Continuous Integration For The Win!
summary
Shows how a GitHub pull request carries automated test, lint and coverage results so merges require green checks: tox + Docker builds a reproducible environment, a wheelhouse of pinned wheels avoids network installs, pytest runs ~8,300 tests that once took ~40 minutes, and flake8 enforces style. CI services used include CircleCI + codecov (previously Shippable/TravisCI), a docker-run-test script maps volumes to export JUnit/coverage XML, and failing checks block merges and surface regressions instantly. This produces enforced workflow and rapid feedback.
AI-written summary · about
chapters (9)
- 0:00Intro
- 1:25Code Ratio
- 5:20Our Process
- 9:42More than a Process It's about culture
- 12:14Eat the Elephant
- 15:10In Virtualenv
- 16:54With Code Coverage
- 22:02Docker The heart of the docker-run-tests scrint
- 28:02Questions?
related talks
06mContinuous Integration
34mContinuous Integration: A Bittersweet Love Story
01h 39mMastering git collaboration -A DevOps necessity
31mContinuous Integration 101 for Devs
43mTransformative Culture
55mContinuous Deployment Workflows for Modern JavaScript
30mAutomating Code Quality
46mComic book continuity and Git rebase