Why Testing Matters

What testing gives you

• Quality: bugs are found while they are cheap to fix, on your machine or in a pull request, instead of in production.
• Confidence: a green test run tells you a change did not break existing behaviour, which makes it safe to ship small changes often.
• Maintainability: tests describe what the code is supposed to do, and they let you refactor without fear. Code without tests is often called legacy code for exactly this reason.

Kinds of tests

You will meet these names often:

Kind	Checks
Unit test	One small piece in isolation
Integration test	Several components working together
End-to-end test	A complete user flow through the system
Regression test	That a fixed bug stays fixed

A common rule of thumb is to have many fast, small tests and few slow, broad ones. This is often drawn as a “test pyramid”.

Common misunderstandings

• Tests do not prove the absence of bugs. They show that the cases you thought of work.
• Coverage numbers measure which lines ran, not whether the checks are meaningful.
• Not everything should be automated. See Test Automation Principles.

Where it fits

Automated tests usually run in a CI/CD pipeline on every push, and reviewers expect them in a pull request. The topics that follow cover unit tests, integration tests, manual testing and reporting defects.

Try it

Pick a small piece of code you wrote or know well. List five ways it could fail: normal input, empty input, a boundary value, invalid input and a failing dependency. Check each by hand, and note which checks you would be happy to see run automatically on every change.