Move fast and make things
summary
Argues for resisting heavyweight tools, managing dependencies, and using short analytics spike in a weekend experiments to shrink big problems into doable pieces. Shows four concrete libraries — Faraday, Celluloid, Metriks, Pow — and demonstrates an actor-based analytics collector that runs Redis-backed counters across Celluloid actors with supervision and a pipeline that emits metrics to logs, CSV and dashboards. Measures latency with Metriks timers including 95th percentile to find outliers and reports that the prototype works and guides production design choices.
AI-written summary · about
chapters (44)
- 0:00Intro
- 1:25Problem with large applications
- 2:36Moving fast
- 3:18Reduce big challenges
- 4:16Do the right things
- 4:57Discover and solve risks
- 5:51Make things
- 6:09Facilitate change
- 7:06Big things resist change
- 7:28Wants and needs
- 8:28Big things
- 9:04Resist change
- 9:46Small things are great
- 10:28Small things are loosely held
- 11:07Small things have to adapt
- 12:09Resist having large software
- 13:46Large tools
- 14:27Manage dependencies
- 15:43How to use spikes
- 17:16Build the right software
- 18:52Spikes
- 19:39Uncertainty
- 21:14Leverage in tools
- 22:36Cost benefit of tools
- 23:56selfish tools
- 26:41versatility
- 26:52starting off easy
- 27:21discover the structure
- 28:05add structure as you go
- 28:51parallelize
- 29:58resisting large software
- 32:43moving fast make things
- 34:14actors
- 35:14analytics
- 36:47analytics collector
- 37:31celluloid
- 38:11Network services
- 38:22More reading
- 39:12Metrics
- 41:40Stats
- 42:36CSV
- 42:49Spreadsheets
- 43:18Dashboard Services
- 43:32Metrics everywhere
