Methodology

An annotated edition is editorial work. The regulation text is free, in the public domain; the apparatus around it — what counts, where it sits, how it interacts — is what we build. This page describes how that work happens.

Data becomes information when it is selected; information becomes knowledge when it is bundled and pin-cited to the point where it bears.

Sources

We work from five primary sources. We name them; we do not partner with them.

EUR-Lex — Canonical regulation text and recitals.
CURIA — Court of Justice and General Court judgements. We adopt CURIA’s outcome classification rather than imposing our own taxonomy.
Rechtspraak — Dutch judgements at every instance.
Autoriteit Persoonsgegevens & sister DPAs — Enforcement decisions, fines and remedies; coverage extends to peer authorities across Member States.
EDPB & WP29 archive — Guidance, recommendations, opinions; the WP29 archive remains relevant for GDPR.

Bundling and pin-citation

Each entry of the apparatus is cited to the paragraph (and where useful, the point) of the article it bears on. A judgement that turns on Article 6(1)(f) appears at 6(1)(f), not at "Article 6." A guidance document that addresses two paragraphs of different articles is mirrored at both. Where we have done less work than we want to, we say so editorially: we improve cite resolution as we go.

Update cadence

Regulation text changes by amendment; we mirror the official consolidated version. Judgements and decisions are picked up on publication. Guidance is picked up on issuance. Where coverage is queued as backlog, the affected article carries an editorial note.

What we do not publish

The schema, the scraping pipeline and the build process are not editorial surfaces. We expose enough for a reader to assess the work; the machinery stays in the workshop.