Reference Vectors
The canonical examples and test cases through which expected Foundational Survey Fabric results become independently reproducible.
Foundational Survey Fabric Reference Vectors define the exact input-and-expected-result cases through which independent implementations prove agreement with the governing Specification and Conformance framework.
Reference Vectors do not select Survey mathematics. The Requirements establish what must be satisfied. The Specification defines the mathematics that satisfies those requirements. Conformance defines how an implementation proves compatibility. Reference Vectors provide canonical examples, negative cases, adversarial cases, and machine-verifiable expected results derived from that already-defined mathematics.
All twelve Reference Vector sections are reconciled against the canonically adopted FSF-SPEC-1.0 and the established Conformance architecture. The corpus demonstrates behavior defined by that Specification and preserves explicit boundaries around capabilities outside the adopted production profile.
Reference Vector Question Given this exact input, what exact answer must every conforming implementation produce?
Current Corpus Standing
Reference Vector Standing
The twelve-section framework demonstrates the adopted FSF production profile through exact executable, negative, compatibility, and adversarial fixtures. Capability outside FSF-SPEC-1.0 remains explicitly unsupported or reserved rather than being invented by the test corpus.
Adopted executable profile: CRPC exact coordinates and arithmetic; canonical Point / Segment / SCPE geometry; exact predicates; deterministic normalization; ECEM meaning preservation; exact SCPE area; canonical orientation frame; adopted exact transformation behavior; FSF-CJSON-1.0 canonical interchange; invalid / unsupported separation; Conformance evidence and compatibility testing; and adversarial stress across those defined behaviors.
Canonical Survey Domain: H = 1,000,000 Pang and D = [-1,000,000,+1,000,000]². Domain-edge and corner fixtures may now use exact canonical numerical values rather than parameterized placeholders.
Governed production placement envelope: P = [-500,000,+500,000]² Pang, with the required relationship World-space ⊂ interior(P) ⊂ interior(D).
Reserved outside the current production profile: unrestricted Boolean / composite-result closure and other extended mathematics not adopted by FSF-SPEC-1.0. Reference Vectors may demonstrate unsupported boundaries but may not create new Specification behavior.
Reference Vector Structure
Addressing Vectors
Tests CRPC Point addressing, exact equality and ordering, ECEM exact-place semantics, frame interpretation, canonical serialization, and version-preserved meaning while keeping numerical Domain capacity separate.
CRPC · Point · ECEM · Equality · Ordering →Survey Domain Limit Vectors
Tests the adopted closed-square Survey Domain D = [-1,000,000,+1,000,000]², including origin, interior, boundary, corners, outside classification, ECEM stability, and separation from World membership.
Origin · Canonical Domain · Closed Boundary · Validity →Pang Measurement Vectors
Tests uniform Pang scale, exact coordinate differences, rational sub-Pang values, exact SCPE area, normalization-invariant measure, and lossless versus lossy measurement representation.
Pang · CRPC Difference · Exact Area · Sub-Pang →Orientation Vectors
Tests Pankor, Panvath, Panoris, Panvel, the right-handed canonical frame, Panoris zero direction, counterclockwise positive rotation convention, and exact determinant orientation.
Axes · Handedness · Zero Direction · Orientation →Geometry Vectors
Tests Point, Segment, and SCPE validity, exact predicates, Segment intersection, Point-in-SCPE, containment, connectedness, normalization, and semantic geometry equality.
Point · Segment · SCPE · Exact Predicates →Normalization Vectors
Tests CRPC reduction, canonical zero, Segment endpoint order, SCPE traversal and start normalization, redundant-collinear removal, idempotence, no snapping, and canonical convergence.
CRPC · Geometry · Idempotence · Canonical Convergence →Transformation Vectors
Tests the restricted exact transformation profile: translation, positive uniform scale, identity, inverse, composition, orientation preservation, exactness preservation, and Survey Domain validity.
Translation · Uniform Scale · Inverse · Composition →Precision Vectors
Tests CRPC exact rational precision, ECEM meaning preservation, sub-Pang exactness, canonical/storage/rendering separation, lossless conversion, lossy reduction, and exact round-trip preservation.
CRPC · ECEM · Sub-Pang · Lossless / Lossy →Invalid-Input Vectors
Tests deterministic rejection and classification for malformed CRPC, invalid geometry, out-of-Domain input, ambiguous input, invalid FSF-CJSON, unsupported versions or profiles, snapping traps, and unsupported operations.
Invalid · Unsupported · Reject · No Silent Repair →Serialization Vectors
Tests FSF-CJSON-1.0 canonical object encoding, exact CRPC preservation, deterministic key order, UTF-8 canonical bytes, required fields, canonical round-trip, and cross-implementation exchange.
FSF-CJSON-1.0 · Canonical Bytes · Round-Trip →Version / Conformance Vectors
Tests compatible meaning preservation, Mandatory Core PASS / FAIL, valid-but-unsupported capability, profile scope, evidence traceability, declaration consistency, independent reproduction, and evaluator agreement.
Compatibility · PASS · FAIL · UNSUPPORTED · Evidence →Pathological Edge Cases
Adversarially stresses exact predicates, normalization fixed points, extreme CRPC, Domain classification, Point-in-SCPE, exact area, supported transformation chains, canonical serialization, hidden-state independence, and Specification ambiguity.
Adversarial · Stress · Hidden State · Ambiguity Exposure →Together, these twelve sections define the Reference Vector architecture for the adopted Foundational Survey Fabric production profile. The corpus covers exact reference, canonical Domain classification, Pang-based mathematics, orientation, geometry, normalization, transformation, precision, invalid-input handling, canonical interchange, version / Conformance behavior, and adversarial stress.
The corpus remains subordinate to Requirements, FSF-SPEC-1.0, and Conformance. It may prove, expose, compare, reproduce, and stress behavior that the Specification defines. It may not invent spatial truth, create unsupported result types, redefine canonical authority, or allow higher Architecture domains to leak into FSF outcomes.
The governing development rule is now straightforward: build exact executable vectors for adopted behavior; classify non-adopted behavior as unsupported or reserved; and never use a fixture to manufacture new Specification mathematics.
Architectural Sequence Requirements establishes what must be satisfied. Specification defines the mathematics that satisfies it. Conformance defines how an implementation proves it follows the Specification. Reference Vectors provide canonical examples and test cases proving the expected results.
Reference Vector Standing Requirements ownership and reverse traceability remain closed across all twelve sections. The corpus now demonstrates the adopted FSF-SPEC-1.0 production profile, including the canonical Survey Domain and canonical serialization, while preserving explicit unsupported / reserved boundaries beyond that profile.