How these guides are produced
We create controlled fixtures, run them through the same conversion or OCR implementation used by NetroDoc, inspect the generated files, and publish only the behaviour that the test actually demonstrated.
Results are deliberately scoped to the tested fixture. A measured percentage is not presented as a universal promise for every document, scan, image, browser, or source application.
- Test date and runtime details are recorded before publishing measured claims.
- Unsupported or untested cases are labelled instead of inferred.
- Failures in the evidence run stop publication until the fixture or implementation is understood.