Test Coverage
Recently, while I was running a TDD workshop with some colleagues at the office, an interesting discussion arose regarding test coverage targets and metrics which reminded me of this article.
Back in 2012, Martin Fowler analyzed the traditional coverage metrics in this short and concise blog entry. Here, he stated that while coverage metrics could be useful to find untested code, it’s numeric value provides very little idea on the quality of the tests.
As an added bonus, you’ll also find links to both the excellent article by Brian Marick on the “Misuse of Code Coverage” and the funny—yet very accurate—commentary of Testivus.