Make NaN != NaN

[coder0xff]

[ProphetLamb]

Not desirable because it would impact performance on all equality checks, because at least a check of the biasedexponent is required.

[Willt125]

Not desirable because it would impact performance on all equality checks, because at least a check of the biasedexponent is required.

It is standardized IEEE 754 behavior. It is not only desirable, but mandatory for compliance, if perfect compliance is the goal.

However, it could be an opportunity to fix mistakes of the committee, if that is desired.

Edit: I realize I poked a dead issue, but I'm going to leave this here for posterity and informational purposes

[coder0xff]

It's an open issue, so I'm glad you commented. I didn't make the change because it seemed to have no support until now.