Software Delivery from First Principles
summary
Argues software delivery is closer to industrial engineering than coding, claims reproducible builds (DAGs, hashes, not mtimes) cut wasted work but hit diminishing returns, and treats artifacts + repositories as the reliable unit of shipment that lets promotion and signing work. Mentions make, just/task, cmake, cargo, Go toolchain, Bazel/Pants/Nix/NX, tar and even Excel as examples. Emphasizes reproducible builds, artifact repositories, and gates as AND checks plus timely feedback and observability (canaries, smoke tests).
AI-written summary · about
chapters (7)
- 0:00Introduction
- 1:03Metaphors in software
- 6:16Software delivery as engineering
- 7:19Builds as blueprints
- 15:13Pipelines as flow systems
- 21:27Feedback loops in pipelines
- 24:44Summary
related talks
31mSoftware Delivery and the Rube Goldberg Machine: What Is the Problem We Are Trying to Solve?
30mBringing Continuous Delivery to Open Source
26mThe Timeless way of DevOps
26mWhere Are We Going? The Future of Continuous Delivery
56mThe Nature of Software Development
30mModern Continuous Delivery
40mContinuous Delivery Anti-Patterns
51mBetter software, faster: principles of Continuous Delivery and DevOps
