Skip to content

Commit 1186dfd

Browse files
authored
Merge pull request #434 from AdaWorldAPI/claude/splat3d-cpu-simd-renderer-maoo0
docs: unified-soa-convergence-v1 integration plan + handover (ONE SoA end-to-end, 9 consumers, Libet −550 ms kanban, Lance 6.0.1 stack)
2 parents 38b633b + 7c28967 commit 1186dfd

6 files changed

Lines changed: 1004 additions & 0 deletions

.claude/board/EPIPHANIES.md

Lines changed: 24 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -1,3 +1,27 @@
1+
## 2026-05-29 — E-SOA-IS-THE-ONLY — there is ONE SoA, never transformed; mailbox SoA mutation IS the hot path; Libet −550 ms anchors the Rubicon kanban in surrealkv-on-lance; SPO-W witness is a *pointer* via the belief-state arc
2+
3+
**Status:** CONJECTURE / design (user-stated 2026-05-29, post-PR-#433). Records five layered rulings; details in `bindspace-singleton-to-mailbox-soa-v1.md` §11.1–§11.5. Author-stated; not council-gated.
4+
5+
1. **One SoA, never transformed (§11.1).** The mailbox SoA is the *single* universal carrier across the stack — never re-encoded into a different shape. Only three operations are allowed on it: (a) cognitive-shader thinking (hot path), (b) cold-path read/write to LanceDB, (c) AriGraph Markov-chain context building. *Any change in any mailbox SoA = the only hot-path activity.* Today's `crates/lance-graph` containers are *cold-path-adjacent thinking* — only accidentally aligned. Realignment (`D-MBX-7`): lance-graph containers ≡ `MailboxSoA` layout ≡ `ndarray::simd_soa.rs` alignment, unlocking **1.4–4.2× SIMD acceleration**. Nice-to-have today; **hard prerequisite** for the SurrealDB transparent view (§2.7).
6+
7+
2. **Mailbox = full BindSpace, reinvented as LE; witness = belief-state arc (§11.2).** Carry everything BindSpace had, but as LE-contract types (never `Vsa16kF32`). The **witness IS the per-row arc of `CausalEdge64` emissions** (`CollapseGateEmission` arc) — that arc *implicitly* documents NARS revision because every emission stamps `confidence_u8` + `inference_mantissa`; reading the arc IS the `(frequency, confidence)` trace. **No separate revision log column.** D-MBX-A1 columns (`edges`/`qualia`/`meta`/`entity_type`) landed; D-MBX-A2 closes remaining BindSpace-expressivity gaps; D-MBX-A3 adds the witness-arc handle column.
8+
9+
3. **Libet −550 ms anchors the Rubicon (§11.3).** Concretises `E-RUBICON-RACTOR` with wall-clock: pre-decisional = counterfactual deliberation; **commit at t = −550 ms** (Libet readiness-potential anchor) = Σ10 commit = ractor START of the actional phase; Libet "free won't" = pre-(−550 ms) CollapseGate veto / ghost-tier preempt; post-actional → ractor STOP → tombstone. The **Rubicon kanban lives in `surrealkv`-on-lance** (a *view* over leading LanceDB, §2.7), columns = the 4 action phases; ractor lifecycle transitions = kanban moves. (`D-MBX-8` timing anchor; `D-MBX-9` kanban view.)
10+
11+
4. **SPO-W witness is a *pointer*, not stored data (§11.4).** The witness is an arc-handle into the belief-state-arc array; the pointer lives equivalently in the **mailbox SoA**, the **kanban row**, or the **mailbox index**. **The SoA itself decides** whether to commit a belief as a fact-with-witness into (a) other mailboxes (inter-mailbox baton handoff carrying the pointer) or (b) cold-path facts (LanceDB SPO-G calcification linking back to the arc). No storage redundancy: the witness lives where the arc lives. (`D-MBX-A5`.)
12+
13+
5. **Counterfactual Staunen × Wisdom = plasticity spreaders (§11.5).** When a mailbox is pre-Rubicon (counterfactual phase), high Staunen × Wisdom *spreads* plasticity beyond the focal row (Hebbian spread). Hot-path-only — the spread is a mailbox-SoA mutation, never a side channel. Radius/decay TBD. (`D-MBX-A4`.)
14+
15+
**Refinements (2026-05-29, same session):**
16+
17+
- **§11.3 kanban refined.** The Rubicon kanban has **4 explicit columns**: **Planning** (ractor mailbox owned SoA, counterfactual) → **Cognitive work** (Σ10 commit + actional SoA mutation) → **Evaluation of goalstate** (post-actional reflection on witness arc) → **Commit · Plan · Prune** (terminal 3-way decision: calcify / re-deliberate / ghost-preempt). Supersedes the earlier 4-Heckhausen-phase mapping in this entry.
18+
- **§11.4 sharpened.** "Witness in other mailboxes" means a *pointer into the AriGraph episodic Markov chain* — the temporal sequence of mailbox states *is* the episodic memory substrate (CLAUDE.md "The Click": *"AriGraph, episodic memory, SPO, CAM-PQ are thinking tissue — not storage"*). No parallel episodic structure exists.
19+
- **§11.6 — the "half-baked nine" all consume THE same SoA from A-Z.** AriGraph · Markov-grammar Vsa16kF32 substrate · BindSpace · `lance-graph` cold containers · `lance-graph-planner` · `cognitive-shader-driver` · `lance-graph-callcenter` · `lance-graph-ontology` (read-only, AS IS) · thinking-styles/atoms — every one consumes THE single SoA byte layout at every boundary it crosses. The SoA gets a **version byte at the layout root** so older bytes stay readable after schema upgrade (`D-MBX-10`, governed by `I-LEGACY-API-FEATURE-GATED`). For SurrealDB to provide a transparent (zero-copy) view, versioning aligns with the **Lance 6.0.1 / LanceDB 0.29 / DataFusion 53** stack — only one bump pending: `lance =6.0.0 → =6.0.1` patch (`D-MBX-11`); arrow/datafusion/lancedb already on target. The kanban/ractor lifecycle requires a **lance-graph-planner DTO surface overhaul** (`D-MBX-A6`) re-expressing planner DTOs as operations on the SoA + 4-phase kanban transitions.
20+
21+
**Cross-ref:** `E-BATON-1`, `E-MAILBOX-IS-BINDSPACE`, `E-RUBICON-RACTOR` (Heckhausen + Libet origin), `E-LADDER-SERVES-MAILBOX` (§5 ghost preempt = Libet veto, §6 tombstone), `I-VSA-IDENTITIES`, `I-LEGACY-API-FEATURE-GATED` (governs the SoA version gate), `causaledge64-mailbox-rename-soa-v1` §10 + E-CE64-MB-8 (Σ10 router), `linguistic-epiphanies-2026-04-19.md` E21 (Σ10 Rubicon tier doctrine), D-CSV-10 (`SigmaTierRouter` shipped #388), `bindspace-singleton-to-mailbox-soa-v1` §11.1–§11.6 (full ruling + D-MBX-A2/A3/A4/A5/A6/7/8/9/10/11/12 + OQ-11.1–11.8); `epiphany-brainstorm-council` (PR #433 pre-merge gate — bypassed here because rulings are author-stated).
22+
23+
---
24+
125
## 2026-05-28 — E-NORMALIZED-ENTITY-1 — `NormalizedEntity<Stage>` is the single typed carrier holding the four-way inheritance chain (Odoo → OGIT → OWL → DOLCE → FIBU/FIBO); stage advancement is typestate, not method calls on a context
226

327
**Status:** FINDING (architectural unification). Drives `normalized-entity-holy-grail-v1`. The carrier is a typed lens into a `MailboxSoA` row — it does NOT own the four cognitive columns; the mailbox does. The 4-way inheritance slots (`odoo`/`ogit`/`owl`/`dolce` + optional `fibu`) populate as stages advance; phantom-typed `Stage` parameter forbids out-of-order traversal at compile time. Consumers chain ON the carrier (`entity.resolve_ogit(ctx).hydrate_owl(ctx)...`), never reach into its internals.

.claude/board/INTEGRATION_PLANS.md

Lines changed: 10 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -1,3 +1,13 @@
1+
## 2026-05-29 — unified-soa-convergence-v1 (THE single LE SoA end-to-end across the workspace: 9 half-baked consumers + SoA version gate + Lance 6.0.1/LanceDB 0.29/DataFusion 53 alignment + 4-phase Rubicon kanban + lance-graph-planner DTO overhaul + Libet −550 ms anchor + Staunen×Wisdom plasticity spreader + SPO-W pointer via AriGraph episodic Markov chain)
2+
3+
**Status:** PROPOSAL / integration plan. Design-spec only, no code. **Plan file:** `.claude/plans/unified-soa-convergence-v1.md`. **Handover:** `.claude/handovers/2026-05-29-1825-soa-convergence-author-to-impl.md`. **Epiphany:** `E-SOA-IS-THE-ONLY` (+ refinements §11.3/4/6 in PR-this-one).
4+
**Owns:** the meticulous integration sequence for the 5 layered user rulings ratified this session — one SoA never transformed; mailbox = full BindSpace as LE; Libet −550 ms ratchets the Rubicon kanban; SPO-W = pointer via AriGraph episodic Markov chain; counterfactual Staunen×Wisdom = plasticity spreaders. Adds D-MBX-A2/A3/A4/A5/A6, D-MBX-7/8/9/10/11/12 (12 sub-PRs); 8 open questions OQ-11.1–11.8; risk matrix; phased dependency graph (P0–P7).
5+
**Anchored iron rules:** I-VSA-IDENTITIES, I-LEGACY-API-FEATURE-GATED (governs the SoA version gate), I-SUBSTRATE-MARKOV (untouched), I-NOISE-FLOOR-JIRAK.
6+
**Predecessors:** `bindspace-singleton-to-mailbox-soa-v1` (§11 rulings); `causaledge64-mailbox-rename-soa-v1` (canonical 5-crate + 7-PR plan); `cognitive-substrate-convergence-v1/v2/v3` (Σ10 Rubicon shipped #388).
7+
**Council bypass note:** the underlying §11 epiphanies are author-stated (user) — `epiphany-brainstorm-council` (shipped #433) is bypassed for them; the plan's *spec content* IS open to council review via PR.
8+
**Workspace stack verified 2026-05-29:** arrow 58 ✓ · datafusion 53 ✓ · lancedb =0.29.0 ✓ · lance =6.0.0 → **=6.0.1** (one patch bump pending, D-MBX-11).
9+
10+
---
111
## 2026-05-28 — normalized-entity-holy-grail-v1 (typed unified normalization + Op chain over OGIT/OWL/DOLCE/Odoo with three-context execution — the trunk that unifies BP-1 + EXT-1..6 + jit + MailboxSoA into one consumer surface)
212

313
**Status:** PROPOSAL. The trunk plan that ties together every prior architectural deliverable into ONE typed consumer pipeline grammar. Closes the structural gap that lets consumers re-implement business logic in regex / hand-rolled pseudo-code by making such interpretation a MISSING FUNCTION (per `E-CONSUMER-CANNOT-INTERPRET-1`) rather than a code-review concern. Re-encodes Odoo's three-regime decomposition (interactive / bulk / periodisch) as compile-time typestate, not stringly-typed `env.context` flags.

.claude/board/STATUS_BOARD.md

Lines changed: 21 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -546,6 +546,27 @@ hot/cold/feedback loop onto our contract types + SoA floor
546546

547547
---
548548

549+
## unified-soa-convergence-v1 — ONE LE SoA end-to-end across 9 consumers + version gate + Lance 6.0.1 stack + 4-phase Rubicon kanban
550+
551+
Plan path: `.claude/plans/unified-soa-convergence-v1.md`. Handover `.claude/handovers/2026-05-29-1825-soa-convergence-author-to-impl.md`. Epiphany `E-SOA-IS-THE-ONLY` (+ §11.3/4/6 refinements).
552+
553+
| D-id | Title | Crate(s) | ~LOC | Risk | Status | PR / Evidence |
554+
|---|---|---|---|---|---|---|
555+
| D-MBX-A1 | migrated thoughtspace columns landed on `MailboxSoA<N>` (`edges`/`qualia`/`meta`/`entity_type`) | cognitive-shader-driver | 60 | LOW | **Shipped** | between #418 and #433 (verified `mailbox_soa.rs` 2026-05-29) |
556+
| D-MBX-A2 | close BindSpace expressivity gaps in `MailboxSoA<N>` (`content_ref`, S/P/O role slices, temporal/expert fold per OQ-2) | cognitive-shader-driver + contract | 140 | MED | **Queued** | gates on D-CE64-MB-1-impl + OQ-1/OQ-2 |
557+
| D-MBX-A3 | `witness_arc: [u32; W]` per-row column (the belief-state arc handle into AriGraph episodic Markov chain) | cognitive-shader-driver | 100 | MED | **Queued** | gates on D-MBX-A2 + OQ-11.2 |
558+
| D-MBX-A4 | Staunen × Wisdom counterfactual plasticity spreader (Hebbian, hot-path-only, Planning-gated) | cognitive-shader-driver | 80 | LOW | **Queued — design** | gates on D-MBX-A3 + OQ-11.1 + `phase` field |
559+
| D-MBX-A5 | SPO-W witness pointer dual-residency (SoA / kanban / mailbox index); SoA decides commit modality (chain pointer vs cold fact) | cognitive-shader-driver + AriGraph SPO-G | 150 | HIGH | **Queued** | gates on D-MBX-A3 + D-MBX-4 |
560+
| D-MBX-A6 | `lance-graph-planner` DTO surface overhaul: DTOs as SoA-row-lenses; planner output = `KanbanMove`s; 5-phase feature-gated cutover (OQ-11.7) | lance-graph-planner + contract | 600 | HIGH | **Queued** | gates on D-MBX-10 + D-MBX-8 + OQ-11.7 |
561+
| D-MBX-7 | `lance-graph` containers ≡ `MailboxSoA` layout ≡ `ndarray::simd_soa.rs`-aligned (1.4–4.2× SIMD payoff; hard prereq for SurrealDB transparent view) | lance-graph + ndarray | 300 | HIGH | **Queued** | gates on D-MBX-A2 + D-MBX-10 + D-MBX-11 + PR-NDARRAY-MIRI-COMPLETE |
562+
| D-MBX-8 | Σ10 commit stamps **t = −550 ms** wall-clock (Libet anchor) in `SigmaTierRouter`; downstream ractor START fires | sigma-tier-router + shader-driver | 120 | MED | **Queued** | gates on D-MBX-A4 + D-MBX-A6 Phase 1 |
563+
| D-MBX-9 | Rubicon kanban view in `surrealkv`-on-lance (4 columns: Planning · Cognitive work · Evaluation · Commit·Plan·Prune); ractor lifecycle hooks = kanban moves | surreal_container + ractor | 250 | HIGH | **Queued** | gates on D-MBX-7 + D-MBX-8 + surreal_container BLOCKED(B/C/D) resolved (OQ-11.6) + D-PERSONA-5 |
564+
| D-MBX-10 | SoA version byte at layout root (`MailboxSoAHeader`); refuse v(N>M) bytes on v(M) reader; field-isolation matrix tests on every column op (`I-LEGACY-API-FEATURE-GATED` discipline) | lance-graph-contract | 100 | HIGH | **Queued** | foundation — should land early in P2; gates on OQ-11.5 |
565+
| D-MBX-11 | Lance `=6.0.0 → =6.0.1` patch bump (5 Cargo.toml files identified) | workspace Cargo.toml | 10 | LOW | **Queued (mechanical)** | none — can land in parallel with par-tile prereq |
566+
| D-MBX-12 | 8-PR workspace-wide consumer alignment: 12.1 AriGraph · 12.2 Vsa16k audit · 12.4 lance-graph · 12.5 planner · 12.6 shader-driver · 12.7 callcenter · 12.8 ontology audit · 12.9 thinking-styles | per-crate | 800 | per-PR | **Queued (multi-PR)** | sequencing per OQ-11.8: 12.4 → 12.5 → 12.6 → 12.7 → 12.1 → 12.9 → 12.2 → 12.8 |
567+
568+
---
569+
549570
## bindspace-singleton-to-mailbox-soa-v1 — dissolve `Arc<BindSpace>` into per-mailbox `MailboxSoA<N>`
550571

551572
Plan path: `.claude/plans/bindspace-singleton-to-mailbox-soa-v1.md`. Epiphany `E-MAILBOX-IS-BINDSPACE`. Migration of the shared singleton address space into mailbox-owned ephemeral thoughtspace (LE-contract SoA columns); drops the 64 KB `Vsa16kF32` `cycle` plane.

0 commit comments

Comments
 (0)