Why is it difficult to maintain a purely top-down integration testing strategy in practice?
If testing is so critical to software function and reliability, why is eliminating it a desirable objective in improving software trustworthiness?