Feature Toggles: The Good, The Bad, and The Ugly
summary
Explains concrete rules for feature toggles with C# examples and real failures like Night Capital where reusing a flag for an eight year old power peg feature plus a new smart implementation causes divergent behavior across eight servers and wipes out about $450M in 45 minutes, and engineering wins like HP moving from compile-time ifdefs to startup config, cutting merge feedback from eight weeks to two days across 400–800 devs and 10M LOC. Covers naming, short lifespans, monitoring, rollout patterns and when to delete toggles.
AI-written summary · about
chapters (12)
- 0:00The knight capital disaster
- 1:50Technical root causes
- 5:20Lessons in toggle management
- 7:31Best practices for toggles
- 9:11Types of feature toggles
- 11:10Case study: HP printer firmware
- 15:21Implementation techniques
- 20:22Toggles for operations
- 25:36Safe rollout strategies
- 31:29Monitoring and observability
- 34:23Managing complexity
- 38:47Testing and recap
related talks
30mFeature toggles: ask us anything
27mFeature Toggles
35mFeature Toggles, Trunk-Based Development and Continuous Delivery
14mWhat they don’t tell you about feature toggles
56mFeature Toggles: Dynamic Configuration at Wirex
41mFeature Flags: A Boon or a Brake for a Team
32mMaster your software delivery with feature flags
51mCode Born to Die