trunkanddrop

Trunkanddrop is a term used in software development to describe a workflow pattern that blends trunk-based development with deliberate removal of code or features. The phrase does not have a single, universally accepted definition, and its meaning can vary by organization.

Core idea: teams commit frequently to the main code line, the trunk, while periodically dropping or pruning obsolete or canceled work. The drop is typically accomplished through refactoring, cleanups, or dedicated removal commits, often aided by feature flags to keep in-flight work from impacting users.

Practices commonly associated with trunkanddrop include small, frequent commits to the trunk, automated tests and continuous integration, and regular code cleanups to remove unused modules or dead code. Projects may use feature toggles to keep incomplete functionality in the codebase while controlling exposure, and then drop the feature entirely once it is canceled or superseded.

Benefits of this approach include reduced integration risk, clearer long-term maintainability, and a leaner codebase. Potential drawbacks are the need for disciplined cleanup, the risk of over-pruning valuable work, and the overhead of maintaining feature flags and cleanup processes.

Variants of the concept exist across teams; some treat drop as removing entire modules, while others see it as disciplined removal of canceled experiments or deprecated code paths. Adoption typically requires clear governance around when and how drops occur, as well as automated testing to prevent regressions.

See also:

- trunk-based development

- feature toggle

- continuous integration

- refactoring

- technical debt