Replaceability is the ability to rewrite, upgrade, or substitute a component without destabilizing the surrounding system. It matters when requirements, technology, teams, or constraints will change, because architecture must preserve safe options rather than optimize only for today. section 4.10 section 4.2
Assess replaceability by asking:
- Which stable contract must the replacement preserve? section 4.2
- Which consumers will be affected, and how can they migrate safely? section 4.2
- Who owns the component and has the knowledge and trust to change it deliberately? section 4.8
- What feedback will reveal whether the change improved, degraded, or unexpectedly altered system behavior? section 4.10 section 4.2
Replaceability depends on modularity, limited assumptions across boundaries, clear ownership, stable contracts, backward compatibility, and incremental replacement. It is not achieved by splitting everything apart or inventing extension points for every imagined future. Good boundaries reduce unnecessary coupling while keeping related behavior together; operational feedback then shows when a boundary is becoming painful or a workaround is becoming permanent. section 4.10 section 4.8
A six-month replacement test
Before committing to a component, ask: if it had to be replaced in six months, would the team know the contract to preserve, the affected consumers, and the migration path? If not, the design is carrying hidden assumptions or unclear ownership. section 4.2
