How we source

Every troubleshooting entry on this site links to where it came from. This page says what that means.

What counts as a source

Public issue trackers — GitHub issues in manufacturer and community repositories; each entry cites the issue number. Manufacturer documentation — error-code tables and status-light pages published by the manufacturer; codes and messages are reproduced verbatim, and the “what to do” advice is ours and is marked as such. Published research and disclosures — arXiv, IEEE Spectrum, security advisories. Published owner reviews and distributor documentation.

What is never published

Anything without a URL we fetched ourselves. Specifications or fixes reconstructed from memory or written by an AI model without a source behind them. Forum, Reddit or Discord posts — we don’t read those yet; when we do, this page will say so.

How entries are built

A weekly script searches the trackers for each robot and files candidates. Candidates must clear an engagement floor and a repository-quality floor before an editor sees them. Entries above a quality threshold are published automatically and reviewed afterwards; everything else waits for an editor. Severity, status and category labels are assigned by us and are judgements, not measurements. Each page shows its last-updated date; manufacturer pages are re-checked weekly for changes.

Corrections

If an entry is wrong, use the contact form and name the entry. Corrections are reviewed by the editor and the entry is updated or withdrawn.

Who

Edited by Ofer Tirosh. Operated by Tomedes LLC, Beaverton, Oregon.