Chapter Three · The Map of Testing
Test Levels & Test Types
The profession's coordinate system. Four levels — unit, integration, system, acceptance — each catching what the others structurally can't. Then the axes that cut across them: what-it-does versus how-well (functional and non-functional), how much you can see inside (black, white, grey box), the vocabulary of testing around change (regression, smoke, sanity, re-testing), and the split between running things and reading them (dynamic and static). By the chapter's end, every kind of testing you'll ever hear named has a place on your map.
Job postings and standups throw testing words around like everyone was born knowing them: "we need regression coverage," "that's caught at unit level," "did anyone smoke the build?" This chapter is where those words stop being noise. It's the ISTQB backbone of the book — the classification every interview assumes — taught the way classifications should be: each term with the job it does, the thing it alone catches, and a Fernway example you'll actually remember.
Five topics build the map. The four test levels, with the date bug FW-312 traced through all of them. The functional/non-functional split, and the nine-second search that passed every written requirement. The three shades of box, ending with the 1:59-before-the-nightly-job test that marks a grey-box thinker. The change vocabulary — and the fix that broke its neighbor, giving birth to FW-329. And static versus dynamic: the discovery that your refinement questions were testing all along.