Skip to content

Commit d55b116

Browse files
authored
docs, comments and diagnostics: describe behaviour in plain words instead of internal tracking labels (#580)
Lands as a member of an ordered engine stack that was verified once on its top: one whole-population parity sweep and every verify step on the top tree, recorded for each member head.
1 parent e4d6703 commit d55b116

267 files changed

Lines changed: 639 additions & 638 deletions

File tree

Some content is hidden

Large Commits have some content hidden by default. Use the searchbox below for content that may be hidden.

‎CHANGELOG.md‎

Lines changed: 58 additions & 58 deletions
Large diffs are not rendered by default.

‎README.md‎

Lines changed: 4 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -61,7 +61,7 @@ runtime supports selected windows and `[]` otherwise.
6161
From 1.6.2, the engine/codegen image advertises selected-window
6262
contract revision `1.3` through `docker/run_json.py --capabilities-json`:
6363
with the section-4 producers and manifest present and value ownership
64-
`complete` (EN-115), it emits the four typed codes (`strategy_value_invalid`,
64+
`complete`, it emits the four typed codes (`strategy_value_invalid`,
6565
`engine_command_value_invalid`, `report_result_invalid` and
6666
`report_post_execution_failed`). A build declaring `1.3` pairs only with a
6767
`complete` receipt; `1.3-parity` remains the parity build's literal, and a
@@ -717,9 +717,9 @@ Documentation: [C ABI reference](https://cdocs.pineforge.dev) · [Getting starte
717717
## Releases
718718

719719
- **[v1.7.0](https://github.com/pineforge-4pass/pineforge-engine/releases/tag/v1.7.0)** (2026-10-25) — report-sink transport, scratch-backed history and bounded storage admission, with new C++ views and command-line report routes. `get_trade()` and `closed_trade()` now return `Trade` by value, working-book, journal and cohort accessors return views, and some reads can throw on scratch I/O; closing a trade writes `TradeLedger` scratch even without a report sink, so a run needs a disk-backed `TMPDIR` or an explicit scratch directory; the C++ epoch namespace is still `engine_script_run_v19` although the layout of `BacktestEngine` changed, so rebuild every C++ object that includes an engine header, regenerate strategy C++ with the matching codegen and relink. Pair with [codegen 1.7.0](https://github.com/pineforge-4pass/pineforge-codegen-oss/releases/tag/v1.7.0). The known limits of this release (slower than 1.6.3 on some scripts, the closed-trade cache and its size by entry point) are in [CHANGELOG.md](CHANGELOG.md#170--2026-10-25), section "Costs that remain".
720-
- **[v1.6.3](https://github.com/pineforge-4pass/pineforge-engine/releases/tag/v1.6.3)** (2026-10-23) — CI, release-gate and test-only changes after 1.6.2: the engine's CI runs on the GCP fleet (`pf-gcp-pr`, `pf-gcp-main`), fork-gated; the PR wall is split from full-event verification (CR-233 A+B); a release requires full verify and parity green first (C2); the live test matches delivery receipts by request sequence (CR-223); a session tape comment no longer names a private repository. Nothing under `src/`, `include/`, `cmake/` or `runner/` changed since v1.6.2, and in `docker/` only the run-failure code catalog diff (`run_failure_codes_diff.json`) did. Pair with codegen 1.6.3. See [CHANGELOG.md](CHANGELOG.md#163--2026-10-23).
721-
- **[v1.6.2](https://github.com/pineforge-4pass/pineforge-engine/releases/tag/v1.6.2)** (2026-10-21) — selected-window contract revision `1.3` again, declared with the EN-115 completeness proof (`capabilities.selected_window_contract_revisions: ["1.3"]` on a supported runtime and `[]` otherwise; 1.6.1 declared `1.3-parity`); proportional annualisation for runs shorter than a day, never null (CR-202, W-73); the ownership probes fixed so proof B reads 118/118 (CR-139) and the ownership manifest-controls test sharded four ways (CR-196); CI changes: Linux x64 jobs on the pf-ci-spot pool and release jobs on pf-ci-release, `timeout-minutes` removed, the actions of `promote-baseline.yml` pinned by SHA, the PR tier skipping the slow rows (D1, CR-195). It ships the [v1.6.2-rc.1](https://github.com/pineforge-4pass/pineforge-engine/releases/tag/v1.6.2-rc.1) candidate (2026-10-19). Pair with codegen 1.6.2, regenerate and relink. See [CHANGELOG.md](CHANGELOG.md#162--2026-10-21).
722-
- **[v1.6.1](https://github.com/pineforge-4pass/pineforge-engine/releases/tag/v1.6.1)** (2026-10-18) — selected-window contract revision `1.3-parity` (`capabilities.selected_window_contract_revisions: ["1.3-parity"]` on a supported runtime and `[]` otherwise) rather than `1.3` until a completeness proof exists (EN-115), with the four typed value-ownership codes (`strategy_value_invalid`, `engine_command_value_invalid`, `report_result_invalid` and `report_post_execution_failed`) emitted and value parity with the 1.4.0 reference on the tested cases; per-stage `startedAt` and `seconds` in `preflight-summary.json`; `noindex` on raw Doxygen source listings and their omission from `sitemap.xml`; verifier-tests sharded across parallel `unittest` subprocesses; the lower-feed timing ratio skipped only under AddressSanitizer. The C layouts stay at ABI 4, `PF_ABI_VERSION` stays 4, the selected-window wire version stays 1, the phase version 1 and the fingerprint/schema version 2. Pair with codegen 1.6.1, regenerate and relink. See [CHANGELOG.md](CHANGELOG.md#161--2026-10-18).
720+
- **[v1.6.3](https://github.com/pineforge-4pass/pineforge-engine/releases/tag/v1.6.3)** (2026-10-23) — CI, release-gate and test-only changes after 1.6.2: the engine's CI runs on the GCP fleet (`pf-gcp-pr`, `pf-gcp-main`), fork-gated; the PR wall is split from full-event verification (the CI split); a release requires full verify and parity green first (C2); the live test matches delivery receipts by request sequence; a session tape comment no longer names a private repository. Nothing under `src/`, `include/`, `cmake/` or `runner/` changed since v1.6.2, and in `docker/` only the run-failure code catalog diff (`run_failure_codes_diff.json`) did. Pair with codegen 1.6.3. See [CHANGELOG.md](CHANGELOG.md#163--2026-10-23).
721+
- **[v1.6.2](https://github.com/pineforge-4pass/pineforge-engine/releases/tag/v1.6.2)** (2026-10-21) — selected-window contract revision `1.3` again, declared with the completeness proof (`capabilities.selected_window_contract_revisions: ["1.3"]` on a supported runtime and `[]` otherwise; 1.6.1 declared `1.3-parity`); proportional annualisation for runs shorter than a day, never null (short-run annualisation); the ownership probes fixed so proof B reads 118/118 and the ownership manifest-controls test sharded four ways; CI changes: Linux x64 jobs on the pf-ci-spot pool and release jobs on pf-ci-release, `timeout-minutes` removed, the actions of `promote-baseline.yml` pinned by SHA, the PR tier skipping the slow rows (D1). It ships the [v1.6.2-rc.1](https://github.com/pineforge-4pass/pineforge-engine/releases/tag/v1.6.2-rc.1) candidate (2026-10-19). Pair with codegen 1.6.2, regenerate and relink. See [CHANGELOG.md](CHANGELOG.md#162--2026-10-21).
722+
- **[v1.6.1](https://github.com/pineforge-4pass/pineforge-engine/releases/tag/v1.6.1)** (2026-10-18) — selected-window contract revision `1.3-parity` (`capabilities.selected_window_contract_revisions: ["1.3-parity"]` on a supported runtime and `[]` otherwise) rather than `1.3` until a completeness proof exists, with the four typed value-ownership codes (`strategy_value_invalid`, `engine_command_value_invalid`, `report_result_invalid` and `report_post_execution_failed`) emitted and value parity with the 1.4.0 reference on the tested cases; per-stage `startedAt` and `seconds` in `preflight-summary.json`; `noindex` on raw Doxygen source listings and their omission from `sitemap.xml`; verifier-tests sharded across parallel `unittest` subprocesses; the lower-feed timing ratio skipped only under AddressSanitizer. The C layouts stay at ABI 4, `PF_ABI_VERSION` stays 4, the selected-window wire version stays 1, the phase version 1 and the fingerprint/schema version 2. Pair with codegen 1.6.1, regenerate and relink. See [CHANGELOG.md](CHANGELOG.md#161--2026-10-18).
723723
- **[v1.6.0](https://github.com/pineforge-4pass/pineforge-engine/releases/tag/v1.6.0)** (2026-10-17) — selected-window contract revision `1.3` (`capabilities.selected_window_contract_revisions: ["1.3"]` on a supported runtime and `[]` otherwise; 1.6.0 declared contract revision `1.3`; corrected to `1.3-parity` in 1.6.1); typed failures across source value admission (`strategy_value_invalid`), derived execution and settlement arithmetic (`engine_command_value_invalid`), non-finite canonical trade or summary doubles (`report_result_invalid`) and post-boundary serialization faults (`report_post_execution_failed`), with every report writer enforcing JSON so no bare `NaN` or `Infinity` token is emitted; Doxygen favicon, 404 page, `robots.txt`, `sitemap.xml`, canonical URLs and tag-or-`unreleased` version label; CI deadline and timeout-flag removals, private-fixture test moves and benchmark comparison/quiet-gate updates. The C layouts stay at ABI 4, `PF_ABI_VERSION` stays 4, the selected-window wire version stays 1, the phase version 1 and the fingerprint/schema version 2. Pair with codegen 1.6.0, regenerate and relink. See [CHANGELOG.md](CHANGELOG.md#160--2026-10-17).
724724
- **[v1.5.0](https://github.com/pineforge-4pass/pineforge-engine/releases/tag/v1.5.0)** (2026-10-12) — recorded outputs: a module can record per-bar doubles in named slots, run-level doubles and events, read through the nine `pf_outputs` runtime exports with coded refusals, off until a caller turns them on and byte-identical when unused; a run's report belongs to the call that began it; the engine's test inputs and strategy scripts moved to a private repository, with tests that need them skipping by name, a LEGAL.md *Data not distributed* note and a maintainers' private-fixture overlay. The C layouts stay at ABI 4, `PF_ABI_VERSION` stays 4 and the new members enter `engine_script_run_v19` without an epoch. Pair with codegen 1.5.0, regenerate and relink. See [CHANGELOG.md](CHANGELOG.md#150--2026-10-12).
725725
- **[v1.4.0](https://github.com/pineforge-4pass/pineforge-engine/releases/tag/v1.4.0)** (2026-10-08) — two additive C failure getters; typed input/override provenance with explicit unresolved reasons and duplicate-title counts; coded failure JSON and checked-setting refusals; Pine exit/fill-location corrections that can change trades, reports and hash values on the affected paths; chart-timeframe live input and final-delivery draining; feed ownership and verification improvements. The C layouts stay at ABI 4, while hash recipes and report content have the disclosed changes. Pair with codegen 1.4.0, regenerate and relink. See [CHANGELOG.md](CHANGELOG.md#140--2026-10-08).

‎docker/README.md‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -308,7 +308,7 @@ tested cases, and no completeness claim (the published 1.6.0 image declares
308308
From 1.6.2, the engine/codegen image's `run_json.py --so
309309
<strategy.so> --capabilities-json` probe declares revision `1.3`: with the
310310
section-4 producers and manifest present and value ownership `complete`
311-
(EN-115), it emits the four typed codes (`strategy_value_invalid`,
311+
(the completeness proof), it emits the four typed codes (`strategy_value_invalid`,
312312
`engine_command_value_invalid`, `report_result_invalid` and
313313
`report_post_execution_failed`). A build declaring `1.3` pairs only with a
314314
`complete` receipt; `1.3-parity` remains the parity build's literal, and a

‎docker/pf_bars.py‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1,4 +1,4 @@
1-
"""CR-150 pf_bar_t/v1 file contract and streaming CSV writer.
1+
"""pf_bar_t/v1 file contract and streaming CSV writer.
22
33
No header: each record is <5dq (OHLCV, open Unix ms), 48 bytes. A feed
44
also has <q closes. The canonical JSON sidecar is the commit record. Only

‎docker/pf_sink.py‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1,4 +1,4 @@
1-
"""Strict, streaming pf-report-sink version-1 reader (CR-150 DESIGN 3.7/4.4).
1+
"""Strict, streaming pf-report-sink version-1 reader.
22
33
Use ``with SinkReader(path) as reader:`` and consume the iterator to its end.
44
For a blocking binary file or pipe, use ``SinkReader(stream=file)`` instead;

‎docker/run_json.py‎

Lines changed: 2 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -5108,7 +5108,7 @@ def main(argv=None) -> int:
51085108
command line argparse refuses. The one exception is a success report stdout
51095109
cannot take whole (_main): exit 1 with the reason on stderr and no line after
51105110
the part written. The csv2bin subcommand instead returns 2 on refusal and
5111-
writes its one failure line to stderr (CR-150's converter contract)."""
5111+
writes its one failure line to stderr (the converter contract)."""
51125112
tokens = sys.argv[1:] if argv is None else argv
51135113
if tokens and tokens[0] == "csv2bin":
51145114
return _csv2bin_main(tokens[1:])
@@ -5651,7 +5651,7 @@ def _run(st, rep):
56515651
return 0
56525652

56535653

5654-
# --- CR-150 L9A/L9B: full and summary JSON from the fixed v1 report sink -------
5654+
# --- Full and summary JSON from the fixed v1 report sink -----------------------
56555655

56565656
class ReportSinkC(ctypes.Structure):
56575657
"""Exact pf_report_sink_v1; the setter copies all three strings."""

‎docker/run_json_bin_test.py‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1,4 +1,4 @@
1-
"""CR-150 L2 preflight discovery entry for tests/test_run_json_bin.py.
1+
"""Binary preflight discovery entry for tests/test_run_json_bin.py.
22
33
CTest names the implementation under tests/ so the private-fixture controls
44
copy the actual driver into their source mirror. Preflight keeps its Docker

‎docker/run_json_bounded_transport_test.py‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -67,7 +67,7 @@ def test_array_transport_is_the_frozen_bytes(tmp_path, frozen, count):
6767
report, trades, curve = population(count)
6868
incarnations = tuple(range(1, count + 1))
6969
reference = build(frozen, report, incarnations)
70-
# CR-202 / revision 3.1 (34f9b3a39f) added the two named forms after
70+
# Revision 3.1 (34f9b3a39f) added the two named forms after
7171
# this oracle froze. Every population here has a positive sub-day span
7272
# and finite cagr/calmar; preserve every other byte of the frozen report.
7373
assert 0 < curve[-1].time_ms - curve[0].time_ms < 86_400_000

‎docker/run_json_converter_metadata_test.py‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1,4 +1,4 @@
1-
"""S4E E-03: the replay converter retains report interpretation metadata."""
1+
"""The replay converter retains report interpretation metadata."""
22
import json
33

44
import pytest

‎docker/run_json_converter_session_test.py‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1,4 +1,4 @@
1-
"""S4E E-03: selected-window calendar nulls survive fd-sink conversion.
1+
"""Selected-window calendar nulls survive fd-sink conversion.
22
33
Only the native library and planner are doubles. run_json, the selected
44
report producer, fd wire writer and standalone converter remain real.

0 commit comments

Comments
 (0)