Skip to content

docs(r3): §3.1 emission interrogation — scope correction + GAP findings at HEAD #2829

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
briansrls merged 3 commits into from
May 13, 2026
Merged

docs(r3): §3.1 emission interrogation — scope correction + GAP findings at HEAD #2829

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
7e7e2fa
docs(r3): §3.1 emission interrogation — scope correction + GAP findin…
briansrls May 13, 2026
9613b9e
docs(r3): §3.1 emission — fix omni-demo wording (3-target not 5) + cl…
briansrls May 13, 2026
1f56e65
docs(r3): §3.1 emission — correct overstated L4 in-CI claim (operator…
briansrls May 13, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
33 changes: 28 additions & 5 deletions docs/r3-close-interrogation.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -147,17 +147,40 @@ A close-eligible R3 has every item PROVEN or R4-DEFERRED with operator-recorded

## §3. The emission promises

### §3.1 Omni-emission (5 targets)
### §3.1 Omni-emission (R3 = 3 Shape-A targets: Rust / Python / Go)

**Promise** (THESIS.md:115 + WISHLIST.md R4.A): "automatic parallelism, memoization, omni-emission ... are consequences of the same structural commitments." R3 scope per recent ratification: Rust + Python + Go + C + C++ emission targets.
**Promise** (THESIS.md:115 omni-emission claim + THESIS.md:180 L5 + r3-program-plan.md:185 gate #15 `l5_cross_target_consistency`): "automatic parallelism, memoization, omni-emission ... are consequences of the same structural commitments." R3 scope is **Rust + Python + Go** (Shape A — programming-language targets per THESIS.md:215). C/C++ are **R4.A**-scope (WISHLIST.md:67-73, operator ratification 2026-05-12), not R3. LLVM IR / assembly / machine-code are **R4.C**-scope (WISHLIST.md:107), not R3. Shape B targets (Markdown / OpenAPI / data-shapes) are tracked separately under T-Omni-Shape-B (r3-program-plan.md:430).

**Probes**:

- [ ] Show me a single `.dag` program. Emit it to all 5 targets. Are the outputs runnable?
- [ ] Behavioral parity: same input through each output target. Do they produce equivalent results?
- [ ] Pick the most complex emission target. Is there a non-trivial example that compiles and runs?
- [ ] Show me a single `.dag` program. Emit it to all 3 Shape-A targets (Rust / Python / Go). Are the outputs runnable?
- [ ] Behavioral parity: same input through each output target. Do they produce equivalent results? (THESIS.md:180 L5 claim.)
- [ ] Pick the most complex Shape-A target (Go's goroutine substrate? Python's dynamic dispatch?). Is there a non-trivial example that compiles and runs?
- [ ] **Falsification probe**: program with feature X. Does target Y handle X correctly, or punt to a stub?

**Findings at HEAD (2026-05-13)**:

- **Target substrate inventory** (`src/v3/spec/`): `rust.dag`, `python.dag`, `go.dag` — 3 declared targets matching R3 L5 scope. C/C++/LLVM/assembly correctly absent (R4-scope).
- **L6 data-coverage substrate** (`src/v3/std/cross_target_coverage.dag`): 41-row `emission_path_projections` over (target × form × behavior) cross-product, populated for Rust/Python/Go. Phase-1 carrier-only per Director ratification gunbc#828 2026-05-05.
- **L4 runtime equivalence at HEAD** (oracle-style stdout-parity: compiled emit-target binary stdout = expected fixture stdout; **not** literal artifact byte-equality):
- **CI integration-binary execution state** (`.github/workflows/ci.yml:478-501`): integration harness is **prebuilt** in CI but **execution is HOT-FIX-SKIPPED** via `__HOT_FIX_NONEXISTENT_FILTER__` (zero tests selected) per operator directive at gunbc#846 ("cut all demos and integration tests for now, get v3 to 10 minutes"). Lane2d (Stage 2d integration module, `m1_5_testgen`-class) also HOT-FIX-SKIPPED at `ci.yml:385`. Restore criteria at `ci.yml:501`: OnceLock/cached_compile amortization → per-test wall ≤ 2s ratchet → re-enable per cluster filter (replace nonexistent-filter with cluster-specific filter). **Until restore, in-CI L4 evidence at HEAD is: integration-binary compilation passes (structural exercise of emit code paths) + non-integration test surfaces (lib + bins + determinism_test + doc) only.**
- **Rust**: per-fixture stdout-parity tests EXIST + are unconditional in source (`src/v3/compiler/tests/boundary/m1_3_emit_rust_test.rs:995–1009` and following — `rustc_roundtrip_*` family) BUT live in the integration binary that is HOT-FIX zero-test-filtered above; **execution at HEAD is local-only** (`cargo test -p v3-compiler --test integration rustc_roundtrip` ... locally). Full-matrix `emit_rust_fixtures_rustc_green` at `#[ignore]` (lines 735, 764, 1199, 1218) — local-only, toolchain-gated.
- **Python**: roundtrip tests at `#[ignore]` (`m1_4_emit_python_test.rs:1003, 1070`) — toolchain-gated (python3); local-only.
- **Go**: roundtrip tests at `#[ignore]` (`m1_3_emit_go_test.rs:252, 279, 324`) — toolchain-gated (go); local-only.
- **Omni demo**: Rust-only slice (`emit_omni_demo_rust_roundtrip` at `m1_5_emit_omni_demo_test.rs:106`) **unconditional in source** BUT also lives in integration binary under the HOT-FIX zero-test-filter — execution at HEAD is local-only. Full 3-target receipt (`emit_omni_demo_fixtures_green` at `:125`, Rust + Python + Go) at `#[ignore]`, requires go + python3 toolchains.
- **L5 corpus status** (gate #15 `l5_cross_target_consistency`): DECLARED, RED at HEAD (r3-program-plan.md:243 + :431) — waits on L4 corpus + Shape A grounding ready.

**Open R3 question (PM-surfaced, not yet routed)**:

What's the close-shape for the omni-emission promise?

- **(a) L6 data-coverage interpretation**: 41 (target × form × behavior) rows declared in v3-side data = ✓ for Rust/Python/Go. Structural-fold property, no runtime evidence needed.
- **(b) L4 runtime stdout-parity interpretation**: compiled emit-target binary stdout equals expected fixture stdout across the corpus. At HEAD this is **runnable locally** across all 3 targets (Rust unconditional + Python/Go via `--ignored`) but **not actually executed in CI for any target** because the integration binary is HOT-FIX zero-test-filtered (`ci.yml:478-501`). Three sub-shapes: (b1) restore integration harness CI execution per `ci.yml:501` restore-criteria before R3 close (depends on OnceLock/cached_compile amortization landing), (b2) explicit acceptance that R3-close evidence-bar is "data-coverage + local-runtime + integration-binary-compilation-passes" with full CI runtime tied to a separate fast-follow gate, (b3) gate-class promotion: add the integration-harness-execution-state restore to a new §1.8 gate explicitly anchored to R3-close.

The two interpretations differ on whether locally-executable runtime roundtrips count as R3-closure-evidence when CI doesn't actually run them. THESIS.md:180 ("L5: **same .dag produces same behavior** in Rust/Python/Go") reads as runtime-shape; current **in-CI** evidence is L6 data-coverage-for-all-three + integration-binary-prebuild-passes; current **runnable** evidence is per-fixture Rust + (with toolchains) Python/Go locally.

**Falsification candidate** (under interpretation (b) only): pick a non-trivial fixture with Python-specific or Go-specific structural sensitivity (e.g., variant-with-payload pattern-matching, fold composition). Run roundtrip locally via `cargo test ... -- --ignored`. Does it pass without toolchain-skip? If not, that's the gap shape.

### §3.2 Workflow-as-data

**Promise** (r3-program-plan.md §1.8 #82 / Cluster F: `ci_workflow_modeled_as_dag`): CI workflow is itself a `.dag` program, not hand-authored YAML.
Expand Down
Loading