Testing is not an exact science. We can test for a class of anomalies, and certify a product does not contain those anomalies under test conditions. We can test for another class of anomalies under operational conditions, but it is never clear when the list is complete. The test engineer has to balance the conflicting requirements between customer's expectations, his boss' expectations, and market forces. If he fails in maintaining this balance, he fails at his job. However, there exists instances where ethics would dictate that a balance is impossible. The author discusses these problematic issues facing the test engineer.