How would your recommendations change if you know that the triggers from the defects in Prob. 9.6 come from three different releases? For simplicity let us assume that all the system-test-triggered defects are from release 1 , the function-test-triggered defects from release 2 , and the others from release 3 . Now, if we were to make a recommendation for the development of release 4, how would it be different from those of Prob. 9.6?