• Added support for Daikanwajiten (Morohashi) references.
• Added support for Korean and Vietnamese readings.
• Added xs and small popup font size options.
• Fixed handling of hidden text.
• Fixed pop-up being unscrollable in some cases (see.
• Worked around a major string substitution bug in Safari.
• Fixed word lookups splitting ruby text incorrectly, while still allowing splits around center dots (・).
• Fixed the popup showing for number-only matches in the name dictionary (e.g. 64).

• Fixed handling of long alt text.
• Made romaji readings correspond to actually shown romaji.
• Added more graceful handling of empty data files.

• Added references for The Kodansha Kanji Dictionary.
• Improved scanning of ruby transcription text.
• Added vertical reading mode to puck, allowing puck earth to orbit on the sides rather than just the top or bottom.
• Revert changes from PR #1773. Puck can hold popup open again.
• Made popup work for Notion popups like the calendar.
• Fixed handling of multi-line text areas in Chrome/Edge due to Chromium bug.
• Fixed showing of title field when looking up alt and title attributes.

• Added support for looking up subtitles in Plex video content.
• Added Shift as a possible keybinding for hold to show.
• Fixed a bug when displaying Nelson radicals.
• Fixed a bug that prevented the flash animation after copying an entry from playing.
• Fixed a bug where changing the font size in the settings wouldn't apply to the Kanji and Names tabs.

• Fixed the popup getting stuck to the top-left in Facebook.
• Switched kanji data to new schema so we can show subcomponents.

• Added precise conversion of Japanese era dates, including those preceding the adoption of the Gregorian calendar in Japan, with accurate calculations for the timespan of specific years and months (e.g. 元治元年, 慶応4年閏4月, 明治5年12月2日).
• Fixed sorting of word results (see.
• Added deinflection for -得る.
• Made the kanji references table more responsive.
• Fixed displaying metadata in the names tab.
• Added support for characters in the Enclosed Ideographic Supplement block (e.g. 🈀, 🈁, 🈴).

• Added support for irregular verb forms of 揺蕩う.
• Made tap detection more reliable.
• Added handling for metric suffixes on currencies (e.g. 40k円) as well as "yen" suffixes (e.g. 200k yen) and "JPY" prefixes (e.g. JPY 1,000).
• Added handling for 戶 and 內 kyūjitai.
• Added deinflection for additional forms of -sugiru.
• Added deinflection for irregular forms of vs-s and vz class suru verbs.
• Removed deinflection of vs-c class verbs, as they are not used in modern Japanese.
• Fixed unreasonable matches caused by duplicates in the deinflection reason chain.
• Fixed sorting of deinflected results in some cases (e.g. 見とれる).
• Fixed sorting when looking up kana in some cases (e.g. なる).

• Added stroke animations to the kanji tab!
• Fixed handling of shadow DOM content in Firefox Nightly.
• Fixed matching of 0年.

• Extended the support for honorific speech.
◦ Added imperative for くださる
◦ Added addtional forms of ます
◦ たら・たり and て-form
◦ Imperative
◦ Added respectful speech
◦ Continuous form (ていらっしゃる, てらっしゃる)
◦ なさる
◦ になる
◦ Added humble speech
◦ Continuous form (ておる, とる)
◦ 致す
◦ Fixed recognition of irregular verb forms of 宣う
• Added support for irregular verb forms of 給う.
• Fixed recognition of continuous forms of verbs with irregular te form.
• Added supported for doing unit conversions to imperial units
• Added links from WaniKani annotations in the words and kanji tabs to the corresponding page on the WaniKani homepage.
• Fixed a condition that could cause some preferences to be overwritten causing, for example, the puck to suddenly disappear.
• Fixed duplicate matching of names with both 新字体 and 旧字体.
• Fixed display of matches on search-only forms so they don't appear as part of the entry.

• Added an option to use system fonts instead of bundled fonts.
• Fixed verb class tags being shown on proverbs.
• Fixed kanji incorrectly encoded as radicals not being displayed in kanji tab.

• Made it possible to change tabs in the popup by swiping horizontally.
• Bundled fonts used by the popup to provide consistent display and avoid issues with sites that patch system fonts.
• Changed puck behavior to no longer hold the popup window like a regular mouse event would.
• Made copying to the clipboard include rare headwords again when not using simplified copy mode.
• Made the popup no longer show when tapping on text boxes and other interactive elements.
• When searching using just kana (e.g. し and ヤク), made kana matches be prioritized.
• Fixed sorting between various deinflected results so that 進む appears before 進ぶ when looking up 進んでいます.
• Made the popup update after the database update is complete.
• Made senses no longer be restricted when matching on search-only headwords.
• Fixed translation of odd numbers with digits and multiple powers of ten like 11,786百万円.

• Made the puck persist its position, orientation, and active state between pages.
• Updated the clipboard copy feature:
◦ Added a new setting to copy a simplified version of the entry that may be more suitable for making flashcards etc..
◦ Definitions are now separated by line breaks.
◦ Rare headwords are excluded from the result unless we matched on them.
• Fixed a case of 10ten not looking up text with ruby correctly in Ttsu reader.
• Stopped popup being inverted when Wikipedia's experimental dark mode is enabled.

• Improved rendering of source language information.
• Added localization and handling for new rk (rarely-used kana form) tag.

• Added option to show Bunpro vocabulary and grammar labels next to words.
• Added localizations for newly-introduced field tags.
• Fixed a bug where it would take two clicks to enable the add-on sometimes.
• Fixed slow lookup for long text spans.
• Fixed options page becoming disconnected from the background page (again).

• Enabled non-persistent event pages.
• Lookup fixes
◦ Fixed recognition of words that end in half-width numerals like 小1.
◦ Added parsing for ill-formed numbers like 39,800万円.
◦ Fixed lookup misbehaving on Firefox with <textarea>s with display: block such as on pastebin.com.
◦ Made matches on search-only headwords show the search-only version too (e.g. 磨ガラス, ペイチン).
◦ Hid the "usually kana" annotation when no kanji headwords are shown (also.
◦ Added ability to recognize (deinflect) a few more irregular verbs like ございません and くださいます.
• Keyboard fixes
◦ Added handling to avoid the x key being assigned to both closing the popup and expanding it.
◦ Fixed the toggle key setting to allow using ",", ., Home, End, PageUp, PageDown, Space, Insert, Delete, Up, Down, Left, and Right.
◦ Made it possible to clear the toggle key.
• Other bug fixes
◦ Added a workaround for a Firefox bug that would cause 10ten to hang when changing dictionary language.
◦ Fixed the popup changing width in some cases (e.g. when tabs are hidden and scrollbars are configured to always show).
• Android fixes
◦ Made the lookup puck handle taps and double-taps correctly on Firefox for Android.
◦ Hid settings button in popup on Firefox for Android until bug 1795449 is fixed.

• Fixed selection styles not applying in some browsers.

• Collapsed the popup window to only show the most relevant results by default.
◦ You can restore the old behavior by choosing "Expand all entries → Words" from the settings screen.
◦ You can view other entries by scrolling the popup window or by expanding it to show all entries by clicking the downward triangle button or pressing <kbd>x</kbd> on the keyboard.
• Made the popup show kanji details for all kanji in the longest matched word.
• Added a setting for changing the font size of the popup.
• Made options page support dark mode.
• Added an option to disable looking up by tapping on them on touchscreen devices.
• Added WaniKani data to word and kanji display. This can be enabled from the options page under the "Popup style" and "Kanji dictionary" sections.
• Added handling for kanji incorrectly encoded as radicals.
• Fixed handling of pages with <frameset> elements and nested <iframe>s.
• Fixed highlight styles on pages with strict CSP settings.

• Fix keyboard detection for some devices.

• Rewrote text lookup to better handle cases of hidden content (including user-select: all content), avoid lingering popups, and improve performance.

• Fixed breakage on Firefox beta.

* Made Rikaichamp better handle situations where there is limited disk space
* Fixed handling of YouTube subtitles with ruby

• Moved display of part-of-speech labels to the start of the definition and grouped them when they are repetitive
• Made Rikaichamp include する in the match range for nouns that take する (vs nouns)
• Fixed recognition of vs-c verbs (suru verbs ending in す instead of する, e.g. 兼した)
• Stop trying to convert years written as a transliteration of digits (e.g. 令和七九年 instead of 令和七十九年)
• Moved Rikaichamp popup to an isolated shadow DOM tree so that it should be more robust when used on pages that set styles in ways that conflict with Rikaichamp's styles

• Made all kanji headwords be shown for an entry, not just the matching ones
• Added recognition of full-width alphanumerics (e.g. ８月, Ｂ級グルメ)
• Allowed Rikaichamp to recognize characters in the CJK Unified Ideographs Extension B range (e.g. 𠏹沢)
• Tweaked name preview feature to match names that start with hiragana (e.g. ほとけ沢)
• Support for recognizing き inflection of i-adjectives
• Fixed an issue with keydown listeners not being unregistered causing the "toggle definitions" feature to misbehave
• Added a few missing word entries (e.g. 印鑑)
• Fixed popup window language tagging so the Chinese localization works better
• Localized a few hardcoded English strings

- Fixed an issue with the SVG star being too large when upgrading and the old stylesheet is still in effect