Fun Fact

[Greg McKaskle]

Equally fun fact:
When you compare anything and one of the elements is NaN, the result is false. Nan > 0 is false and Nan < 0 is false and Nan == 0 is false.

And it isn't just LV. This is what IEEE 754 says to do. And it makes sense, but at the same time will surprise you at times when you least expect it -- possibly causing shooters to misbehave, wormholes to form, etc.

Greg McKaskle

[Alan Anderson]

If you do decide to experiment with exploiting wormholes, remember that game pieces from past or future games should not be allowed to enter the playing field. There's no rule against it yet, but R01 is always in force.

[Kevin Sevcik]

Originally Posted by Greg McKaskle

Equally fun fact:
When you compare anything and one of the elements is NaN, the result is false. Nan > 0 is false and Nan < 0 is false and Nan == 0 is false.

And it isn't just LV. This is what IEEE 754 says to do. And it makes sense, but at the same time will surprise you at times when you least expect it -- possibly causing shooters to misbehave, wormholes to form, etc.

Greg McKaskle

Corollary: The smart programmer uses this fact to some advantage. If there's a possibility of you getting a NaN as result, then make sure your false case is what you'd want to run if you're processing a NaN. Makes things a lot more reliable.

[sur]

Originally Posted by Greg McKaskle

Equally fun fact:
When you compare anything and one of the elements is NaN, the result is false. Nan > 0 is false and Nan < 0 is false and Nan == 0 is false.

And it isn't just LV. This is what IEEE 754 says to do. And it makes sense, but at the same time will surprise you at times when you least expect it -- possibly causing shooters to misbehave, wormholes to form, etc.

Apparently this is how IsNaN is implemented in C#: http://stackoverflow.com/a/1145462.