Specification for typed identifiers that explicitly distinguish references from referents — "the map" (the identifier and any observations recorded about it) from "the territory" (the actual digital object, person, organization, or other entity being referred to). RIDs identify *digital objects as references which point to referents*, which is a deliberate philosophical commitment: multiple RIDs can reference the same referent with different observations, weightings, or contexts, without conflicting, because identification and reference are kept architecturally distinct. RIDs serve as the data-layer naming scheme for knowledge graphs across the KOI (Knowledge Organization Infrastructure) ecosystem, used by BlockScience's internal KOI, Metagov's KOI Pond, and RMIT/Telescope's deployment, with reference libraries `rid-lib` and `rid-registry`. Co-designed with KOI-net Protocol as part of the broader KOI suite.
LicençaOpen source — MIT License (rid-lib and rid-registry repositories at github.com/BlockScience)
Status de Dev 🔬 Beta
Detalhe do Status de DevActive research / Beta (current iteration is KOI v3; rid-lib and rid-registry libraries published and in use across multiple deployments; specification continues to evolve through working group review)
ProprietárioBlockScience (primary steward — complex systems engineering, R&D, and analytics firm); collaborative research with Metagov (Metagovernance Project) and RMIT University (via the Telescope project); rid-lib and rid-registry hosted under github.com/BlockScience
PaísUSA (BlockScience based in Albany, New York); collaborators include Metagov (international) and RMIT (Australia)
Ano de Início 2023
StackSpecification + Python reference libraries (rid-lib for core RID handling; rid-registry for registered RID type definitions); language-agnostic specification consumable by any implementation
FinanciamentoBlockScience operational funding (private firm); Metagov nonprofit funding for KOI Pond integration; Australian Research Council Future Fellowship funding for Ellie Rennie's RMIT-side work ("Cooperation Through Code" FT190100372, 2020–2025); collaborative research without dedicated RID-specific funding tranche
OriginsKnowledge organization / sensemaking infrastructure (originated within BlockScience's internal KMS as the naming primitive enabling distinction between observations about the world and the world itself; emerged from work on participatory digital ethnography and computer-aided ethnography, and from the practical need for an identifier scheme that could carry weighted, contextual references rather than imposing a fixed data structure)
DatabaseN/A — RIDs are an identifier specification; persistence is implementation concern (knowledge graphs hosted by KOI nodes typically use graph databases or hybrid stores, but the RID specification does not mandate any particular storage backend)
Query LanguageN/A — RIDs are identifiers, not a query mechanism; query patterns over knowledge graphs containing RIDs are implementation-layer concerns of the consuming KOI node
Data FormatsRID structure includes a type prefix (the kind of referent) plus the reference itself; precise encoding specified in rid-lib; designed for embedding into JSON, CBOR, and other structured data formats consumed by knowledge graphs and LLM interfaces
Mobile SupportLibrary-level (Python rid-lib runs on any platform supporting Python; mobile bindings not explicitly documented but architecturally feasible)
Web SupportLibrary-level (rid-lib usable from any Python web service; no dedicated web client; KOI nodes typically expose RID-aware HTTP APIs for web consumption)
Native AppsNo first-party native apps (RIDs are consumed by KOI nodes and downstream applications such as KOI Pond, Metagov's deployment, and Project Interlay; not a standalone end-user product)
TermsOpen Source / Free — MIT License
FundsUnknown — combined funding across BlockScience, Metagov, and RMIT operational budgets; specific RID development allocation not separately published
Based OnConcept of references-pointing-to-referents (drawn from semiotic distinction between map and territory); knowledge-graph traditions; conceptually adjacent to but explicitly distinct from W3C DIDs (DIDs identify controllers; RIDs reference any kind of digital object including but not limited to identity); designed to compose with KOI-net Protocol for inter-node knowledge exchange
PermissionsPermissions are handled at the consuming KOI node level rather than at the RID specification level. A node can decide which RIDs and which observations about RID-referenced referents to accept, retain, weight, or expose. The RID specification itself imposes no permission model
Authentication & IdentityN/A — RIDs are an identifier specification that names digital objects; authentication and verification of any identity claim associated with an RID-referenced entity is a concern of the consuming KOI node or downstream identity layer, not of the RID specification itself. RIDs are explicitly distinct from DIDs: DIDs identify controllers with key-rooted self-certifying semantics; RIDs reference any digital object without prescribing authentication semantics
Storage ModelRIDs themselves are values, not stored entities; in deployment, RIDs are used as the naming primitive within knowledge graphs hosted by KOI nodes. Storage is implementation-defined — typical KOI nodes use graph databases or hybrid stores; the RID specification does not mandate any backend. Reference-vs-referent distinction means multiple knowledge graphs can hold different observations about the same referent without identifier conflicts
InteroperabilityHigh at the identifier layer — RIDs are embeddable in JSON, CBOR, RDF, JSON-LD, and other structured data formats; designed to compose with KOI-net Protocol for inter-node knowledge exchange; bridges to W3C DIDs feasible (RIDs can reference DID-identified entities) and to RDF/JSON-LD via URI embedding; consumable by any implementation language given the language-agnostic specification
Data PortabilityHigh — RIDs are inherently portable as typed identifier values; the same RID resolves to the same referent across KOI nodes; observations and weighted references about a referent can be exchanged between knowledge graphs via KOI-net Protocol without identifier translation; consuming nodes retain ownership of their own observation sets
Governance & Decision MakingBlockScience-led working group with collaboration from Metagov and RMIT/Telescope; "Beneficent Dictator" pattern with primary developer guiding contributions; weekly Govbase Labs calls for cross-organizational coordination; ongoing specification refinement through KOI working group review; no formal standards-body governance at present
Protocol Maturity / StandardizationActive research / Working specification (codified into rid-lib and rid-registry; evolving through BlockScience-led working group with Metagov and RMIT collaboration; not yet submitted to formal standards bodies; governance follows "Beneficent Dictator" pattern with primary developer guiding contributions)
Identity StandardsRIDs are themselves the standard at the identifier-naming layer of the KOI ecosystem; explicitly distinct from but composable with W3C DIDs (which identify controllers via key rooting); not aligned to traditional identity standards (X.509, OAuth, SAML) since RIDs do not perform authentication or credential semantics
DID Methods SupportedN/A — RIDs are not a DID method and do not implement the W3C DID specification. RIDs and DIDs serve complementary purposes: a DID can be referenced by an RID, and an RID can name a DID-identified entity, but the systems are architecturally distinct
Key ManagementN/A — RIDs are identifiers, not key holders; the RID specification does not include cryptographic key management. Identity claims about RID-referenced entities (if any) are handled by downstream identity layers or by the consuming KOI node
Credential TypesN/A — RIDs are not a credential system. Credentials about RID-referenced entities, if issued, would be handled by downstream identity layers (e.g. W3C VCs anchored to a DID that an RID references)
Verification MethodN/A at the identifier level — RID values themselves are not cryptographically verified; they are typed identifier strings whose meaning is established by the registered RID type and consuming KOI node. Observations and references made about a referent through an RID may be cryptographically signed by the asserting node, but this is implementation-defined and not part of the core RID specification
Privacy FeaturesThe reference-vs-referent distinction itself supports privacy by enabling contextual and weighted observations: a knowledge graph can record references and observations about a referent without those records being intrinsically linked to the referent's other appearances elsewhere. Different KOI nodes can hold different observation sets about the same referent. Stronger privacy guarantees (selective disclosure, zero-knowledge proofs) would be implemented at downstream layers, not in the RID specification itself
Authentication MethodsN/A — RIDs do not perform authentication. Authentication of parties asserting observations or references in a KOI knowledge graph is handled at the KOI-net Protocol layer or by the consuming node's identity infrastructure
Revocation MechanismN/A at the identifier level — RID values are stable typed identifiers; they are not revoked. Observations or references made about a referent via an RID may be retracted, superseded, or down-weighted by the consuming KOI node, but this is a knowledge-graph-layer concern rather than an RID specification concern
Agent Types SupportedAny digital object — broader than the typical identity system. RIDs can reference humans, organizations, software agents, documents, knowledge artifacts, datasets, observations, and other digital objects. Type extensibility through rid-registry allows new referent categories to be added without changing the core specification
Wallet/Client TypesN/A — RIDs are not consumed through a wallet pattern. KOI nodes are the primary consumers; downstream applications (BlockScience's KOI, Metagov's KOI Pond, RMIT/Telescope, Project Interlay) interact with RID-aware HTTP APIs exposed by KOI nodes
Recovery MechanismsN/A — RIDs are stable typed identifiers, not credentials or keys that can be lost. The concept of "recovery" does not apply at the RID specification level
Compliance / RegulationsNot specifically aligned to identity-focused regulations (GDPR identifier provisions, eIDAS, etc.) since RIDs are a more general identifier mechanism than the typical SSI identity system; compliance posture is determined by what the consuming KOI node does with RID-referenced data and observations, not by the RID specification itself
Credential Exchange ProtocolsN/A — RIDs are not a credential system. Knowledge exchange between KOI nodes is handled by the KOI-net Protocol, which operates at the community-data-and-governance layer above the identifier-naming layer where RIDs sit
Trust FrameworkNo formal trust framework at the RID specification layer. Trust posture in a KOI deployment is established by the participating nodes and the KOI-net Protocol's inter-node coordination patterns, not by the RID specification. Composable with broader trust frameworks (e.g. ToIP, GLEIF vLEI) at downstream layers if an RID references entities anchored in those frameworks
Cost ModelFree — MIT-licensed reference libraries; specification is open; operational costs are deployment-specific (consuming KOI nodes' infrastructure, storage, compute). No protocol-level fees
Censorship ResistanceIdentifier-level only — RID values themselves are not censorable since they are stable typed identifier strings that any node can mint and use. Network-level censorship resistance depends on KOI deployment topology and the underlying KOI-net Protocol's federation pattern. Individual KOI nodes can choose to ignore or down-weight references and observations from other nodes
Viu algo faltando ou que poderia ser melhorado? Nos diga →