Skip to main content

Terminology

Last updated: 2026-03-01

This handbook uses the following terms in a precise, protocol-engineering sense. When terms are overloaded in other ecosystems, this page is the source of truth for the handbook.

Design goal: terminology should make disputes resolvable and implementations interoperable.

Core concepts​

Presence​

A private, real-world observation that a subject was at a location during a time interval. In Scintilla Locate, presence is proven as a statement, not published as raw coordinates.

Location privacy​

The protocol treats raw coordinates as sensitive. The system aims to make presence legible to smart contracts without making location histories public.

Claim​

A protocol artifact asserting a statement of presence (typically “within region R” and “within time window T”). A claim is structured so that it can be committed to, verified, and challenged.

Claim leaf (claim leaf bytes)​

A canonical byte encoding of a claim suitable for hashing and inclusion in Merkle commitments. A claim leaf is the unit of commitment for round datasets.

Claim ID​

A stable identifier derived from hashing the canonical claim-leaf bytes.

Regions and spatial indexing​

SemanticRegion​

A human-meaningful description of space (and optionally altitude) used for intent: polygons, circles, H3/S2 unions, geofences, etc.

ProofRegion​

A deterministic, discrete execution format derived from a SemanticRegion and used for verification. For Scintilla Locate v1 the primary ProofRegion execution format is LocateTile unions.

LocateTile​

A deterministic, integer-only spatial index identifying a rectangular geodetic cell by (level, ix, iy), with canonical encoding rules.

Membership​

The property that a point (or private tile ID) lies within a region. In ZK form, membership is proven without revealing the underlying coordinates or private index.

Rounds and publication​

Round​

A batching interval over which protocol inputs are collected and committed. Rounds exist to scale throughput while keeping attested on-chain artifacts small.

Epoch​

A larger time or configuration interval used to interpret round artifacts. Examples include committee snapshots and TEE certificate epochs.

Round Publisher​

A deterministic bulletin-board service that collects inputs, orders them by fixed rules, builds per-round datasets, computes commitments, and publishes a compact per-round header. It is liveness-critical, not trust-critical: misbehavior must be detectable by recomputation.

Round Header​

A small per-round JSON document that commits to large off-chain datasets by hash and Merkle root. The Round Header is designed to be suitable for Web2Json attestation and to be independently replay-verifiable.

Datasets and commitments​

Dataset​

A binary blob representing per-round inputs (claims, evaluations, committee snapshot, TEE certs, etc.) with a normative byte encoding specification.

Dataset reference​

A tuple that identifies how to fetch and verify a dataset (e.g., CID + sha256 + byte length + bytes_schema).

Canonical encoding​

A unique, deterministic serialization for a structure (no alternative encodings). Canonical encodings enable byte-for-byte reproducibility and dispute resolution.

Canonical JSON (JCS)​

RFC 8785 JSON Canonicalization Scheme used where protocol rules define JSON signing and hashing.

Canonical CBOR​

Canonical CBOR byte encoding used for dataset payload bytes when specified by Locate v1 dataset specs.

Merkle tree​

A hash tree constructed over dataset leaves that yields a single 32-byte commitment (the Merkle root).

Merkle root (commitment)​

A 32-byte value committing to an entire dataset. Any party can verify inclusion of an element via an inclusion proof.

Inclusion proof​

A proof that a leaf is included in a committed Merkle tree, usually by providing a sibling path.

Actors and roles​

Prover​

An entity that produces proofs and/or submits claims. A prover may be a user device, a service, or a delegated agent depending on the deployment model.

Verifier​

An entity that checks proofs, validates datasets, and participates in economic security mechanisms (staking, challenges, and dispute resolution).

Challenger​

An entity that disputes or challenges a claim/round outcome and may earn rewards if the challenge succeeds.

Publisher​

The operator identity responsible for Round Header publication and signing. Publisher identity must make tampering detectable even prior to on-chain attestation.

Attestation and settlement​

Flare Web2Json attestation​

A mechanism that attests a Web2-hosted JSON artifact on-chain. The protocol uses this to anchor compact per-round commitments rather than large datasets.

Settlement layer​

The L1 blockchain layer (Flare) where commitments are anchored and disputes are resolved according to rules.

Economic security​

Self-sizing economic security envelope​

A protocol rule that constrains accepted exposure so that the system remains economically secure as stake changes.

Exposure (E_per)​

Per-claim economic exposure (e.g., bond + reward cap).

Security budget (E_max)​

The maximum acceptable exposure for a round given effective stake and assurance constraints.

Effective stake (S_effective)​

Stake counted under protocol rules (often vesting-weighted or otherwise adjusted).

Assurance multiplier (M_assurance)​

A multiplier representing the desired safety margin; higher values reduce usable exposure budget.

Engineering doctrine​

Spec-Driven Development (SDD)​

A development doctrine where code is a compiled artifact of versioned specifications. Normative authority resides in ADRs, specs, schemas, canonical encodings, and conformance vectors.

Determinism boundary​

A rule that consensus-/dispute-critical transforms must be deterministic and reproducible across implementations.

Evidence bundle​

A complete, replayable set of artifacts (inputs, prompts, outputs, hashes, logs) sufficient to reproduce a run.

Notes​

If you encounter a term not defined here, treat it as a documentation bug and add it to this page.

Last updated on