The Dawn of Liquid Software
summary
Argues that software must become continuously updatable rather than versioned, because threats like Spectre require patches faster than traditional releases and Meltdown patches cost 5–30% performance; cites a Google Wi‑Fi push that factory‑reset thousands of home routers as a cautionary failure. Describes tooling and patterns used at scale — Spinnaker Kubernetes Netflix OSS — and proposes a coordinated release shuffle to achieve zero‑downtime updates while calling out schema and data migrations as the hardest problem and overcoming physics (binary transfer, scale, trust, visibility).
AI-written summary · about
chapters (11)
- 0:00Start
- 1:01Evolution of Software
- 1:23Nokia Software Updater
- 9:18Fred Salmon
- 14:24How the Pipes of Liquid Software Are Different from Pipes of Continuous Integration Continuous Delivery
- 21:04Java Release Cadence
- 26:51What's the Next Frontier
- 27:54Overcoming Physics
- 28:20Question of Trust
- 28:35Question of Visibility
- 31:48Breakout Sessions
related talks
50mDevOps Patterns & Antipatterns for Continuous Software Updates
30mDevOps Patterns and Antipatterns for Continuous Software Updates
27mDevOps Patterns and Antipatterns for Continuous Software Updates
17mDevOps Patterns & Antipatterns for Continuous Updates
38mAutomate or Stagnate: Surviving the Era of Continuous Updates
48mDevOps Patterns and Antipatterns for Continuous Software Updates
47mContinuous Acceleration: Why Continuous Everything Requires A Suppl...
45mSoftware Update for IoT: The Current State of Play