what is consequences in arithmetic operations with normalized floating point and pitfalls in computing?

Question

What is difference between them?

Answer 1

One important consequence is that it is not possible to exactly represent some numbers. Where this often shows up is in comparisons. Sooner or later you will get into trouble if you write something like

if (fValue == 3.025)

Instead you have to write something like

if (abs(fValue-3.025) < fThreshold)

where fThreshold is some small value.