Big Sur Compatibility

• Signing and notarization for Big Sur and later.
• Fix for #61: mapping overrides take effect in App UI.

Feature enhancements

• Ability to manage languages; rename, disable and reorder them
• Shortcut support for changing languages in IME
• Cleaner and more functional UI for LipikaApp
• Ability to export and import custom mappings
• Deprecated SCM format in lieu of above

General Availability v2.1

• Fix for issue #50
• Automatic session improvements

Stable Beta Release

• Major overhaul of Engine algorithm to fix ratreya/lipika-engine#1
• Fixed major memory leak in Engine due to strong reference cycle
• Defensive code when interfacing with system IMK
• Fixed a whole class of crashes (#47, #48 and #49)
• Fixed issue with system bar menu not updating correctly

Completely redesigned v2.0 [Alpha Release]

• Swift instead of Objective C - lines of code reduced from 6395 to 4407 (~45%)
• Broke out lipika-engine into a separate repository
• Uses an order of magnitude less memory
• Comes with a companion app called LipikaApp
• LipikaApp allows transliteration between any combination of Indic scripts and schemes
• Built-in mappings can be fully edited using LipikaApp
• Ability to start a new transliteration session from existing words
• Reverse-transliteration with 100% fidelity
• Correct ordering of ZWNJ which enables one to type Kannada words such as ಸೂರ‌್ಯ
• Support for Marathi Chandra or Ardha Chandra
• Ability to escape transliteration: `3`netra -> 3नेत्र
• Numerous bug fixes and performance improvements

Minor Improvements

• Minor fixes to Grantha and ISO-15919
• Fix for candidate selection with input string (#26)
• Preserve ordering of mappings

Transliteration Fixes

• Nuqta for all languages - Issue #19
• Minor fixes to Grantha and IPA
• Complete ITRANS mappings
• Consolidated Tamil scripts
• Updated code to Mac OS X 10.12

Improved Reliability

• Bug fix for issue #14
• Ported tests to work on Xcode 7.0
• More debug logging

Indic Support

• Highly maintainable proprietary scheme format
• Schemes standerdized for applicability to all Indian languages
  • ITRANS, Baraha, Harvard Kyoto, Barahavat and Ksharanam
  • Minimized exceptions for individual language
• Unicode mapping to all major Indian language scripts
  • Bengali, Devanagari, Gujarati, Gurmukhi, Hindi, Kannada, Malayalam, Oriya, Tamil and Telugu
  • Standerdized naming for consistent transliteration