Sources we rely on
We prefer the original document over anyone's description of it. In practice that means:
- Primary documents and data: government agencies and regulators, standards bodies, laws and court texts, company filings and official announcements, peer-reviewed journals, universities and national laboratories, and intergovernmental bodies.
- Established reporting: news organizations and trade publications with editorial standards and a record of correcting their mistakes.
Sources we do not cite
- Wikipedia, which is a useful starting point but not a source of record.
- Content farms and blogs written mainly to rank in search engines.
- AI-generated summaries of other people's work.
- Press releases that quote market forecasts from firms with no track record.
- A vendor's marketing, when it is presented as neutral fact.
- Forums and social media posts.
- Any page that was not actually opened and read during research.
How claims are handled
- Every number, date, named event and attribution must be supported by a listed source. If it cannot be supported, it is left out.
- What an organization says about its own products, results or plans is reported as its claim, with its name attached.
- Anything likely to change carries a date, such as "as of October 2026".
- Disagreement and uncertainty are stated, not smoothed over.
- Announced plans and targets are described as plans and targets, not as facts.
Credit and originality
Information on this site belongs to the people who produced it, and we credit them. Each page ends with its sources: the title of every document used, the organization that owns it, and a link to the original. Reference pages such as trackers go further and attach a source to each individual entry.
We write in our own words. We do not copy passages from sources, and we keep direct quotations rare and short. The purpose is to point you to the original work, not to stand in for it.
The full list of what we have cited, grouped by owner, is on Sources we cite.
Last updated