Testing, Testing, 1, 2, NaN
summary
Argues that testing myths are false and shows concrete tooling and tradeoffs: unit tests with Jest and fetch-mock, snapshot tests via Storybook, property testing with checkjs, integration caution, end-to-end with Cypress (recordings work) and mobile with Detox (works but has gotchas) while Appium often fails. Gives surprising data point — 67% Windows class fails on renderer tests — and offers built tools: solidarity for environment checks, Danger for PR warnings, greenkeeper/updater for deps, and recommends type checking and chaos tools for resilient apps. Jest snapshots Cypress recordings
AI-written summary · about
chapters (22)
- 0:00Intro
- 0:53Testing... Testing... 1, 2, NaN
- 3:40TALK ON TESTING
- 4:32Teams
- 5:00The Tester's Crisis
- 5:32Testing Myths aka Excuses
- 7:08Think Functional
- 7:53If you're doing the same thing in two places, DRY your code.
- 9:39The Testing Domain
- 15:54Snapshot Unit Tests
- 20:52Cypress for Web
- 21:09Appium
- 23:51Complexity Audits
- 27:49Environment Tests? wat?
- 29:27API Snapshots?
- 30:40Perf Testing
- 30:52Copy-Paste Testing? JS Inspect DRY code is maintainable code
- 31:48Coverage Reporting
- 33:14Static Type Checking PLEASE!
- 34:02Localization pseudo-localization
- 37:02Production Testing Crowd-sourced Errors
- 37:52The question is: "What needs to be durable and effective for your software?"
