>I won't need math, so I'll blindly trust this 20 year old 32 bit half precision floating point thing someone coded up in an afternoon before their holiday brea

>I won't need math, so I'll blindly trust this 20 year old 32 bit half precision floating point thing someone coded up in an afternoon before their holiday break. Don't get me started on engineers who think floating point numbers can be treated like real numbers for equality testing.