Sources and Citations
Our standard is that every article carries a sources section. Articles that are still in preparation may not have one yet: a sources section is added when sourcing review is completed, before publication. This page explains what we treat as a source, in what order we prefer them, and what we will not cite.
Preferred sources
Standards and specifications
- RFC Editor — the authoritative source for the IETF RFC series.
- ISO/IEC — international standards, including the MPEG and JPEG families.
- ITU-T — Recommendations, including the P-series used for quality of experience assessment.
- 3GPP — mobile system specifications.
- ETSI, DVB, EBU — European broadcasting and media standards.
- IEEE — technical papers.
Law and regulation
- Court websites and published judgments.
- Legislation databases.
- Regulator publications, including EUIPO and Europol reports and national regulators such as Ofcom.
Product and project material
Datasheets, official documentation and release notes for hardware and software. We state the actual licence of any project we describe, as published by that project. Marketing claims are not treated as sources.
What we do not cite
- Search-engine snippets, summaries or aggregators, as an authority for a fact.
- Automated tool output, as a source for a figure, quotation or standard number.
- Another blog’s citation, where the original document can be found directly.
- A source we have not actually opened.
- Marketing material presented as an independent finding.
If no source exists
Where no published figure exists, we say so in the article instead of estimating, and instead of carrying a number over from an unsourced page. An acknowledged gap is more useful than a confident guess, and it is the only honest option.
Attribution in our articles
Where an article has completed sourcing review, its sources are listed at the end, with a link to the document where one exists and the date we checked it. Where that review is not yet complete, the article carries no sources section: an acknowledged gap is more honest than a partial list presented as if it were whole. Standards change, so a check date matters as much as a citation.