Textbook publishers offer on-line computer homework and assessment educational materials. Such materials abound with positive and negative features. Some of the strengths are immediate feedback, unlimited practice, and randomized variables. One weakness is that students adopt a “trial-and-error strategy” instead of understanding the problem. Of special concern, for static and randomly generated exercises, are the real difficulties in assessing student answers. Overly strict methods generate complaints, while lax schemes may give credit for incorrect work. In this roundtable, we share our joint experiences. Our goal is identification of practices that we should adopt and those we should avoid.