Just adding a float cast (since a float is being assigned to)
allows MSVC to compile and should not affect other compilers.