Version 2.3.1

• Fix missing macro error when deprecation is disabled

Version 2.3.0

• Added missing 2D/4D variants of some vector arithmetic
• Added support for Windows min/max macros
• Added vector_get_component/vector_set_component
• Added matrix_get_axis/matrix_set_axis and component versions
• Added qv/qvs/qvv from matrix
• Improved type traits
• Improved code generation for float64 types
• Added camera related matrix utilities
• Deprecated scalar/vector coersion for some arithmetic functions in favor of as_scalar and as_vector suffixes
• Other improvements and cleanup

Version 2.2.1

• Fix matrix 4x4 multiplication and add unit tests

Version 2.2.0

• Added support for Clang 15
• Added support for GCC 12 and 13
• Added support for XCode 14
• Added support for VS2022
• CI now uses docker where possible
• Updated Catch2 to 2.13.7
• Tons of static analysis fixes and polish
• Added new QV transform type
• Added new QVS transform type
• Tons of new minor additions
• Tons of optimizations, especially on ARM64
• Many other improvements and cleanup

Version 2.1.5

• Force macro expansion in version namespace identifier

Version 2.1.4

• Add support for clang 12, 13, and 14
• Add support for GCC 11
• Add support for XCode 12 and 13
• Add support for Arm64 development on OS X and Linux
• Misc CI improvements

Version 2.1.3

• Add versioned namespace to allow multiple versions to coexist within a binary

Version 2.1.2

• Fix MSVC/clang static analysis warnings
• Improve MSVC ARM64 intrinsic support detection

Version 2.1.1

• Avoid enabling NEON when just __arm__ is present

Version 2.1.0

• Force inlined most functions to improve MSVC code generation
• Migrated from Travis to GitHub Actions
• Added support for clang 11
• Lots of minor additions, fixes, and improvements