Handout / ros-171-2026-10-10Research note

ROS 1.7.1 · Canonical Materialization

Causal completeness, constraints and a consistent current research state.

Normative release with a local reference implementation: causal coverage and feasibility from 1.7.0, deterministic projection and adversarial regressions from 1.7.1.

  • ROS
  • canonical state
  • red team
  • constraints
  • release handout
01

From 1.6 to 1.7.0

Causal Coverage Audit looks for missing material drivers and tests the sensitivity of central claims. HARD, SOFT and UNKNOWN govern feasibility by pathway, time and jurisdiction. Causal claims receive an identification level M0–M4; Impact Ledgers record countervailing effects. LEVEL 2 requires operational transitions, countermodels and observation boundaries rather than symbolic labels alone.

02

History ≠ Current State

Append-only applies to research history; consolidation applies to current state. An ordered EventHistory is materialized into exactly one active revision per UID. Sources, Claims, Hypotheses and Expectations/Predictions are projected in full from the same fingerprint; change appendices must not form competing main registers.

03

QG-RS-01

Canonical State Materialization & Referential Closure checks resolvable typed references, unique short IDs, reciprocal hypothesis/test links, source provenance and unchanged ex ante conditions. A SYSTEM_MODEL_UPDATED with obsolete main registers must receive FAIL.

04

QG-RT-01

Red-team findings, too, are claims subject to verification. Preconditions, counterexamples, dispositions, remedies and regressions must be retained. ACCEPTED, PARTIALLY_ACCEPTED, REJECTED and UNRESOLVED must not collapse into each other; refuted criticism must not be accepted as fact.

05

Operational Projection & Migration

The new root schema requires projection_seq, event_head_hash and canonical_fingerprint; state_projection and register_integrity_assessment extend the object model. ros_reference.py provides a local deterministic reference, not a transactional production backend. Migration preserves IDs and historical predictions; missing history is documented as a provenance gap. A hash chain without an externally anchored head is not authenticated tamper protection.

06

Latest, Pins & Components · Historical Status

The local resolver supplied on October 10, 2026 mapped ROS@latest to ROS@v1.7.1. Since the promotion of October 15, 2026, ROS@latest=1.7.2; this handout remains a historical 1.7.1 snapshot. Explicit historical pins remain intact; the operational 1.6.1 overlay is not migrated retrospectively. NIS adopts applicable ROS gates without automatic version upgrade; BIS remains independent at 0.3.1.

07

Quality & Limitations

The release reports 40/40 passing synthetic pytest regressions. Schema validation alone proves neither semantic closure nor truth. QG-AR-01 independently examines each export. According to the release audit, the older diesel revision 1.1 remains QG-RS-01 FAIL; the new methodology does not automatically repair that research artifact.