> After reading the paper, I'm inclined to change the definition of UB in D to not mean it can be assumed to not happen and not be unintended.

What's the current definition of UB in D?