I Don’t Need No Tests
summary
Claims building a no QA from day one engineering model that ships faster by forcing team ownership and automation and by pairing shift left tests with shift right chaos, canary and canary rollout practices. Examples include CI/CD pipelines, feature flags, infrastructure as code, SDK pipelines that run across common customer configs, and postmortem debriefs; a multi-zone distributed system that detects and mitigates automated tax in real time runs under this model and the approach has been iterated for six years.
AI-written summary · about
chapters (24)
- 0:00Introduction
- 2:00Engineering without QA
- 3:50Blame for bugs
- 5:59Shift left vs shift right
- 8:10Perks and risks of shift left
- 9:35Testing web app behaviors
- 10:18SDK testing rigor
- 12:05Engineers own tests
- 14:12Manual QA need
- 15:40What to test and BDD
- 18:09Engineer ownership
- 19:15Fixing features and post-mortems
- 20:40Failing successfully
- 23:31Testing infra as code
- 26:26Quality metrics and alerts
- 27:30Notifications vs alerts
- 30:22Preventing burnout
- 32:15Burnout metrics and noise
- 34:23Incident debriefing
- 36:29Fresh start vs transition
- 37:11Shift left in practice
- 38:38Shift right and production
- 40:27Engineer culture fit
- 41:55Closing thoughts
related talks
51mWhy Companies Without Testers are the Best Place to Be One
46mQA to QE: Scaling Quality with Ownership, Code, and Culture
44mAchieving Excellence in a Startup without a Dedicated QA
47mHow Testing is Evolving in DevOps
42mTesting in a Continuous Delivery World
53mIs having a QA a symptom of a dysfunctional team?
44mQA in a Pipeline Based Organization
01h 09mScaling up with Embedded Testing