Basque guidance adapts Microsoft's Basque summary and selected supplied PDF pages, with Mozilla-derived shared review material under the adaptation terms acknowledged above. The adaptation corrects changed actions, states and quantities, preserves contextual grammar and abbreviations, and replaces unsupported blanket rules with task-specific review. Selected word checks use Elhuyar's entries for erabili, aukeratu and guraso, plus Labayru's senide. Number checks use Euskaltzaindia's rule 203, sections 4.2 and 7.1. Its supplier bibliography identifies further grammar, terminology and EIMA references; these were not all independently consulted. Selected readings do not certify complete translations.

Xhosa and Zulu guidance adapts Microsoft's respective summaries and selected supplied PDF pages, plus selected sections of the Mozilla Xhosa and Zulu guides. Mozilla-derived material follows the source notices and adaptation terms above. The adaptations preserve grammatical and task distinctions, correct changed actions and technical identities, and qualify inconsistent number examples and unsupported blanket rules. The complete Mozilla guides and supplier bibliographies were not all independently reviewed.

Selected Xhosa noun-class checks use SADiLaR's ALGraP noun discussion. Selected Zulu title and borrowed-word checks use the national language board's 2021 orthography, PDF pages 3 and 202. Number separators were checked in CLDR 48 Xhosa, Zulu and inherited root data. South African phone formatting uses sections 13 and 15 of the numbering regulations hosted by ITU. RFC 1034 above supports the DNS correction. These selected readings do not certify complete translations or every runtime's locale behavior.

Sesotho sa Leboa and Setswana guidance adapts Microsoft's respective summaries and selected supplied PDF pages, with Mozilla-derived shared review material under the source notices and adaptation terms above. The adaptations correct character codes, units, lost placeholders and action distinctions; preserve local calendar and grammar examples; and qualify inconsistent formatting and blanket grammatical rules. The supplier bibliographies were not all independently consulted.

Selected Sesotho sa Leboa spelling and calendar checks use PanSALB's Melao ya Mongwalo le ya Mopeleto, rules 34–35 and 109. Setswana grammar checks use SADiLaR's noun classes and verb inflection. Unicode's Latin Extended-A chart supports the caron-letter correction. CLDR 48 Sesotho sa Leboa, Setswana and inherited root data support selected formatting comparisons. The NIST unit and ITU numbering references above also apply. These targeted checks do not certify complete translations, runtime formatters or every grammatical example.

Hausa and Igbo guidance adapts Microsoft's respective summaries and selected supplied PDF pages, plus the archived Mozilla Hausa and Igbo community guides. Mozilla-derived material follows the source notices and adaptation terms above. Both community guides were read; imported noun-class, Afrikaans, unit and resource-format templates were excluded. The adaptations preserve local examples and correct character identities, ownership, warning certainty, task scope, dynamic counts and units. The supplier bibliographies were not all independently consulted.

Selected Hausa checks use UCLA's writing discussion, noun modifiers and verb tense discussion; indexed text was available when direct retrieval failed. Igbo vowel, pronoun and word-division checks use the Onwu Committee's 1961 spelling notes, PDF pages 2–5. CLDR 48 Hausa, Igbo and inherited root data support selected calendar and number comparisons. The NCC numbering notice in ITU Bulletin 1280 identifies the obsolete Nigerian fixed-line patterns. The cited NIST and RFC references support unit and DNS distinctions. These targeted readings do not certify complete translations or actual-product locale behavior.

Swahili and Wolof guidance adapts Microsoft's respective summaries and selected supplied PDF pages, plus the archived Mozilla Swahili and Wolof community guides. Both community guides were read in full; their adaptations follow the source notices and terms above. The pages preserve local examples while correcting damaged characters, grammatical templates, changed actions and quantities, obsolete contact claims and inconsistent units. Supplier bibliographies were not all independently consulted, and source calendar forms are distinguished from CLDR data.

Selected Swahili grammar checks use the University of Georgia's KIKO Unit 1 grammar and Unit 4 time glossary. Wolof checks use selected pages of the Peace Corps Senegal manual, including possession, imperatives, aspect and currency counting, and a republication of the 2005 orthography decree. The republication has transcription defects and is not a current official legal-status check. CLDR 48 Swahili, Wolof, root and supplemental data support selected calendar, separator and currency-precision comparisons. TCRA's postcode guidance, its numbering notice in ITU Bulletin 1304, and the ITU-hosted Senegal and Mali numbering notices support contact-pattern corrections. These targeted readings do not certify complete product translations, current number allocations or actual runtime formatting.

Catalan and Valencian guidance adapts the respective Microsoft summaries. No locale-specific Mozilla community guide was present at the supplied archived collection paths, so the summaries' generic Mozilla-derived review material is integrated only as shared review practice. The adaptations preserve locale-specific vocabulary and accepted variants while removing unsupported blanket grammar rules. Selected source PDF pages were visually checked; supplier bibliographies and complete external grammars were not all independently reviewed.

Catalan orthography checks use the Institut d'Estudis Catalans' 2021 orthography update and its verb paradigms. Valencian variant checks use the Acadèmia Valenciana de la Llengua's report on Generalitat language criteria and the Softvalencià proposal, treated as an editorial proposal rather than an authority. Ordinal checks use Optimot and the EAPC's ordinal-abbreviation guidance. These targeted readings do not certify complete translations, every accepted regional form or runtime formatter behaviour.