In the session I will walk over and show examples of how, with an open-source tool JCov, one could build a software development process where code coverage would not be just another report generated but would really be used for estimating quality of the software and also help with setting up the testing process and planning further test development.
Few subjects would be touched during the session:
- 100% - is it meaningful, achievable?
- selecting representative subsets from code coverage
- ranking code coverage info with other characteristics
- monitoring of state of the test base and the source code
- controlling quality of the incoming source code
- testbase reduction