Scaling Mobile Development
summary
Twitter shifts from a few mobile engineers to full stack permanent teams, trains hundreds through Twitter University, and buys Crashlytics for realtime crash data. Committer count climbs 10x while patch throughput rises 5x, exposing branch chaos that forces a move to trunk, CI with Jenkins, and a server-driven feature switching system so client code ships dark and ramps per device, OS, or locale. Nightly builds feed a public beta program and device lab to catch platform-specific crashes early.
AI-written summary · about
chapters (22)
- 0:00Intro
- 0:30What is Mobile at Twitter
- 4:30ServiceOriented Architecture
- 10:00Getting the Company Educated
- 12:00Culture Shift
- 13:25Network Effect
- 14:23Experimentation
- 15:36Lean Startup
- 16:00Logistical Issues
- 17:54Recruiting Mobile Engineers
- 23:58Code Reviews
- 27:31Logistics
- 29:22Feature Switching
- 30:59Slice and Dice
- 32:17Photo Filters
- 33:34RighttoLeft Rendering
- 35:23Fear
- 36:59Challenge
- 37:26Automation Testing
- 39:37What has this done
- 40:47Device Lab
- 41:29Questions
related talks
25mBetter Together
32mBUILDING MOBILE PRODUCTS THE HAPPY WAY!
34mBeing more than an Android developer
32mShipping at the Speed of Life
43mFrom Developer to Development Manager
37mWhat I wish I knew when I was a junior developer that I would tell as a staff software engineer
35mGetting called up to the Majors
49mMobile Developer Productivity Panel