What do these operators do in D 2.0: <>= !<>= !<= !>=

Question

What do these operators do in D 2.0:

• <>=
• !<>=
• !<=
• !>=

Answer 1

They are used for values that could be unordered, such as NaN for floats and doubles. 1 <>= NaN evaluates to false, whereas x <>= y evaluates to true for any pair of numbers, as long as neither number is NaN. The other operators you mention work the same, mutatis mutandis.

Answer 2

The long answer:

When dealing with floating point, two values will compare as one of A<B, A=B, A>B or unordered (if one is NaN).

The operators represent every interesting (non constant) row in the truth table. They can be interpreted as testing true for each of the cases for which the operator has the corresponding char, unless it has ! in which case the value is inverted.