Your floating-point calculator won't give you infinite trailing nines, but it may very well give you something like 3.99999976158 (approximately the largest flo

Your floating-point calculator won't give you infinite trailing nines, but it may very well give you something like 3.99999976158 (approximately the largest float below exactly 4) or 3.99999999999999955591 (approximately the largest double below 4).