Observed, reported, and estimated are separated
Direct readings such as viewport size are observed; processor count supplied by navigator is reported; inferences such as rounded memory or device class are labeled estimated.
See how results and guidance are produced.
This page records how scoring rules, test explanations, and troubleshooting guidance are reviewed before publication.
Last reviewed: September 1, 2026Editorial standard
Four rules applied to every explanation
Missing detail is not a fault
A browser hiding battery, GPU name, connection type, or memory does not create an automatic deduction. Privacy reduction and hardware failure are different states.
Recommendations start with low-risk steps
Permission review, repetition, and controlled comparison come before profile resets, operating-system installation, or hardware replacement.
Measurement limits stay visible
DeviceTriage does not claim to measure temperatures, malware, physical disk health, exact free memory, or definitive CPU/GPU performance.
Pre-publication review
How a diagnostic rule is reviewed
The goal is not to imitate certainty; it is to interpret a browser-exposed signal inside honest limits.
- 1
A new rule must map to a signal the browser can actually expose.
- 2
Mobile, tablet, and desktop resource signals do not share one threshold.
- 3
Permission denial, unavailable APIs, and real test errors are tested as separate states.
- 4
Critical scoring and route behavior must pass automated checks before publication.
- 5
Content and methodology are reviewed when browser behavior changes.
Change log
Material product and content updates
Input, storage, and WebRTC diagnostics
Added mouse, touchscreen, browser storage, and privacy-focused local WebRTC tests, plus three deep troubleshooting guides and topic-cluster navigation.
Diagnostic knowledge center
Added a three-step problem-narrowing flow, eight detailed test guides, and eight expanded troubleshooting guides.
English and Turkish coverage
Localized tools, device check, methodology, privacy, and fourteen troubleshooting guides in both languages.
Device-class-aware scoring
Separated mobile and desktop resource interpretation and made the non-benchmark boundary visible in results.
Initial diagnostic release
Introduced the browser-based device check, explainable scoring, and focused hardware tests.
Review the score rules
Category weights, device-class interpretation, and browser limits are documented in the methodology.