Fun with Flags: How OpenFeature Solves Your Feature Flag Heada...
summary
Starts from a concrete failure — environment-variable flags cause a half a billion dollars lost incident — then shows how OpenFeature replaces vendor SDKs with a single evaluation API, pluggable providers and Flagd and Flagsmith providers so swapping backends is one line. Demonstrates deterministic rollouts with an evaluation context and targeting key, OpenTelemetry hooks and metrics for observing stale flags, and cites adoption signals like 2.4 million NPM downloads/week plus Dynatrace reporting ~2,000 flags with >1,000 inactive.
AI-written summary · about
chapters (14)
- 0:00Introduction and welcome
- 1:27Feature flag horror story
- 7:06What are feature flags?
- 10:15OpenFeature introduction
- 14:09Evaluation API explained
- 15:33Problems with homegrown solutions
- 17:18Providers and vendor lock-in
- 19:05Dynamic evaluation and context
- 25:24Percentage-based rollouts
- 27:11Obsolete feature flags
- 29:17Hooks for observability
- 32:22Key takeaways summary
- 34:26Why OpenFeature matters
- 35:51Call to action and community
related talks
26mBuild Progressive Delivery and Safety Into Your Apps with Feature Flags
23mFrom “What Broke?” To “What Changed?”: Unlocking Feature Flag Observability
29mCooking up o11y w/ Feature Flagging
33mSupercharging Observability with Feature Flagging
27mHands-Off Features Releases With Keptn, OpenFeature, And OpenTelemetry
21mFeature Store Observability: What it is, and why it matters
50mOpen Telemetry: The Rise of Open Standards in Observability
10mLightning Talk: Is OpenTelemetry Too Complicated to Get Started?