bulletphysics/bullet3

btScalar functions are not consistently used

SirNate0 opened this issue · 0 comments

btScalar.h introduces a number of functions to operate correctly on btScalar regardless of single vs double precision. However, these functions are not used consistently in the library, as can be seen with a simple case-sensitive search of fabs, for example.

This isn't that big of an issue - it does still compile and work, after all. However, it does make it much harder to replace btScalar with something other than float or double, such as a fixed point number, so it would be nice if the btFabs and similar functions could be consistently used.