Draft: Fix #156: NaN comparisons incorrect
Don't convert <= and >= to > and < via the source transforms.
Instead add vops to handle <= and >=.
This requires adding vops both for fixnum arithmetic and floating-point arithmetic.
Don't convert <= and >= to > and < via the source transforms.
Instead add vops to handle <= and >=.
This requires adding vops both for fixnum arithmetic and floating-point arithmetic.