Deprecation

Code is a liability, not an asset. Every line requires maintenance — bug fixes, dependency updates, security patches, cognitive overhead. When equivalent functionality requires less code or better abstraction, the old version should be retired.

Principles

Code as liability

Value comes from functionality, not code volume. Less code serving the same purpose is strictly better.

Hyrum's Law makes removal hard

All observable behaviors become dependencies. Users rely on bugs and undocumented side effects. Deprecation requires active migration, not just announcement.

No transitional architecture

Don't maintain two systems in parallel. Land the replacement, migrate consumers, remove the old system. Dual systems double maintenance cost.

The Churn Rule

If you own the infrastructure being deprecated, you are responsible for migrating your users — or providing backward-compatible updates that require no migration.

Workflow

1. Build the replacement first. Never deprecate without a working alternative.
2. Identify all consumers. Grep for usages, check imports, trace call paths.
3. Migrate incrementally. Move consumers one at a time, verify each migration.
4. Verify zero usage. Confirm no remaining references before removing.
5. Remove completely. Delete code, tests, documentation, configuration. No commented-out remnants.

Zombie code

Code with no owner but active dependents. Signs:

• No commits in months with active consumers
• Failing tests left unfixed
• Outdated dependencies

Either assign an owner and maintain it, or deprecate it with a migration plan. Zombie code cannot remain suspended.

Red flags

• Deprecating without a replacement available
• Multi-month "soft" deprecations with no progress
• New features added to deprecated systems
• Code removal without verifying zero active consumers
• "We'll maintain both indefinitely"