Continuous Delivery of HTML5 Applications
summary
Rangle.io moves front-end projects to continuous delivery so every merge is deployable, aiming for one-week sprints and most stories finished in days. Continuous integration runs on CircleCI, builds use gulp, pull-request environments auto-deploy to Heroku (migrating to Docker/Kubernetes for parity with production). Developers write unit and webdriver end-to-end tests, testers coach and debug the process, cordova apps use a live-update wrapper called cordwood, and teams track rejected stories to keep iterations closing with no lingering bugs.
AI-written summary · about
chapters (18)
- 0:00Intro
- 0:37Continuous What?
- 1:18Continuous Integration
- 2:58Continuous Delivery
- 4:04Continuous Deployment
- 6:52High Level Objectives
- 10:55Deliver Early for Value and/or Learning
- 12:27User Stories
- 16:48Downsizing Stories
- 19:06Avoid Horizontal Slicing
- 20:28Clear Acceptance Criteria
- 21:54Buggy Software = Work in Progress
- 24:22Deferred Bug-Fixing Costs More
- 27:45Build Environments
- 29:36Cordova Apps
- 31:34What Do the Testers Do All Day Then
- 33:49Automated E2E Tests
- 35:53PR Environments vs Production
related talks
55mContinuous Deployment Workflows for Modern JavaScript
46mContinuous Delivery for Native Apps
42mCONTINUOUS DELIVERY FOR RUBYISTS
01h 03mYet Another Continuous Delivery Talk
49mThe Whole Team Approach to Testing in Continuous Delivery
11mWorking with sharp tools - A case study of Continuous deployment
26mContinuous Everything in Mobile
41mAll Things Continuous