manual-testinglisted

# Manual Testing ## When to reach for this - Before automation exists, or when automation isn't worth the cost (one-off, highly visual, exploratory). - New feature needs a test plan. - Investigating a reported bug and need a repeatable, structured approach. ## Test case design techniques Pick techniques based on the input space, don't just guess values: - **Equivalence partitioning** — split inputs into classes that should behave the same; test one representative per class. - **Boundary value analysis** — test min, min-1, min+1, max, max-1, max+1, and just-outside-range values. Most bugs live at boundaries. - **Decision table testing** — for features with multiple independent conditions (e.g. role × subscription tier × feature flag), build a table of condition combinations and expected outcomes. - **State transition testing** — for anything with a lifecycle (order status, account state), test valid transitions, invalid transitions, and transitions from every state. - **Pairwise/combinatorial testing** — when full combinatorial coverage is too large, use pairwise coverage to catch most interaction bugs cheaply. - **Error guessing** — based on experience: empty strings, nulls, huge inputs, unicode/emoji, SQL-special chars, double-submits, back-button, concurrent edits. ## Exploratory testing (session-based) 1. **Charter** — one sentence: what area, what risk you're hunting for. e.g. "Explore checkout discount code entry for validation and edge-case bugs, 30 min." 2. **Time-b