Trunk-based development is a branching model where developers commit small, frequent changes directly to a single main branch that deploys automatically.
Successful trunk-based development relies on small frequent commits, robust continuous integration, and feature toggles to hide incomplete work from users.
Unlike feature branching which leads to merge conflicts and delayed feedback, trunk-based development fosters a collaborative culture through pair programming and real-time feedback.
Breaking down yearly revenue to per-minute losses highlights the significant financial risk of pushing breaking changes to a high-traffic production environment.
The testing pyramid's base layer consists of fast, easy-to-write unit tests and static linting to catch issues early for every line of code.
The middle of the testing pyramid uses integration tests, contract tests for APIs, and load tests to prevent performance regressions and breaking changes between services.
The top of the testing pyramid includes visual regression tests to catch unintended style changes and end-to-end tests to validate critical user flows.
A comprehensive testing strategy includes automated Lighthouse checks for performance and accessibility compliance, alongside periodic security and penetration testing.
Teams can transition to trunk-based development by auditing branch lifetimes, enforcing smaller pull requests, introducing feature flags, and fostering a strong review culture.
Hubert Burda Media
München, Germany
20:33Nadeen Hussein
28:12Luise Freese
23:33Maria Korneeva
56:49Jens Happe
58:26Markus Oberlehner
47:22Christian Kühn
29:26Bastian Heilemann & Bruno Margula
25:36Eric Deandrea
Jobs that call for the skills explored in this talk.
Wolters Kluwer
Alphen aan den Rijn, Netherlands
DEKRA AG
Stuttgart, Germany
Bertrandt AG
Tübingen, Germany
