Code:
float a=.7;
(.7f>a)?printf("Yes\n"):printf("No\n");
by default .7 is stored as double that are accurate upto 14 decimal points, but float is accurate upto 6 decimal place only,
so .7(double) > a(float) /* Result is yes */
and .7f(float) > a(float) /* Result is no because both are equal now */