On Naming Things

There are only two hard things in computer science: cache invalidation and naming things. Phil Karlton said this decades ago, and it remains stubbornly true.

1. Names as Commitments

A name is a compressed explanation. When you call a variable data, you have explained nothing. When you call it unvalidatedUserInput, you have told the next reader exactly what they are holding and what still needs to happen to it.

2. The Refactoring Signal

When a name no longer fits, that is not a cosmetic problem. It is a design signal. The concept has drifted from its label, and the gap between name and reality is where bugs hide.