Editorial policy

Codepoints, Unicode names, CLDR short names, groups, subgroups and version numbers are facts. They carry the label Source: Unicode Consortium and are never rewritten or paraphrased.

Meanings, tone, common uses, misreadings, tags and collections are ours. They carry the label DBEmoji editorial interpretation. If you are citing something as authoritative, cite the first group and treat the second as our reading.

Where we have no confident basis for a section, we leave it out. An emoji page with a meaning, the technical data and nothing else is a better page than one padded with plausible-sounding sentences.

In practice: tone is shown only for emoji that carry a clear register — an object emoji gets none. “Common uses” is derived from Unicode’s own subgroup and omitted when that gives us nothing. “Easy to misread” appears only for emoji where a misreading is genuinely common, not on every page.

A page never describes an emoji as two opposite things. Where an emoji has a recorded mood, that mood overrides any tag that contradicts it, and use cases are taken from what the emoji depicts rather than from its keyword list.

This rule exists because of a real defect: 😂 was tagged both happy and sad by an upstream keyword match, and the page generator turned that into a suggestion to use a laughing emoji in condolence messages. Contradictory mood tags are now removed at build time and the validator refuses to ship them.

• No invented authors. Pages are written by the DBEmoji editors, collectively.
• No invented dates. A “verified” date comes from the build that produced the dataset, not from a template.
• No statistics without a source, and no usage rankings we cannot measure.
• No questions added to a page purely to chase a rich result. The Q&A on an emoji page exists because those are the questions people actually arrive with.
• No structured data describing something the reader cannot see on the page.
• No claim that the database is “complete” or covers “the latest Unicode”. We state the version: Unicode Emoji 16.0.

Interface text, category names and emoji names are translated for every locale we publish. Longer editorial prose is not machine-translated: where a locale has no reviewed translation, the page shows the factual summary and the structured data rather than English paragraphs dropped into a non-English page.

Blog articles are published in English only. Their non-English URLs are canonicalised to the English version and carry noindex, so translated-looking duplicates never enter the index.

Emoji meanings shift, and we get things wrong. Corrections are welcome through the report-issues link in the sidebar. Factual errors in Unicode data are treated as blocking; disputed interpretations are revised toward the more conservative reading.