↓ Ir para o conteúdo principal

← todas as notas

📎 Webclip

Me: "I tested and it works, why do I have to write tests?"

The author says manual testing used to be the only way to check code, but it could not cover all possible cases. He presents automated testing as a way to cover use cases systematically, keep software stable, and ship with more confidence.

Reading notes
#

  • Manual testing in the browser was the only way he knew at first.
  • He remembers launch checklists in spreadsheets, yet bugs still appeared.
  • Automated testing with Continuous Integration is described as systematic and confidence-building.
  • He sees resistance to testing as fear of leaving the comfort of “I coded it and it works.”
  • Tests are presented as a way to cover cases that the developer or reviewer might overlook.
  • He says tests define the specification of the code and make clear what it can and cannot do.
  • He browses test folders in libraries to understand behavior when API or demo pages do not explain it.
  • When code changes, all tests should still pass because the code must remain within its specification.
  • New code should come with new tests because it adds new specifications to the software.
  • Shipping code without tests means shipping code without a specification.
  • Open source repositories are described as treating tests seriously because code stability matters to many people.
  • He concludes that the article may not apply to everyone, but it may help readers rethink why testing matters.