When considering edge cases (via custom tests) for the assessed lab, should I consider that the object references provided (for Union or Intersection classes, for example) as arguments when testing them could be null, or is that outside the scope of the assessed lab? Also, for the Range and the Union/Intersection classes, do they need to have it actually compute the values (in Arrays or something similar) or is it enough to ensure the two methods being implemented in each of them (describe, contains) have their logic work as expected and tested in the provided tests?