Three kinds of article
- Field notes: observations from a named type of environment, with date, method and scope. A short sample is not a sustained load test.
- Local experiments: isolated tests with recorded software versions, steps, result files and explicit limits. Synthetic data is identified.
- Practical guides: methods and checklists. Examples are labelled, and a guide does not claim measured performance.
Evidence before claims
We keep original results separate from interpretation. A website returning HTTP 200 does not establish real traffic, a good experience on every device or a successful business. A local test does not establish production readiness.
Reproducibility
Where useful, articles link to reduced public results and scripts. Access credentials, private configuration and other people's data are excluded. Version changes may affect results; dates and relevant versions remain visible.
AI disclosure
AI may assist with research organization, language and illustrations. Commands, figures and product limitations require checking. Images are conceptual unless explicitly identified as screenshots. We do not generate imaginary benchmark data or user quotes.
Corrections and updates
When we identify a material error, the affected article should describe the correction and update its modified date. New measurements are kept distinct from the earlier observation. The RSS feed carries published notes; it does not imply that a page is monitored continuously.
Commercial relationships
Launch articles contain no paid recommendations. Any future sponsored content or qualifying referral links must be labelled on the relevant page. A commercial relationship is not evidence of product performance.