Forum Discussion
Regression tests, by definition, make sure that what worked in a previous version are not "broken" in the next release candidate. To me, this means that you're doing functional testing based upon feature set and requirements. Technically speaking, this should also cover any previous bugs that were found and fixed because, if they were fixed properly, then the test cases that found them originally would pass.
When I did automated regression testing at a prior company, we did not create specific test cases to re-test the bugs. We adjusted existing test cases to make sure that they included the features that needed to be tested. The whole reason why a bug happened in the first place was that the previous test cases were insufficient in their code coverage. So, we simply adjusted those test cases to make sure that they covered the necessary code. This way we kept the number of test cases to a manageable amount but increased our regression code coverage.
Related Content
- 12 years ago
- 6 years ago
Recent Discussions
- 3 days ago
- 3 days ago
- 4 days ago
- 4 days ago