Stress has typified the last three months. Completing the testing stage of the product in just three months has drained most of your employees. Even though the testing stage is officially over, a pernicious bug remains in the code. In certain trial runs, KAREL fails to send out a message to turn patients over to the right; instead, KAREL requests that a patient be turned over three times to the left. The bug seems to occur infrequently.

After giving the matter much thought, you decide there are one of three routes you can take.

First, you can redo the code that deals with the flaw. Estimates tell you that rewriting the code may take three months.

Alternatively, you can document the flaw in the program, advising hospitals of possible problems with KAREL's instructions to turn patients. The already printed literature about Karel would have to be republished, requiring perhaps an additional month to reassemble the KAREL package. In the process, you can work on a patch to cover the program's flaw and release that as soon as it's developed.

But maybe you should just ignore the bug. After all, it only occurs in very rare situations; as a whole, the code seems fine. Besides, the extra time saved by assembling the product now will give you a greater advantage in capturing your market before LOGO the MEDIC saturates hospitals.

You make the call. Do you want to: