M12.4: isolate continued population replay after transfer
Foundation / checks (deploy) (push) Successful in 5s
Foundation / checks (ecs) (push) Successful in 2m23s
Foundation / postgres-tools (push) Canceled after 0s
Foundation / server-image (push) Canceled after 0s
Foundation / image (push) Canceled after 0s
Foundation / checks (client) (push) Canceled after 14m44s
Foundation / checks (checks) (push) Canceled after 14m44s
Foundation / checks (network) (push) Canceled after 14m43s
Foundation / checks (postgres) (push) Canceled after 14m42s
Foundation / checks (tsan) (push) Canceled after 14m43s
Foundation / checks (deploy) (push) Successful in 5s
Foundation / checks (ecs) (push) Successful in 2m23s
Foundation / postgres-tools (push) Canceled after 0s
Foundation / server-image (push) Canceled after 0s
Foundation / image (push) Canceled after 0s
Foundation / checks (client) (push) Canceled after 14m44s
Foundation / checks (checks) (push) Canceled after 14m44s
Foundation / checks (network) (push) Canceled after 14m43s
Foundation / checks (postgres) (push) Canceled after 14m42s
Foundation / checks (tsan) (push) Canceled after 14m43s
This commit is contained in:
@@ -1,5 +1,17 @@
|
||||
# Project status and next steps
|
||||
|
||||
## M12.4e continued population replay after transfer - 2026-10-17
|
||||
|
||||
Portable population batch 2 binds envelope 2/owner 7 and preserves boundary flows
|
||||
and nonresident history through biological intervals and manual conversions.
|
||||
Batch 1 retains its old meaning/bytes; cross-version replay is refused. A fresh
|
||||
process also reproduces a 64-tick destination interval after cohort transfer.
|
||||
Native Release passes 78/78; Clang analysis/format, ASan/UBSan and GCC checks pass.
|
||||
The private PostgreSQL suite proves both prepared C++ and SQL reject batch 2 under
|
||||
retained bundle 9, without moving the head. [Scope](engineering/M12_EXACT_BIOLOGICAL_TRANSFER.md)
|
||||
keeps durable version admission, paired escrow/activation and later M12 ecology
|
||||
gates open. No SQL, live flag, biological rule or deployment changed.
|
||||
|
||||
## M12.4d paired biological transfer replay - 2026-10-17
|
||||
|
||||
Canonical bounded migration trace 1 binds both predecessor cuts, both resulting
|
||||
|
||||
@@ -72,6 +72,12 @@ the trace's embedded request cannot supply that trust. Independent-process repla
|
||||
qualifies the raw and population representations without implying durable escrow.
|
||||
See [the trace schema](../../schemas/cellular-migration-trace-v1.md).
|
||||
|
||||
Continuing a migrated population uses separately identified population batch 2
|
||||
with the same bounded intention/member layout, bound to envelope 2/owner 7.
|
||||
Batch 1 continues to mean envelope 1/owner 6. Cross-version replay is refused;
|
||||
the portable addition does not admit batch 2 under durable bundle 9. Boundary
|
||||
accounting persists through normal biology and no-tick representation conversion.
|
||||
|
||||
This primitive changes ownership at an explicitly compatible cut and preserves
|
||||
coordinates. It does not choose a physical route or let current clamped movement
|
||||
cross an arena edge. Automatic linked-habitat movement, boundary contacts and
|
||||
|
||||
@@ -70,3 +70,16 @@ has no freeze, fence, durable retry outcome or activation rights. Independently
|
||||
establishing manifest/payload SHA-256 bindings and journal authority remains an
|
||||
I/O obligation; neither a checksum nor a UUID grants ownership. The durable
|
||||
escrow and linked-geometry limits above remain unchanged.
|
||||
|
||||
## M12.4e portable population continuation
|
||||
|
||||
Population batch 2 records ordinary biological intervals and whole-group no-tick
|
||||
conversions for envelope 2/owner 7. Batch 1 retains its old meaning and bytes;
|
||||
cross-version replay and mismatched magic/version are refused. Continued biology
|
||||
retains boundary flows/nonresident history while conserving actual current stock.
|
||||
The restart fixture now also records and independently replays a 64-tick interval
|
||||
after cohort transfer, comparing the complete continued destination snapshot.
|
||||
|
||||
This closes the portable continuation format boundary needed by a future durable
|
||||
adapter. It does not add a durable bundle or SQL admission: both the current C++
|
||||
prepared validator and SQL checker continue to admit only population batch 1.
|
||||
|
||||
@@ -50,6 +50,10 @@ history. They do not yet supply durable paired head publication or activation.
|
||||
Paired transfer trace 1 now supplies bounded, independently replayable input and
|
||||
member/material evidence, including raw/cohort separate-process reproduction.
|
||||
It does not grant ownership or close database escrow/retry/partition qualification.
|
||||
Population batch 2 continues owner-7 biology and conversions with retained flows;
|
||||
cross-version replay is refused and a new process reproduces a post-transfer
|
||||
64-tick interval. Its durable admission/upgrade remains intentionally unsupported
|
||||
by bundle 9 and schema 45.
|
||||
The transfer tests also expose the current rules-8 maturation gap: buds retain
|
||||
their subadult realization while reproduction requires a fully realized parent.
|
||||
Finite extra food alone does not close repeated wildlife generations. A versioned
|
||||
|
||||
@@ -1,5 +1,46 @@
|
||||
# Validation report
|
||||
|
||||
## 2026-10-17 - M12.4e versioned population continuation
|
||||
|
||||
Portable population batch 2 is bound to envelope 2/owner 7. Batch 1 keeps its
|
||||
original bytes and meaning. The existing durable prepared validator still admits
|
||||
only batch 1; no SQL, biological rule, dependency, live admission or deploy changed.
|
||||
|
||||
Executed:
|
||||
|
||||
- MSVC 19.44 Release configure/build: **78/78 CTests passed**, 13.64 s.
|
||||
- Clang 19 analysis build with clang-tidy: migration/checkpoint subset **3/3
|
||||
passed**, 18.71 s; final reverse-version refusal test **1/1 passed**, 5.31 s.
|
||||
Final repository clang-format 19 gate passes.
|
||||
- Clang 19 ASan/UBSan: the three-contract subset **3/3 passed**, 64.26 s,
|
||||
no reports, including a fresh process continuing a transferred cohort.
|
||||
- Actual GCC 14 server/integration-test build: portable subset **3/3 passed**,
|
||||
2.43 s. Final private `postgres-check.sh --population-only` **passed, exit 0,
|
||||
6 s**, including explicit C++ prepared retry/resolve and direct SQL rejection
|
||||
of batch 2 mislabeled as bundle 9. The SQL test requires the invalid-format
|
||||
category, so an unrelated duplicate-commit conflict cannot count as proof.
|
||||
The durable head remains identical after refusal.
|
||||
- Explicit clang-tidy 19 on the optional population store and its integration
|
||||
tests passes, with no owned warnings.
|
||||
|
||||
Tests replay ordinary 64-tick intervals for both sides after a real transfer,
|
||||
matching complete outputs while retaining each ledger and current stock. Manual
|
||||
whole-group conversion preserves exact biology. Wrong version/owner pairing and
|
||||
magic/version confusion are refused. Separate-process replay reproduces the full
|
||||
continued destination snapshot, including history, flows and cohort state.
|
||||
Existing batch-1 tests still pass under NDEBUG; no golden was replaced.
|
||||
|
||||
Limits: this qualifies portable continued replay and retained admission refusal.
|
||||
It does not add a durable bundle/upgrade, two-head transaction, transfer freeze,
|
||||
destination activation or migration-specific database fault proof. Those remain
|
||||
required before live admission. See [scope](M12_EXACT_BIOLOGICAL_TRANSFER.md).
|
||||
|
||||
The maintainer's `Build_Errors.txt` also records CI failures on the preceding
|
||||
78-test build: the instrumented cohort qualification exceeded 180 s, and the two
|
||||
new restart tests tried to create fixtures under a read-only `/source` mount.
|
||||
Those CI execution constraints are not covered by the local passes above and
|
||||
are being repaired as a separate validation segment; no CI pass is claimed.
|
||||
|
||||
## 2026-10-17 - M12.4d paired biological transfer trace and replay
|
||||
|
||||
ADR-0144's portable trace segment affects lifecycle replay/serialization, its
|
||||
|
||||
Reference in New Issue
Block a user