Monotone floating point property

Question

I was reading a book. The book says that -

with the addition of a floating point satisfies the following monotonicity property:
if a>=bthen (x + a) >= (x+b)for any value a, band x, except NaN. This property of real (and integer) addition is not subject to unsigned or two additions to the complement.

How does a floating point obey it?
Why is an addition without a sign or two additions not subject to it?

Answer 1

C , .. . , 1 , , , , .

, , C undefined, .

C, IEEE745, ( , NaN, NaN ) , : , , , , . , , .