Answer:
Testing is done to show that a program is capable of performing all of its functions and also it is done to detect defects.
It is not always possible to deliver a 100 percent defect free program
A program should not be completely free of all defects before they are delivered if:
1. The remaining defects are not really something that can be taken as serious that may cause the system to be corrupt
2. The remaining defects are recoverable and there is an available recovery function that would bring about minimum disruption when in use.
3. The benefits to the customer are more than the issues that may come up as a result of the remaining defects in the system.