Thank you James, and you're very welcome. I agree that developers on large projects may sensibly conclude that 100% coverage may not be a worthy near-term goal. Concretely, one of the very large codebases to which I contribute is very high quality by most measures yet also has significant technical debt and low test coverage. The value of test coverage is recognized by key developers as a way to improve their code review process, and so coverage reporting is just now being introduced. There is no talk of goals to reach specific coverage percentages, rather the expectation is simply that coverage should improve over time. I am interested to see, over the coming couple of years, how the new code coverage reporting will impact not only the design and quality of codebase but also how it will influence us as developers, patch reviewers, and architects.
Authored Comments
Thank you James, and you're very welcome. I agree that developers on large projects may sensibly conclude that 100% coverage may not be a worthy near-term goal. Concretely, one of the very large codebases to which I contribute is very high quality by most measures yet also has significant technical debt and low test coverage. The value of test coverage is recognized by key developers as a way to improve their code review process, and so coverage reporting is just now being introduced. There is no talk of goals to reach specific coverage percentages, rather the expectation is simply that coverage should improve over time. I am interested to see, over the coming couple of years, how the new code coverage reporting will impact not only the design and quality of codebase but also how it will influence us as developers, patch reviewers, and architects.