Existing "eventual" semantics for time can yield broad time ranges that may not accurately capture timing relations. For a sequence: A ##[t1:t2] B, the interval [t1:t2] may be too broad allowing B to move between true and false states many times in the defined interval. Furthermore, the computation of [t1:t2] is based on abstract interpretation techniques that weakens the interval. Is there a better way to reason about time using the present framework based on interval arithmetic?
Existing "eventual" semantics for time can yield broad time ranges that may not accurately capture timing relations. For a sequence: A ##[t1:t2] B, the interval [t1:t2] may be too broad allowing B to move between true and false states many times in the defined interval. Furthermore, the computation of [t1:t2] is based on abstract interpretation techniques that weakens the interval. Is there a better way to reason about time using the present framework based on interval arithmetic?