Ranke-Graph

Read the research

Two papers so far. The first works out what a provenance-first data structure has to be; the second, what it takes to serve one. Both are drafts, and both are readable on their own.

The papers

Paper 01 · draft

Ranke-Graph: A Provenance-First Data Structure

Defines the graph as an abstract data type: what a claim is, how its identity is computed, and which guarantees follow from that alone. The minimum contract an implementation has to keep.

Read the PDF
Paper 02 · draft

RankeDB: Serving the Ranke-Graph

Takes that contract as given and specifies a server for it: how a graph is persisted across interchangeable backends, composed, queried, verified, and bounded for access.

Read the PDF

Three more are sketched and unwritten: workers that populate the graph, retrieval strategies over it, and coordination between several agents working on one archive. They exist as titles and notes, nothing more.

Companions

Glossary The series' terminology in one place.
Normative specification The rules an implementation follows. Written for implementers rather than readers.

Citing them

The papers carry no DOI. Each release is tagged and permanently downloadable, so cite the version you read: the release tag, the paper number, and its title. The changelog records what changed between versions.