No dev. No staging. No problem.
summary
Claims trunk based development with every PR deploying to production and averages 639 deployments per month, uses strict typing and runtime API response validation with Zod and TanStack Query, keeps infrastructure minimal via Forge and Envoyer, and runs hundreds of feature flags (Pennant) plus a zero bug policy Tests and quick rollbacks handle failures A live demo merges a PR and deploys a PDF export behind a flag and it works on stage
AI-written summary · about
chapters (16)
- 0:00Start
- 0:26Shipping straight to production, no staging
- 1:03Who is James Mooring and what Astalty builds
- 2:53The real-world constraints: mission-critical, compliance, sensitive data
- 3:45git-flow and staging branch complexity vs trunk-based development
- 6:06“Real app” scale and deployment frequency stats
- 7:12Why staging servers are harmful
- 8:50Why deploy fast
- 10:00The four pillars
- 10:22Stable code tactics
- 15:06Zero-bug policy, tests as non-negotiable
- 16:11Simple infrastructure
- 17:55Empower engineers
- 19:24Feature flags in practice
- 25:06Live demo: merge to production
- 28:24Do you really need a staging server?
related talks
01h 03mMoving Fast - without crashing
36mContinuous Deployment with Rails
01h 00mReliability and speed by 'push to main' mentality
18mTrunk Based Development and Feature Toggles
30mPushing to master - adopting trunk based development
35mIncrease Agility and Quality Using Trunk-Based Development in Continuou...
46mContinuous Deployment Workflows
48mHow to deliver continuously