On January 24, after nearly four years of hiatus, version 1.0.0 of the header-only SIMD-optimized library for C++ was released GLM (OpenGL Mathematics), based on specifications GLSL (pdf) (OpenGL Shading Language).
Changes:
- The module GLM_EXT_scalar_reciprocal has been added with tests;
- The module GLM_EXT_vector_reciprocal has been added with tests;
- The module GLM_EXT_matrix_integer has been added with tests;
- Functions glm::iround and glm::uround have been added to modules GLM_EXT_scalar_common and GLM_EXT_vector_common;
- The function GLM_FORCE_UNRESTRICTED_FLOAT has been added to prevent static assertions when using other scalar types with a function expecting a float type;
- A constexpr classifier has been added to the cross and dot functions;
- An incorrect assertion for glm::min and glm::max has been fixed;
- The orientation of quaternions in the glm::decompose function has been corrected;
- Singularity when converting quaternion to Euler roll angle has been fixed;
- The operation of glm::pow with small quaternions has been fixed;
- A compilation error in glm::fastNormalize has been fixed;
- A compilation error in glm::isMultiple has been fixed;
- Computation in the glm::adjugate function has been fixed;
- The sign truncation of the result in the glm::angle function for angles in the range (2pi-1, 2pi) has been fixed;
- The prohibition of using glm::string_cast in CUDA host code has been removed;
- Github Actions have been added.
Source: linux.org.ru
