How cases are produced

Editorial and testing policy

Cases are selected from distinct user problems and published only with reproducible original evidence.

Selection

A case must solve a specific responsive-layout failure and have a mechanism distinct from existing cases. Framework keyword variants and lightly rewritten duplicates are not separate content.

Original evidence

Fixtures, test strings, screenshots, measurements, and regression tests are created by the project. Failing and fixed states share the same content and differ only where needed to isolate the cause.

Sources and review

Technical claims prioritize specifications and first-party browser or testing documentation. A browser run records its environment and remains partial until every browser required by the case has been executed and reviewed.

AI assistance

AI may assist with outlines, code review, wording, and omission checks. It does not replace browser execution, evidence capture, source verification, operator identity, or final editorial responsibility.

Updates

Pages use real modification and review dates. A date is not changed merely to make content appear recent. Material corrections are recorded in the changelog.