This section connects the vocabulary of the testing literature to the project's framework. Most developers arrive with that vocabulary and its habits: the unit/integration/e2e pyramid, test-driven development, code coverage, the relationship between tests and types. The connection is what lets the Methods catalog and the Quality dimensions guide read naturally from that starting point.

• Terminology maps the conventional terms — unit, integration, e2e, mock vs stub vs fake, and the rest — onto the axes and onto Google's scope/size taxonomy.
• Testing folklore and Design folklore set out what the empirical record does and doesn't support for popular practices: the 80% coverage target, TDD as a design discipline, "100% coverage subsumes types", small-files rules.
• Test smells does the same for the test-code smell catalog: which of the 196 cataloged types have been measured, how far the detectors can be trusted, and what the vocabulary is still good for.

Further reading

• Book-length treatment of the same subject, outside testing — The Leprechauns of Software Engineering takes claims the field repeats as settled (the 10x spread in developer productivity, the cost-of-change curve, the software crisis) and follows each citation chain back to what the primary sources measured (Bossavit 2015)¹.

Referenced by

• Software Quality · Overview

References

---

1. Bossavit, Laurent. 2015. The Leprechauns of Software Engineering: How folklore turns into fact and what to do about it. Leanpub. https://leanpub.com/leprechauns. ↩