It solves a very specific kind of documentation noise: timestamp-only churn.

The Problem

If your project publishes generated YARD documentation from a checked-in docs/ directory, you have probably seen this:

You run the docs task.

YARD regenerates HTML.