There is an ambiguity in Moore's use of "proper" in "proper length." Unlike "proper time," which is the difference in time between two events (the measurements of a clock at rest), the length of a rod is not the difference between two unique events. If I refer to a rod at rest and define "proper length" to be the distance between the two ends of the rod as measured simultaneously, those measurements are indeed two unique events. [The measured distance then corresponds to the "proper distance" as defined above by DaleSpam.] However, the confusion comes when this "proper distance" is used as Moore's "proper length," determined by measurements performed at arbitrary times. The length of the rod at rest is not a "proper" distance in the same sense as "proper" time.
So, what about cc(dt)^2-(dx)^2=cc(dt')^2-(dx')^2 ? Both dx and dx' can be equal to a proper length by Moore's definition. If they refer to the same rod then dx=dx', and the equation can only be true if dt=dt'. If dx and dx' are equal to the proper lengths of two different rods then dx may not equal dx', but the equation can still be true for special combinations of dt and dt'. But if the equation refers to a measurement of DaleSpam's "proper distance" between two events, then |dx|=|dx'|=ds and dt=dt'=0.