The Changelog

This changelog covers only the modern WebExtension, beginning with version 2.0.0. It does not include the release history of the legacy Mozilla XUL Add-On (versions 1.x).

• 2026-08-04 ver 2.0.0 -- initial WebExtension release.
• 2026-08-16 ver 2.0.1 -- fixed Ancient Lexicons lookups for pointed Hebrew text.
• 2026-08-17 ver 2.0.2 -- updated the privacy policy.
• 2026-08-17 ver 2.0.3 -- increased lookup history size to 64 entries.
• 2026-08-24 ver 2.1.0 -- improved Cyrillic-language translation routing, added an option to force Cyrillic-to-Russian translation, and improved Google AI translation availability handling.
• 2026-08-24 ver 2.2.0 :
  • added a Google AI Mode browser-tab action for Google Search results.
  • added language-neutral word segmentation for unspaced scripts, preserving punctuation and choosing the appropriate Google AI or ordinary Google Translate route.
• 2026-10-09 ver 3.0.0 :
  • added direct SyriacDictionary results between Syriac and configured English or Arabic sources, and made Translate to Aramaic independently enable English-to-Syriac results.
  • added Forvo pronunciation lookups for any one-to-three-word query and improved their result positioning.
  • refined Google AI and Google Translate routing for mixed-script and Cyrillic text, so unreliable non-Russian Cyrillic detection is not treated as Russian.
  • added optional Google AI etymology requests and automatic Show More expansion, with fallback to ordinary Google Translate when Google AI is unavailable.
  • added a circular progress indicator while enabled lookup sources are being checked.
  • fixed Etymonline lookups for the provider's current search result pages.
  • improved Arabic transliteration of hamza, shadda, and final-letter handling.
  • fixed input dialogs to restore normal zoom after opening while preserving their classic 400x200 size.
  • added persistent zoom controls for the transliteration pane.
  • moved transliteration settings into a dedicated Translit Options tab.
  • fixed Wikipedia and Wiktionary lookups to keep original Arabic input and respect enabled Syriac-to-Hebrew compatibility routes.
  • made package builds print the current date.
  • added selectable transliteration-only, original-only, and combined display modes that keep every word clickable.
  • removed redundant Options headings and made Reset to Defaults available in the Other Options tab.
  • added configurable Living Arabic lookups across Classical and eight dialect dictionaries, with selectable root, word, meaning, and example searches.
  • refined Lookup Sources with two-column language options and a bordered Living Arabic group with a clickable title.
  • added a transliteration option for whether Hamza sitting on Alif is rendered with Hataf Patah.
  • improved one-word dictionary lookups by retrying same-script stem variants across lookup sources, while keeping multi-word text exact.
  • refined Google translation routing: ordinary Google Translate retains automatic source detection alongside enabled source-language overrides, while word lookups can use their paragraph's language context and Syriac is sent in its original script.
  • made lookup history show original input and script instead of transliterated labels.
  • improved initial result positioning for Forvo and SyriacDictionary.
  • added recovery controls for Google verification and rate limits, including retry, browser-tab, and skip actions.
  • improved Arabic transliteration of hamza, Alef Madda, honorific ligatures, contextual prepositions, and vocalized definite articles.
  • kept preview zoom independent for each lookup source, preserved direct preview refreshes, and prevented browser pinch zoom from changing result views.
  • made the browser-tab action open a preview's current navigated page while preserving the original POST submission when it has not navigated.
  • excluded pure dictionary sources from lookups longer than three words while keeping translation and combined sources available.
  • made Google AI results retry temporary failures and finish expanding when Google's Show More control stalls.
  • improved Arabic word analysis to distinguish final Ya from Alif Maksura and recover omitted Dagger Alif and article Shadda when the spelling supports them.
  • corrected Arabic transliteration of definite articles, silent Waw, Tanwin, missing Fatha before Alif, and repeated long-vowel marks.
  • added an option to render Hamza sitting on Waw with Hataf Qamatz, including in Hamza with Tanwin.
  • raised Shadda in transliterated Hebrew text for clearer display.
  • improved Hebrew-source routing for Syriac and Imperial Aramaic queries, including Daat and Hebrew Wikipedia.
  • recognized more temporary Google AI Overview failure messages so retries and translation fallback can run.
  • added the Use Shva for Hamza, WaW and Yud only option for Arabic transliteration while preserving reduced vowels and Shva produced by doubled consonants.
  • added a link to Principles of the Transliteration from the Translit options tab.
  • refined Hamza transliteration with Tanwin and following Waw or Yud, preserving long vowels and placing silent seat annotations before Hamza.
  • kept accusative Tanwin's supporting Alif silent and ignored standalone Maddah marks without changing Alif Madda handling.
  • corrected Alif and Yud vowel lengthening, including Dagger Alif on Waw and Shadda on Yud, and removed redundant long-vowel Alefs before vocalized Alefs.
  • improved definite-article analysis for ambiguous unpointed Arabic nouns and prevented an extra Fatha before Ta Marbuta when a preceding vowel is already present.
  • recovered lexical vowels for common Arabic particles, prepositions, and short nouns while preserving conflicting written vowel marks.
  • added recovery of missing Arabic long vowels using Hamza, Shadda, Tanwin, and Maddah evidence across the full original text, while preserving written vowels and pronounced initial Alifs.
  • recovered vowels on confirmed attached Arabic prepositions without guessing ambiguous word splits.
  • preserved a clicked word's displayed transliteration and original text context through new lookups, history, and preference refreshes.
  • kept each transliterated word together when text wraps onto another line.
  • kept automatic Google AI translation prompts and source labels language-neutral unless a source language is explicitly forced.
  • fixed Unicode information for supplementary characters, showing each character as one code point with correct hexadecimal references and secure character links.
  • fixed Sureth POST searches to stay in the result preview without opening an extra browser tab.
  • distinguished Syriac Marhetana and Mtalqana marks using written vowels and known word forms, showing reduced vowels or silent letters only when supported and leaving ambiguous pronunciation unresolved.
  • corrected Syriac Hbasa vowels to avoid an extra Yud, preserved letters bearing musical marks, and marked pronounced final Syriac He with Mappiq.
  • added optional Ancient Hejazi reading reconstruction, with research references and attribution in the Translit options tab.
  • added the option to treat word-initial Alef as always consonantal, enabled by default.
  • disabled replacement of Arabic Shadda by default.
  • improved Hamza and Alif Madda transliteration after consonantal Sukun and recovered missing pre-Hamza Sukun when supported by word analysis or spelling.
  • disabled Hebrew-to-Syriac compatibility lookups by default while keeping the option available.
  • fixed Iris previews to scroll to the requested article inside the provider's translation pane.
  • excluded Pshita results when CET redirects to its inactive-site page, without treating the redirect as a missing word.
  • clarified the lookup dialog with Enter a word or phrase and Look up wording.
  • expanded Principles of the Transliteration with updated Hamza, Alif Maqsura, and word-initial Alef explanations.
  • reduced spacing in the Translit options tab to accommodate its settings and research links.
  • switched ordinary Google Translate previews to the live desktop site, preserving loaded translations and scrolling below Google's fixed header.
  • made Syriac translation use the original Syriac text in Google AI at every text length, with automatic fallback to AI Mode when an AI answer is missing or disappears.
  • improved Google AI answer detection and verification recovery, and retried AI Mode language-availability refusals after removing Google's restrictive URL flag.
• 2026-10-10 ver 3.0.1 :
  • fixed Arabic-script transliteration so word-initial Ya written without dots remains pronounced after a preceding word, including in Persian names.
• 2026-10-14 ver 3.1.0 :
  • skipped redundant Google translations to Arabic for Arabic-script text unless a non-Arabic source is explicitly forced, distinctive non-Arabic letters are present, or another Arabic-script language is reliably detected, including language context inherited by clicked words.
  • corrected Arabic transliteration to omit silent Alif before Hamza on Ya or Waw when the written vowels support it, including in ordinary and Ancient Hejazi reading modes.
  • recovered missing vowels in Arabic words for one hundred and its plurals, dual forms, and compound numerals, including attached prefixes and suffixes, while preserving written case endings and leaving ambiguous readings unresolved.
  • improved Arabic word analysis to recognize attested broken plurals such as رِجَال, كُتُب, and مَسَاجِد, including their case forms, while retaining compatible alternative readings in unpointed text.
  • preferred the numeral 200 reading for unpointed مائتان and مائتين, recovering its vowels and omitting the silent Alif while preserving conflicting written vowel marks.
  • fixed the Use Shva for Hamza, WaW and Yud only option to respect its disabled state, while keeping Sukun on letters that serve only to lengthen a vowel suppressed independently of the option.
  • restored the Suppress comments setting in the Translit options tab and moved the online translation setting to Other.
  • added an option to ignore standalone Maddah, enabled by default; turning it off lets typed Alif plus Maddah be read as Alif Madda.
  • made separately typed Hamza marks behave like precomposed Hamza letters in Arabic transliteration, including Ancient Hejazi reading.
  • fixed Suppress comments to omit a silent Waw seat for Hamza without removing the Hamza or its vowel.
  • kept Arabic Tanwin attached once to the final letter before trailing punctuation and invisible marks.
  • read marked Quranic initials such as Alif Lam Mim by their letter names while leaving ambiguous unmarked spellings to ordinary Arabic transliteration.
• 2026-10-15 ver 3.1.1 :
  • added Yiddish translation routes for Hebrew-script text when the Yiddish source is enabled, including translation to Hebrew while retaining automatic routes.
  • used reliable Yiddish language detection for ordinary Google Translate and Google AI fallback, while keeping automatic Google AI prompts language-neutral and avoiding duplicate results.
• 2026-10-16 ver 3.2 :
  • removed redundant plain Shva from word-initial Alef in Arabic transliteration while preserving Hataf vowels, including the Shva+Hiriq notation.
  • clarified the Hamza options for short /a/, /u/, and /i/ vowels and added explanatory tooltips for Hataf Patah and Hataf Qamatz.
  • disabled Hataf Qamatz for Hamza followed by short /u/ by default, using Qubuts unless the option is enabled.
  • Refined Hamza Transliteration.
• 2026-10-17 ver 3.2.1 :
  • marked an unknown vowel after medial or final Hamza with an Arabic Hamza sign above Hebrew Alef (אٔ), while retaining Shva for explicit Sukun and leaving word-initial Alef unmarked.
  • preserved separate consecutive Hamza consonants when simplifying word-initial transliteration, including in Ancient Hejazi reading.
  • updated Principles of the Transliteration with examples distinguishing unknown Hamza vowels from confirmed absence of a vowel.
• 2026-10-17 ver 3.2.2 :
  • added support for all Arabic Presentation Forms-A and Forms-B in transliteration, including contextual letter forms, compound ligatures, and vowel marks.
• 2026-10-19 ver 3.2.3 :
  • added a Force translation from Japanese option for Google AI and Google Translate, including kanji-only words, while retaining automatic source-language routes.
  • restricted forced Japanese translation to text containing Han, Hiragana, or Katakana characters and prioritized default Google translations before additional forced-source requests.
  • quoted Japanese and Chinese text in Google AI translation requests, including Russian etymology requests, except when translating to Hebrew.
  • localized Google AI translation prompts and source-language names for Hebrew and Arabic, and used Russian prompts for ordinary Russian translation as well as etymology.
  • placed an Arabic-language instruction at the start of Arabic Google AI prompts, requiring the introduction, translation, and dictionary explanation to be written entirely in Arabic.
• 2026-10-21 ver 3.2.4 :
  • added separate Exact matches only options for all Wiki-tab providers and Wiktionary, enabled by default; turning them off includes full-text search results without changing Wikipedia translation.
  • added a button beside the transliteration zoom controls for entering new text in a modal dialog and updating transliteration and lookup results in the same window.
• 2026-10-21 ver 3.3.0 :
  • fixed the Ayn Ayn context-menu label to show selected text from the result window's transliteration pane.
• 2026-10-22 ver 3.3.1 :
  • improved Vasmer lookup to recognize Russian no-entry responses and continue checking query variants before excluding the result.
  • displayed changelog records newest first while preserving their version order in the source document.
• 2026-10-24 ver 3.4 :
  • improved Vasmer lookups to find accented headwords from unaccented queries using case-insensitive substring matching, and corrected the dictionary's displayed name to Vasmer.
  • increased the wait for Google AI translation answers from seven to ten seconds.
  • changed Arabic ḍād (ض) transliteration to use the traditional Judeo-Arabic dot above Tsadi (צׄ) instead of dagesh for clearer reading, and updated Principles of the Transliteration.
  • added an optional omission of dagesh from ungeminated Arabic ד and ת, including Ta Marbuta, while retaining dagesh for Shadda and Syriac Qushshaya; the option is disabled by default.