test(ecology): qualify unscripted maturing wildlife generations
Foundation / checks (ecs) (push) Successful in 2m22s
Foundation / checks (deploy) (push) Successful in 4m16s
Foundation / checks (network) (push) Successful in 20m53s
Foundation / checks (client) (push) Successful in 22m3s
Foundation / postgres-tools (push) Canceled after 0s
Foundation / server-image (push) Canceled after 0s
Foundation / image (push) Canceled after 0s
Foundation / checks (tsan) (push) Canceled after 22m4s
Foundation / checks (postgres) (push) Canceled after 24m24s
Foundation / checks (checks) (push) Canceled after 24m33s

This commit is contained in:
dingo
2026-10-19 12:35:18 -04:00
parent d3a867a662
commit 3a1b8c7923
15 changed files with 677 additions and 5 deletions
+5
View File
@@ -659,9 +659,14 @@ endif()
add_executable(evo-cellular-reference experiments/resolution/cellular_reference.cpp)
target_link_libraries(evo-cellular-reference PRIVATE evo::lifecycle)
evo_owned_target(evo-cellular-reference)
add_executable(evo-cellular-generations experiments/resolution/cellular_generations.cpp)
target_link_libraries(evo-cellular-generations PRIVATE evo::lifecycle)
evo_owned_target(evo-cellular-generations)
if(BUILD_TESTING)
add_test(NAME cellular-population-reference COMMAND evo-cellular-reference --verify)
set_tests_properties(cellular-population-reference PROPERTIES TIMEOUT 180)
add_test(NAME cellular-unscripted-generations COMMAND evo-cellular-generations --verify)
set_tests_properties(cellular-unscripted-generations PROPERTIES TIMEOUT 180)
endif()
add_executable(evo-cellular-cohorts experiments/resolution/cellular_cohorts.cpp)
+7
View File
@@ -196,6 +196,13 @@ checks rules-8 encoded replay and offline closure. `evo-cellular-reference --ver
runs the short CTest matrix; without arguments it runs the full research matrix
with local cost observations. This is an exact control, not production aggregation.
[Unscripted generations](docs/engineering/M12_UNSCRIPTED_GENERATIONS.md) uses real
funded player-origin offspring under the unchanged controller. Build
`evo-cellular-generations`; `--verify` runs the short CTest contract and `--qualify`
runs the declared 80,000-tick development/hold-out matrix with raw/cohort equality,
whole-genome conversions, restores and resource controls. This is bounded ecology
evidence, not actual-days or product acceptance.
M5.1 local voxel/address API and scope: [local voxel contract](docs/world/LOCAL_VOXEL_CONTRACT.md).
M5.8 process restart proof and measured costs: [world proof](docs/engineering/WORLD_PROOF.md).
@@ -12,7 +12,7 @@ spec:
env:
- {name: REGION_SESSION, value: "dev"}
- {name: REGION_SESSION_LISTEN, value: "0.0.0.0:7400"}
- {name: REGION_SESSION_WINDOW, value: "dev-20261019T155548-13dfba50:1792439748"}
- {name: REGION_SESSION_WINDOW, value: "dev-20261019T163247-193bdb60:1792441967"}
- {name: REGION_M10_MAX_GRANT_LIFETIME, value: "14400"}
ports:
# QUIC (ADR-0090): UDP only; nothing listens on TCP 7400.
+6
View File
@@ -647,6 +647,12 @@ the later multicellular idea are not ratified game rules.
## M12 - Persistent ecology
2026-10-19: [bounded unscripted generations](engineering/M12_UNSCRIPTED_GENERATIONS.md)
qualify four wildlife generations in the declared finite rich four-founder
matrix, including hold-outs, with exact raw/cohort biology, conservation,
predation/twins and restores. This supplies ecological evidence for M12; joint
environmental budgets, target-host costs, actual-days and product gates remain open.
2026-10-17: [12.2c rules-8 reference](engineering/M12_CELLULAR_REFERENCE.md)
validates 65,536 individual ticks and 70,656 encoded replay boundaries on MSVC/GCC,
with mutation, birth/death and offline-state coverage. This is the active control;
+17
View File
@@ -1,5 +1,22 @@
# Project status and next steps
## Bounded unscripted wildlife generations - 2026-10-19
[ADR-0162](adr/0162-bounded-unscripted-wildlife-generations.md) measures actual
funded player-origin wildlife under the unchanged controller. All players freeze
after the first bud. Four declared rich four-founder cases, including two
hold-outs, reach four wildlife generations with 97–102 births and 31–33 living
wildlife after 80,000 ticks. Complete raw/cohort biology and reports, conservation,
whole-genome conversions and periodic restores agree. Actual predation and twins
are observed; rules-8 and scarce controls retain their different outcomes.
Native Release passes **82/82**; the short ASan/UBSan contract, analysis and format
checks pass. [Results and reproduction](engineering/M12_UNSCRIPTED_GENERATIONS.md)
retain the weaker exploratory one-founder outcomes and exact scope. This is
4,000 simulated seconds per case, not actual days or sustainable producer ecology.
M12 remains in progress; environmental coupling, target-host/multi-host evidence,
global reconciliation and returning-player/product gates remain open.
## Durable migration of maturing wildlife - 2026-10-19
[ADR-0161](adr/0161-durable-maturing-wildlife-transfer.md)/schema 54 admits exact
@@ -0,0 +1,52 @@
# ADR-0162: Bounded unscripted wildlife generation qualification
## Status
Accepted for an offline M12 ecological qualification instrument, 2026-10-19.
Extends ADR-0125/0133/0139/0157. No biological, controller, persistence or default
deployment behavior changes. Thresholds below are experiment criteria, not game
rules or a product acceptance decision.
## Decision
Use the existing finite cellular ecology fixture, with real funded player
founders. Founders request budding every 100 ticks until the first successful
bud; thereafter every player is frozen. All wildlife decisions then come from
the unmodified production controller. No reserve grants, injected organisms,
manual wildlife actions, mutation selections or replenishment are permitted.
Generation is the actual retained parent chain, not genetic divergence.
Before measuring previously untested seeds 2 and 3, declare the following matrix:
rules 9, rich finite opening, four founders, 80,000 ticks, development seeds 0/1
and hold-outs 2/3. Each must retain living wildlife, reach at least generation 3,
and produce wildlife births in the final quarter. Every tick must conserve all
five channels and have identical complete canonical biology and action/growth
reports in raw and cohort execution. Whole-genome conversions alternate every
257 ticks; both representations restore every 4,096 ticks. Check that restores
retain bytes and controller intentions. Require actual conversions and restores.
Report all outcomes, including failure, and per-founder ancestry. Existing
80,000-tick exploratory seeds 0/1/7/19 with one and four founders are development
evidence, not unseen hold-outs. Include rules-8 and scarce-resource controls;
neither is required to sustain generations. Add a short NDEBUG-safe executable
contract exercising real birth, growth, conversion and recovery on both rules.
## Alternatives and consequences
Scripted parent/offspring actions establish funding correctness but not autonomous
reproduction. Rewriting the controller before measuring its longer trajectory
would conflate a new biological behavior with qualification. Retain both existing
contracts and measure the unchanged controller instead. The instrument costs a
separate bounded executable and full canonical comparisons; it supplies stronger
reproducible evidence without a runtime dependency or a production speed claim.
## Limits
80,000 ticks are 4,000 simulated seconds. This finite, resource-rich opening
does not establish perpetual sustainability, varied-world balance, useful
compression, target-host cost, environmental coupling or actual elapsed days.
Inherited genomes are unchanged; automatic mutation and evolving behavior remain
deferred under ADR-0133. A seed with two generations remains a measured outcome,
not grounds for silently changing the controller or weakening the declared
four-founder qualification. Automatic population tiers stay disabled. No live
world, database migration, extinction judgment or World0 operation is authorized.
+2
View File
@@ -1,5 +1,7 @@
# Architecture decision records
- [ADR-0162: Bounded unscripted wildlife generations](0162-bounded-unscripted-wildlife-generations.md) - a predeclared finite player-origin raw/cohort measurement with production controller, hold-outs and controls; no biological or deployment change.
- [ADR-0156: Census pending migration evidence](0156-census-pending-migration-evidence.md) - schema 52 captures active handoff pointers with registered heads and fully verifies retained capsules/ghosts; source owns stock and ghosts add none, including across settlement races.
- [ADR-0155: Database history domains](0155-database-history-domain.md) - schema 51 binds registered census evidence to a database namespace; supported quarantined restores retain parent domains and exact fork cuts before readmission. Cross-domain clone provenance and accepted extinction remain open.
+2 -2
View File
@@ -4,7 +4,7 @@ Requested scope: ecological identity, population qualification, durable automati
tier transitions, unattended scheduling/migration and environmental coupling.
Commit and push each validated scoped segment separately.
## Gate status (2026-10-18)
## Gate status (2026-10-19)
| Subset | State | Evidence or condition for completion |
|---|---|---|
@@ -15,7 +15,7 @@ Commit and push each validated scoped segment separately.
| 12.5 | Blocked on 12.2-12.4 | Finite M6 process fixtures do not qualify coupled persistent cellular/cohort atmosphere/water budgets. |
| 12.6 | Regional queries, replayed history, scoped multi-region queries and consistent database head snapshots implemented (12.6a-d) | ADR-0150–0153; short read-only snapshot capture prevents atomic pairs from straddling declared head sets, with exact immutable replay. Durable world/branch coverage, outside transfer/viability reconciliation, target-host query/index costs and client path remain required. |
| 12.7 | Evidence policy, registered coverage, history domains and verified pending handoff evidence implemented (12.7a-c) | ADR-0154–0156/schema 52 captures namespace/registry/heads/pending handoffs together, with exact capsule/ghost replay and supported restore forks. Cross-domain clone provenance, other viable-store coverage and durable accepted global reconciliation/freshness remain required. |
| 12.8 | Not qualified | Actual multi-day integrated ecology/restart/clone evidence remains required. |
| 12.8 | Bounded unscripted generation evidence; actual-days gate not qualified | ADR-0162 measures four real wildlife generations with exact raw/cohort recovery, predation and twins. Actual multi-day integrated ecology/restart/clone evidence remains required. |
| 12.9 | Not accepted | Returning-player product evidence and expansion review remain required; World0 stays separately blocked. |
The [remaining-gate audit](M12_REMAINING_GATES.md) identifies the exact publication
@@ -0,0 +1,119 @@
# Cellular/environment ownership resolution
2026-10-19 engineering design review for M12.5. This document proposes the next
implementation boundary; it does not declare a new biological rule, accepted
producer chemistry, working adapter or completed gate.
## Concrete conflict
`CellularLives` exclusively owns its finite habitat, nursery, living bodies and
reserves, corpses, heat, archive and boundary ledger. Its complete restore checks
five-channel closure against the opening and retained boundary flows. A habitat
configuration copied from environmental fields would create a second spendable
stock. `EnvironmentalCommunity` already owns its source, environment and exactly
two configured resource actors, with its own atomic history and checkpoint.
Advancing those two owners separately cannot publish one conserved biological tick.
The M6 ten-channel basis distinguishes water, dissolved N/P, organic energy and
C/N/P/other, incident energy and heat. Cellular `other` has no water or gas tag.
An elemental sum can therefore close while an invalid adapter turns water into
food or heat into usable energy. No current API authorizes that conversion.
These are implementation conflicts with ADR-0050/0055/0101/0144 and the
engineering constitution, not a request to reopen M11.9. Stop expanding the
affected integration until the following boundary has its own scoped contract.
## Proposed resolution
1. Introduce a distinct composite ecological owner and checkpoint profile. It
consumes exclusive ownership of cellular biology and tagged environmental
stocks, exposes read-only views, and prepares a complete detached tick. Keep
the existing M6 two-actor synthetic process unchanged as a reference fixture;
it is not the cellular producer implementation.
2. Initially support one explicitly declared axis-aligned habitat within one
environmental voxel, using the existing checked micrometre geometry. Persist
world/frame/region, source identity, origin, area and positive effective depth.
Reject overlapping authorizations, cross-voxel mappings and missing fields.
Camera dimensions and organism count never determine depth or supply.
3. Adopt a versioned compartment map. Only owned organic energy/C/N/P/other can
enter the existing edible deposit/reserve basis. Water remains separately
tagged even inside organisms; incident energy, heat and dissolved nutrients
cannot be credited as edible material through a generic retag operation.
An actual conversion requires an explicitly versioned process with all debit,
credit, work, waste and residual terms. Atmospheric species need their own
tagged elemental composition and storage contract; absent support is an error,
not zero, unlimited oxygen or an implicit photosynthesis recipe.
4. Give each source one persisted entitlement over a declared area and interval.
Capture and nutrient inflow need explicit finite source debit or an audited
external import, never both. Disjoint shares or one arbiter resolve contention.
Retain rate denominators, fractional residuals, interval watermarks and skipped
outcomes. Restart, retries and habitat refinement cannot reset entitlements.
5. Keep old cellular format 7/rules 8 and 9, population envelopes and durable
bundles 10/11 exact. Add separately identified composite format/rules/profile
only when the entire codec and replay contract exist. Do not attach new fields
to an old bundle or allow a bare nested cellular snapshot to publish a joint
result. Historical census and migration must either understand the whole joint
owner or refuse it explicitly.
6. Publish source debit, deposits, growth/maintenance, death/recycling, heat,
residuals and the complete nested biological result atomically at one fixed
quantum. No independently acknowledged environmental or cellular prefix.
Cross-worker sources require owned escrow and paired durable settlement before
activation; a timeout grants no release or second spending right.
## Delivery segments and acceptance evidence
Worked conservation examples below are integer accounting fixtures, not proposed
metabolic rates or a physical photosynthesis model:
| Step | Incident energy | Environmental organic energy | Cell reserve energy | Heat | Total pJ |
|---|---:|---:|---:|---:|---:|
| Opening | 100 | 30 | 10 | 0 | 140 |
| Funded capture of 20, retaining 15 | 80 | 45 | 10 | 5 | 140 |
| Cell maintenance spends 4 | 80 | 45 | 6 | 9 | 140 |
| Contact transfers 10 organic energy | 80 | 35 | 16 | 9 | 140 |
| Explicit external import of 7 | 87 | 35 | 16 | 9 | 147 |
The final row additionally retains exactly 7 pJ of cumulative external imports;
the closed rows retain zero. Neither heat nor source entitlement is an extra
spendable balance. Replaying the import must retain 147, not credit another 7.
Source-off executes no capture, but may spend already retained reserves.
For the `other` basis, water 100 fg + environmental organic other 20 fg + cell
organic other 6 fg totals 126 fg. Moving 2 fg of organic other into the cell gives
100 + 18 + 8 = 126. Moving water into the existing untagged food reserve must be
refused even though 98 + 20 + 8 also sums to 126. This is why elemental closure
alone cannot validate the adapter. Dissolved N/P incorporation and atmospheric
conversion likewise need actual typed reaction contracts, not a matching sum.
For geometry, 1,000 × 2,000 × 10 micrometres gives 20,000,000 cubic micrometres.
Doubling depth changes the declared sampling volume; it does not double any
owned water, nutrient or source stock. Source-area entitlement cannot double
merely because another cell overlaps the same illuminated area.
| Segment | Concrete deliverable | Required proof before commit |
|---|---|---|
| 12.5a | Accepted compartment/geometry/source and producer-process contracts, complete format declaration | Worked closed and open balances; water/heat/atmosphere refusals; no ambiguous ownership or unversioned coefficient |
| 12.5b | Portable composite owner and full codec from its first advancing operation | Real player-origin raw/cohort trajectories, atomic failed ticks, malformed/old formats, exact retry/restore/process continuation, closed-box and resource exhaustion comparisons |
| 12.5c | Dedicated durable profile, trace, explicit upgrade and complete recovery | Source/head/fence and representation races, uncertain replies, process kills, object failures, old bundle refusal and exact history |
| 12.5d | Worker/census/migration/client integration and source regimes | Registered complete-stock audits, no offline capture/growth, preserved lineage and escrow, useful source-off/open/nutrient-limited comparisons |
| 12.5e | Target-host and integrated-world qualification | Publication and contention tail costs, recovery faults, actual-days varied worlds and returning-player review |
The first segment must specify producer costs and atmosphere/water semantics
before implementation. The proposed infall/photosynthesis sections of ADR-0136
are design input, not authority to silently supply missing chemistry or free
matter. Trial coefficients must be clearly revisable and distinguish capture of
external energy from nutrient acquisition and body construction.
## Alternatives and consequences
A copied environmental view duplicates supply. Reusing M6's configured producer
as a cellular ancestor invents player-origin biology and treats synthetic growth
units as developed body mass. Independently committing two existing checkpoints
creates torn cuts. Adding a generic inventory setter bypasses source authority.
All four alternatives are rejected as implementation shortcuts.
A dedicated owner costs another complete format, replay path, durable admission
and integration matrix. It avoids changing historical biological bytes and makes
unsupported physical states explicit. Existing maturation, migration, census and
bounded generation qualification can proceed independently. No environmental
runtime, live upgrade, World0 operation or M12.5 completion is claimed here.
+14 -2
View File
@@ -23,13 +23,13 @@ or authorize production enablement.
| Gate | Existing useful basis | Work/evidence still required |
|---|---|---|
| 12.2 | Exact cellular model and paired matrix; local samples through 256 lives | Deployment-host tail costs including publication; larger wildlife cohorts, predation/twins and contention coverage; useful compression remains unimplemented. |
| 12.2 | Exact cellular model and paired matrix; local samples through 256 lives; bounded four-generation rules-9 predation/twin matrix | Deployment-host tail costs including publication; larger wildlife cohorts and fuller contention coverage; useful compression remains unimplemented. |
| 12.3 | Dedicated population profile/trace/store/owning host; exact recovery and private PostgreSQL fault tests | Population-specific executable/network fault drills and target-host publication/capacity qualification. |
| 12.4 | Owned intervals; exact transfer/accounting; durable atomic pair/recovery; independent loopback TLS path loss and owner expiry/replacement/heal; local object sync/read-failure matrix | Multi-host/kernel-network, storage-host and orchestrator drills, linked-boundary contracts and integrated ownership/stock audits at scale. |
| 12.5 | Finite cellular contact feeding/decay and separate M6 producer/consumer resource owner | One jointly owned cellular/environment budget, explicit effective depth and source entitlements, versioned water/atmosphere species mapping and producer coefficients, plus closed/open and nutrient-exhaustion comparisons. |
| 12.6 | Retained archive, regional census/lineage queries, replayed history, scoped multi-region queries and consistent database head snapshots (12.6a-d); registered coverage/history domains follow in 12.7a-b | Cross-domain clone provenance, retained historical coverage/other viable stores, target-host query/index qualification and client permission/presentation; no account power or invented precise statistical ancestry. |
| 12.7 | Viability/unknown policy; consistent cuts, registered coverage, history domains/restore forks and verified pending handoff evidence (12.7a-c/schema 52) | Cross-domain clone provenance, other viable-store coverage and durable accepted global evidence records/freshness. Missing/stale/unsupported owners must prevent terminal extinction. |
| 12.8 | Deterministic seed/regime drivers and separate-process exact continuation | Reproducible integrated worlds running for actual days, real restarts/clone comparisons and retained accounting/diversity/cost audits. Accelerated tests cannot supply elapsed days. |
| 12.8 | Deterministic seed/regime drivers, separate-process exact continuation and bounded unscripted generations | Reproducible integrated worlds running for actual days, real restarts/clone comparisons and retained accounting/diversity/cost audits. Accelerated tests cannot supply elapsed days. |
| 12.9 | M11 feedback and current UI review retained | Returning-player samples on the integrated ecology and explicit product/3D readiness decision after soak evidence. No M12 product acceptance or World0 promotion is inferred. |
Do not couple the current cellular habitat to a copied M6 stock view: two owners
@@ -39,6 +39,11 @@ finite deposits. Their geometry, ownership and chemical units need a versioned
mapping, not a renamed fixture. ADR-0136 step A is implemented; proposed infall
and photosynthesis steps B/C do not silently become accepted M12 mechanics.
The [ownership resolution proposal](M12_ENVIRONMENT_OWNERSHIP_RESOLUTION.md)
records the actual public-API conflict, a composite-owner/checkpoint direction,
typed compartment and source requirements, worked balances and delivery segments.
It is a concrete design for review, not an implemented adapter or completed 12.5.
The engineering constitution requires stopping expansion at this architectural
boundary and documenting the resolution. That does not prevent the independent
worker scheduling correction or the dedicated exact population adapter above.
@@ -192,6 +197,13 @@ add none. Prepare/stage/abort/commit races preserve the selected view. Unavailab
or false evidence fails the whole query. Other viable stores, external clone
provenance and accepted durable global reconciliation/freshness remain open.
ADR-0162's [bounded unscripted generations](M12_UNSCRIPTED_GENERATIONS.md)
qualify unchanged-controller rules-9 reproduction through four wildlife generations
in the declared finite rich four-founder matrix, with raw/cohort equality,
conservation, actual predation/twins and periodic exact restores. Scarce and rules-8
controls retain their different outcomes. These are 4,000 simulated seconds per
case, not sustainable producer ecology or actual elapsed days.
Current expansion readiness: **no-go for M13 integration or World0 promotion**
while these M12 integration/evidence gates remain open. This records the present
engineering evidence, not a substitute for the later maintainer product review.
@@ -0,0 +1,109 @@
# Bounded unscripted wildlife generations
Authority: [ADR-0162](../adr/0162-bounded-unscripted-wildlife-generations.md).
This offline instrument measures real funded player-origin organisms under the
existing controller and rules 9. No controller, biology, persistence, default
opening or live region behavior changes.
## Protocol
The shared finite ecology fixture supplies the entire opening. Player founders
request a bud every 100 ticks until a real funded bud succeeds. All players then
freeze for the rest of the run; their biology is checked unchanged. Wildlife uses
only the production controller. There are no scripted wildlife actions, mutation
selections, injected adults, reserve grants or replenishment. Ancestry follows
actual retained parent IDs, including actual twins; it is not genotype diversity.
The predeclared matrix uses four founders, rich finite resources and 80,000 ticks
on development seeds 0/1 and previously unmeasured hold-outs 2/3. Each case must
reach generation 3, retain living wildlife and produce births in its last quarter.
Raw and cohort execution compare complete canonical biology and bud, mutation and
growth reports every tick, with zero error permitted. Every tick checks complete
energy and C/N/P/other closure in both representations. Whole-genome conversion
alternates every 257 ticks. Both owners restore every 4,096 ticks, preserving
canonical bytes and the controller's next intentions. Controlled cells never enter
cohorts. The instrument reports conversion counts, member-ticks and ancestry by
founder. Rules-8 and resource-scarce controls run separately.
## Exploratory baseline
Before declaring the hold-outs, 80,000-tick raw rules-9 runs measured:
| Seed | Founders | Wildlife births | Deepest generation | Peak wildlife | Final wildlife |
|---|---:|---:|---:|---:|---:|
| 0 | 1 | 12 | 2 | 7 | 5 |
| 1 | 1 | 21 | 3 | 9 | 8 |
| 7 | 1 | 30 | 4 | 10 | 9 |
| 19 | 1 | 28 | 3 | 11 | 8 |
| 0 | 4 | 102 | 4 | 37 | 32 |
| 1 | 4 | 97 | 4 | 39 | 31 |
| 7 | 4 | 94 | 4 | 35 | 28 |
| 19 | 4 | 124 | 4 | 39 | 36 |
The first funded bud was tick 6,100 in every case. A rules-8 seed-7/one-founder
control had one birth, generation 1 and zero growth. The weaker single-founder
seed 0 is retained; no general all-opening or minimum-survival claim follows.
## Declared matrix results
The complete GCC 14.2 and MSVC 19.44 Release matrices pass with exact raw/cohort
biology, complete step/bud/mutation/growth reports and closure on every tick.
Their complete JSONL outputs match after newline normalization. Retained outputs:
[GCC](results/m12-unscripted-generations-gcc14.jsonl) and
[MSVC](results/m12-unscripted-generations-msvc1944.jsonl).
| Seed | Role | Births | Generation | Final wildlife | Final-quarter births | Engulfments | Twin events | Funded growth steps |
|---|---|---:|---:|---:|---:|---:|---:|---:|
| 0 | Development | 102 | 4 | 32 | 42 | 70 | 26 | 17,860 |
| 1 | Development | 97 | 4 | 31 | 37 | 66 | 20 | 16,796 |
| 2 | Hold-out | 102 | 4 | 33 | 35 | 69 | 19 | 17,718 |
| 3 | Hold-out | 100 | 4 | 33 | 37 | 67 | 17 | 19,468 |
Every founding ancestry produces wildlife: total births by root 100/101/102/103
are 12/27/36/27, 25/12/26/34, 20/30/27/25 and 23/25/29/23 respectively.
The rich runs contain no starvation deaths; their 272 deaths are actual
engulfments. The 82 twin events count actual same-parent newborn pairs, including
any founder twins. These measurements qualify those interactions in this matrix;
they do not establish balance or all contention states. No failed spatial bud
placement was observed, so the run does not qualify that refusal path.
Including controls, each full matrix covers **480,000 paired ticks**, 23,292 member
conversions, 2,993,603 cohort member-ticks and 114 paired restore checks (228 owner
restores). The rules-8 control retains one generation and zero growth. The scarce
control produces no buds or wildlife; its four active founders starve through the
normal biology. This is not a terminal-extinction acceptance record.
Native MSVC 19.44 Release CTest passes **82/82, 19.29 s**, including the short
12,000-tick rules-8/rules-9 contract. The same short contract passes Clang 19.1
ASan/UBSan. clang-tidy 19 and repository clang-format 19 pass. The first standalone
tidy invocation lacked the new target's compiler module map; building that analysis
target generated it and the final analysis passed. No production implementation
changed, so the already-qualified private persistence matrix was not repeated.
## Reproduction
After configuring the repository's Release preset and compiler environment:
```text
cmake --build --preset release --target evo-cellular-generations
ctest --preset release -R cellular-unscripted-generations
build/release/evo-cellular-generations --qualify
build/release/evo-cellular-generations --seed 7 --founders 1 --ticks 80000 --rules 9 --rich 1 --paired 1
```
On Windows the executable has the `.exe` suffix. `--verify` is the short CTest
contract: rules 8 and 9 each run 12,000 ticks with real funded birth, alternating
conversions and two restores. Rules 9 must grow; rules 8 must not. Failure is a
nonzero exit under NDEBUG as well as debug. Options and work are strictly bounded.
`--qualify` runs all four declared cases and both controls, retaining all case
results before reporting a failed ecological threshold.
## Limits
80,000 ticks equal 4,000 simulated seconds, not elapsed days. Rich finite contact
food does not establish sustainable producers, a complete nutrient cycle or a
joint cellular/environment owner. The inherited genome and production controller
are unchanged. This does not qualify automatic mutation, useful compression,
target-host costs, multi-host recovery, actual-days soak, terminal extinction,
returning-player enjoyment or M12 completion. No live/default upgrade or World0
operation follows from this instrument.
+51
View File
@@ -1,5 +1,56 @@
# Validation report
## 2026-10-19 - Bounded unscripted wildlife generations
[ADR-0162](../adr/0162-bounded-unscripted-wildlife-generations.md) adds an offline
measurement executable and bounded CTest contract. Real funded player founders
bud; all players then freeze and unchanged production wildlife intentions drive
the rest. No biological, controller, persistence, default-opening or live-world
change. Full protocol, outputs, controls and limits are in
[the report](M12_UNSCRIPTED_GENERATIONS.md).
Executed locally:
- MSVC 19.44 Release full CTest **82/82 passed, 19.29 s**, with NDEBUG. The new
contract runs 12,000 ticks each on rules 8 and 9 with actual birth, growth,
conversions and periodic recovery.
- GCC 14.2 and MSVC 19.44 Release complete declared matrices passed **exit 0**
with identical full JSONL after newline normalization: six 80,000-tick
raw/cohort cases, including development seeds 0/1, hold-outs 2/3, rules-8 and
scarce controls. All rich cases reach generation 4, retain 31–33 wildlife and
produce 35–42 births in their final quarter. Their actual deaths include 272
engulfments, and actual same-parent births include 82 twin events.
- All **480,000 paired ticks** compare complete canonical biology and full step,
bud, mutation and growth reports with zero error. Both owners conserve every
channel every tick; 23,292 conversions and 114 paired restore checks retain
exact biology, presence, history and next controller intentions. Player biology
remains unchanged after freezing. The final frozen GCC executable hash verifies;
both full runs used separately frozen executables.
- Clang 19.1 ASan/UBSan passes the same short rules-8/rules-9 contract, without
sanitizer reports. clang-tidy 19 and repository clang-format 19 pass. The first
direct tidy invocation lacked the new target's compiler module map; building
the target generated it, then final analysis passed.
The weaker exploratory one-founder seed-0 outcome (generation 2) remains in the
report. No failed bud placement occurred in the declared rich matrix; that case
is not claimed covered. The scarce control has zero wildlife and four normal
founder starvation deaths, not accepted global extinction. The rich finite source
fixture demonstrates bounded reproduction, not self-sustaining producers or
actual elapsed days. Genomes and controller remain unchanged. No production
persistence code changed, so its previously passing full private matrix was not
repeated for this instrument.
The same scope records the M12.5 environmental public-API/ownership conflict and
[a concrete resolution proposal](M12_ENVIRONMENT_OWNERSHIP_RESOLUTION.md), with
closed/open and water/food accounting examples. This is design review, not an
accepted chemistry implementation or a completed environmental gate.
The DEV listener window is refreshed in the same commit to an absolute four-hour
deadline, **2026-10-19 20:32:47 UTC**. No live cluster action or World0 operation.
Offline rendered deployment policy passes: 75 objects in 74 files, 138 rendered
objects in 17 overlays, zero failures. Policy code is unchanged; the existing
46-case policy self-test result was not rerun for this value-only refresh.
## 2026-10-19 - Durable migration of maturing wildlife
[ADR-0161](../adr/0161-durable-maturing-wildlife-transfer.md)/schema 54 admits
@@ -0,0 +1,23 @@
{"kind":"generations","seed":0,"rules":9,"founders":4,"ticks":80000,"first_bud":6100,"births":102,"deepest":4,"peak":37,"living":32,"adults":13,"growth":17860,"blocked_buds":0,"engulfed":70,"starved":0,"twin_events":26,"rich":true,"paired":true,"late_births":42,"conversions":5627,"cohort_member_ticks":723294,"restores":19}
{"kind":"ancestry","root":100,"births":12}
{"kind":"ancestry","root":101,"births":27}
{"kind":"ancestry","root":102,"births":36}
{"kind":"ancestry","root":103,"births":27}
{"kind":"generations","seed":1,"rules":9,"founders":4,"ticks":80000,"first_bud":6100,"births":97,"deepest":4,"peak":39,"living":31,"adults":12,"growth":16796,"blocked_buds":0,"engulfed":66,"starved":0,"twin_events":20,"rich":true,"paired":true,"late_births":37,"conversions":5528,"cohort_member_ticks":709956,"restores":19}
{"kind":"ancestry","root":100,"births":25}
{"kind":"ancestry","root":101,"births":12}
{"kind":"ancestry","root":102,"births":26}
{"kind":"ancestry","root":103,"births":34}
{"kind":"generations","seed":2,"rules":9,"founders":4,"ticks":80000,"first_bud":6100,"births":102,"deepest":4,"peak":39,"living":33,"adults":11,"growth":17718,"blocked_buds":0,"engulfed":69,"starved":0,"twin_events":19,"rich":true,"paired":true,"late_births":35,"conversions":5873,"cohort_member_ticks":754582,"restores":19}
{"kind":"ancestry","root":100,"births":20}
{"kind":"ancestry","root":101,"births":30}
{"kind":"ancestry","root":102,"births":27}
{"kind":"ancestry","root":103,"births":25}
{"kind":"generations","seed":3,"rules":9,"founders":4,"ticks":80000,"first_bud":6100,"births":100,"deepest":4,"peak":39,"living":33,"adults":14,"growth":19468,"blocked_buds":0,"engulfed":67,"starved":0,"twin_events":17,"rich":true,"paired":true,"late_births":37,"conversions":5976,"cohort_member_ticks":768763,"restores":19}
{"kind":"ancestry","root":100,"births":23}
{"kind":"ancestry","root":101,"births":25}
{"kind":"ancestry","root":102,"births":29}
{"kind":"ancestry","root":103,"births":23}
{"kind":"generations","seed":7,"rules":8,"founders":1,"ticks":80000,"first_bud":6100,"births":1,"deepest":1,"peak":1,"living":1,"adults":0,"growth":0,"blocked_buds":0,"engulfed":0,"starved":0,"twin_events":0,"rich":true,"paired":true,"late_births":0,"conversions":288,"cohort_member_ticks":37008,"restores":19}
{"kind":"ancestry","root":100,"births":1}
{"kind":"generations","seed":0,"rules":9,"founders":4,"ticks":80000,"first_bud":0,"births":0,"deepest":0,"peak":0,"living":0,"adults":0,"growth":0,"blocked_buds":0,"engulfed":0,"starved":4,"twin_events":0,"rich":false,"paired":true,"late_births":0,"conversions":0,"cohort_member_ticks":0,"restores":19}
@@ -0,0 +1,23 @@
{"kind":"generations","seed":0,"rules":9,"founders":4,"ticks":80000,"first_bud":6100,"births":102,"deepest":4,"peak":37,"living":32,"adults":13,"growth":17860,"blocked_buds":0,"engulfed":70,"starved":0,"twin_events":26,"rich":true,"paired":true,"late_births":42,"conversions":5627,"cohort_member_ticks":723294,"restores":19}
{"kind":"ancestry","root":100,"births":12}
{"kind":"ancestry","root":101,"births":27}
{"kind":"ancestry","root":102,"births":36}
{"kind":"ancestry","root":103,"births":27}
{"kind":"generations","seed":1,"rules":9,"founders":4,"ticks":80000,"first_bud":6100,"births":97,"deepest":4,"peak":39,"living":31,"adults":12,"growth":16796,"blocked_buds":0,"engulfed":66,"starved":0,"twin_events":20,"rich":true,"paired":true,"late_births":37,"conversions":5528,"cohort_member_ticks":709956,"restores":19}
{"kind":"ancestry","root":100,"births":25}
{"kind":"ancestry","root":101,"births":12}
{"kind":"ancestry","root":102,"births":26}
{"kind":"ancestry","root":103,"births":34}
{"kind":"generations","seed":2,"rules":9,"founders":4,"ticks":80000,"first_bud":6100,"births":102,"deepest":4,"peak":39,"living":33,"adults":11,"growth":17718,"blocked_buds":0,"engulfed":69,"starved":0,"twin_events":19,"rich":true,"paired":true,"late_births":35,"conversions":5873,"cohort_member_ticks":754582,"restores":19}
{"kind":"ancestry","root":100,"births":20}
{"kind":"ancestry","root":101,"births":30}
{"kind":"ancestry","root":102,"births":27}
{"kind":"ancestry","root":103,"births":25}
{"kind":"generations","seed":3,"rules":9,"founders":4,"ticks":80000,"first_bud":6100,"births":100,"deepest":4,"peak":39,"living":33,"adults":14,"growth":19468,"blocked_buds":0,"engulfed":67,"starved":0,"twin_events":17,"rich":true,"paired":true,"late_births":37,"conversions":5976,"cohort_member_ticks":768763,"restores":19}
{"kind":"ancestry","root":100,"births":23}
{"kind":"ancestry","root":101,"births":25}
{"kind":"ancestry","root":102,"births":29}
{"kind":"ancestry","root":103,"births":23}
{"kind":"generations","seed":7,"rules":8,"founders":1,"ticks":80000,"first_bud":6100,"births":1,"deepest":1,"peak":1,"living":1,"adults":0,"growth":0,"blocked_buds":0,"engulfed":0,"starved":0,"twin_events":0,"rich":true,"paired":true,"late_births":0,"conversions":288,"cohort_member_ticks":37008,"restores":19}
{"kind":"ancestry","root":100,"births":1}
{"kind":"generations","seed":0,"rules":9,"founders":4,"ticks":80000,"first_bud":0,"births":0,"deepest":0,"peak":0,"living":0,"adults":0,"growth":0,"blocked_buds":0,"engulfed":0,"starved":4,"twin_events":0,"rich":false,"paired":true,"late_births":0,"conversions":0,"cohort_member_ticks":0,"restores":19}
@@ -0,0 +1,246 @@
// ADR-0162: bounded real player-origin generations, not a game rule.
#include "../../tests/fixtures/cellular_ecology.hpp"
#include "evo/server/lifecycle/cellular_population.hpp"
#include <charconv>
#include <map>
namespace {
using namespace evo;
using namespace evo::server;
using namespace evo::fixture;
std::uint64_t number(std::string_view text) {
std::uint64_t value{};
const auto parsed = std::from_chars(text.data(), text.data() + text.size(), value);
require(parsed.ec == std::errc{} && parsed.ptr == text.data() + text.size(), "invalid number");
return value;
}
struct Result {
std::uint64_t births{}, late_births{}, living{}, growth{}, conversions{}, restores{};
unsigned deepest{};
};
Result run(std::uint64_t seed, std::uint64_t ticks, std::uint32_t founders, bool maturing,
bool rich, bool paired) {
auto state = opening(seed, rich, founders);
if (maturing) {
state = checked(state->migration_upgrade_candidate(), "generation boundary upgrade");
state = checked(state->maturation_upgrade_candidate(), "generation maturation upgrade");
}
auto population =
paired ? checked(CellularPopulation::create(restore(*state)), "generation population")
: nullptr;
Result result;
std::map<std::uint64_t, unsigned> generations;
std::map<std::uint64_t, std::uint64_t> roots, root_births;
for (const auto& life : state->lives()) {
generations.emplace(life.identity.life, 0);
roots.emplace(life.identity.life, life.identity.life);
}
std::uint64_t first_bud = 0, blocked_buds = 0, peak = 0, cohort_ticks = 0;
std::uint64_t engulfed = 0, starved = 0, twin_events = 0;
std::vector<CellularOwnedLife> frozen_players;
// Once the first real funded bud succeeds, all controlled cells stay frozen.
// No wildlife intention, mutation, birth or resource is injected by this driver.
for (std::uint64_t index = 0; index < ticks; ++index) {
if (population && index % 257 == 0) {
std::set<std::uint64_t> genomes;
for (const auto& life : state->lives())
if (life.account == LifeAccount{})
genomes.insert(life.identity.genome);
for (const auto genome : genomes) {
if ((index / 257) % 2 == 0)
result.conversions +=
checked(population->collapse(genome), "generation collapse");
else if (std::ranges::any_of(population->members(), [genome](const auto& member) {
return member.genome == genome;
}))
result.conversions +=
checked(population->materialize(genome), "generation materialize");
}
require(population->state().snapshot() == state->snapshot(),
"generation conversion changed biology");
}
auto intents = checked(wildlife_intents(*state, state->tick() + 1), "generation wildlife");
std::vector<CellularLifeIntent> players;
for (const auto& life : state->lives()) {
if (life.account == LifeAccount{})
continue;
CellularLifeIntent controlled{life.identity.life, {}};
controlled.intent.frozen = first_bud != 0;
controlled.intent.bud = first_bud == 0 && (state->tick() + 1) % 100 == 0;
intents.push_back(controlled);
players.push_back(controlled);
if (population)
require(!population->in_cohort(life.identity.life), "player entered cohort");
}
std::ranges::sort(intents, {}, &CellularLifeIntent::life_id);
const auto executed =
checked(state->advance_tick_with_mutations(intents, {}), "generation biology");
if (population) {
const auto candidate =
checked(population->advance_tick(players), "generation paired tick");
require(candidate.step == executed.step && candidate.buds == executed.buds &&
candidate.mutations == executed.mutations &&
candidate.growth == executed.growth,
"generation paired events diverged");
require(population->state().snapshot() == state->snapshot(),
"generation paired biology diverged");
require(population->state().current_total() == population->state().opening_total(),
"generation paired stock did not close");
cohort_ticks += population->members().size();
}
result.growth += executed.growth.size();
for (const auto& bud : executed.buds)
blocked_buds += bud.result == CellularBudResult::no_placement;
for (const auto& corpse : executed.step.starved) {
engulfed += corpse.cause == biology::CellularDeathCause::engulfed;
starved += corpse.cause == biology::CellularDeathCause::starved;
}
std::map<std::uint64_t, unsigned> parent_births;
std::uint64_t living = 0;
for (const auto& life : state->lives()) {
if (life.account != LifeAccount{})
continue;
++living;
if (generations.contains(life.identity.life))
continue;
const auto parent = generations.find(life.parent_life);
require(parent != generations.end(), "wildlife lacks retained player-origin parent");
const auto generation = parent->second + 1;
generations.emplace(life.identity.life, generation);
const auto root = roots.at(life.parent_life);
roots.emplace(life.identity.life, root);
++root_births[root];
result.deepest = std::max(result.deepest, generation);
++result.births;
++parent_births[life.parent_life];
result.late_births += index >= ticks * 3 / 4;
if (first_bud == 0)
first_bud = state->tick();
}
for (const auto& [parent, count] : parent_births) {
static_cast<void>(parent);
twin_events += count == 2;
}
peak = std::max(peak, living);
require(state->current_total() == state->opening_total(), "generation stock did not close");
// Presence may change during disconnection. Frozen biology must not.
for (const auto& before : frozen_players) {
const auto after =
std::ranges::find(state->lives(), before.identity.life,
[](const auto& life) { return life.identity.life; });
require(after != state->lives().end() && after->identity == before.identity &&
after->growth == before.growth && after->position == before.position &&
after->body == before.body && after->reserve == before.reserve &&
after->mutation == before.mutation,
"frozen player biology changed");
}
if (first_bud != 0 && frozen_players.empty())
for (const auto& life : state->lives())
if (life.account != LifeAccount{})
frozen_players.push_back(life);
if (index % 4096 == 4095) {
const auto bytes = state->snapshot();
const auto input =
checked(wildlife_intents(*state, state->tick() + 1), "before restore");
state = restore(*state);
require(state->snapshot() == bytes &&
checked(wildlife_intents(*state, state->tick() + 1), "after restore") ==
input,
"generation raw restore changed bytes or controller");
if (population) {
const auto envelope = population->snapshot();
population =
checked(CellularPopulation::restore(envelope, state->world(), state->region(),
state->frame(), state->tick()),
"generation population restore");
require(population->snapshot() == envelope &&
population->state().snapshot() == bytes &&
checked(wildlife_intents(population->state(), state->tick() + 1),
"population restored controller") == input,
"generation population restore diverged");
}
++result.restores;
}
}
std::uint64_t adults = 0;
for (const auto& life : state->lives())
if (life.account == LifeAccount{}) {
++result.living;
adults += life.growth.realized_per_thousand == 1000;
}
std::cout << "{\"kind\":\"generations\",\"seed\":" << seed
<< ",\"rules\":" << state->rules_version() << ",\"founders\":" << founders
<< ",\"ticks\":" << ticks << ",\"first_bud\":" << first_bud
<< ",\"births\":" << result.births << ",\"deepest\":" << result.deepest
<< ",\"peak\":" << peak << ",\"living\":" << result.living << ",\"adults\":" << adults
<< ",\"growth\":" << result.growth << ",\"blocked_buds\":" << blocked_buds
<< ",\"engulfed\":" << engulfed << ",\"starved\":" << starved
<< ",\"twin_events\":" << twin_events << ",\"rich\":" << (rich ? "true" : "false")
<< ",\"paired\":" << (paired ? "true" : "false")
<< ",\"late_births\":" << result.late_births
<< ",\"conversions\":" << result.conversions
<< ",\"cohort_member_ticks\":" << cohort_ticks << ",\"restores\":" << result.restores
<< "}\n";
for (const auto& [root, count] : root_births)
std::cout << "{\"kind\":\"ancestry\",\"root\":" << root << ",\"births\":" << count << "}\n";
std::cout.flush();
return result;
}
} // namespace
int main(int argc, char** argv) try {
if (argc == 2 && std::string_view{argv[1]} == "--qualify") {
bool passed = true;
for (std::uint64_t seed = 0; seed < 4; ++seed) {
const auto result = run(seed, 80'000, 4, true, true, true);
passed = passed && result.deepest >= 3 && result.living > 0 && result.late_births > 0 &&
result.conversions > 0 && result.restores > 0;
}
const auto legacy = run(7, 80'000, 1, false, true, true);
require(legacy.births > 0 && legacy.deepest == 1 && legacy.growth == 0,
"legacy generation control changed");
const auto scarce = run(0, 80'000, 4, true, false, true);
require(scarce.births == 0 && scarce.living == 0,
"scarce opening unexpectedly funded wildlife");
require(passed, "declared rich generation qualification failed; retain all results");
return 0;
}
if (argc == 2 && std::string_view{argv[1]} == "--verify") {
for (const bool maturing : {false, true}) {
const auto result = run(7, 12'000, 1, maturing, true, true);
require(result.births > 0 && result.conversions > 0 && result.restores > 0,
"generation contract missed funded births/conversions/restores");
require(maturing ? result.growth > 0 : (result.growth == 0 && result.deepest == 1),
"generation rule control failed");
}
return 0;
}
std::uint64_t seed = 7, ticks = 40'000, founders = 1, rules = 9;
bool rich = true, paired = false;
require(argc % 2 == 1, "expected option/value pairs");
for (int index = 1; index < argc; index += 2) {
const std::string_view option{argv[index]};
const auto value = number(argv[index + 1]);
if (option == "--seed")
seed = value;
else if (option == "--ticks")
ticks = value;
else if (option == "--founders")
founders = value;
else if (option == "--rules")
rules = value;
else if (option == "--rich" || option == "--paired") {
require(value <= 1, "expected boolean zero or one");
(option == "--rich" ? rich : paired) = value != 0;
} else
throw std::runtime_error("unknown generation option");
}
require(seed <= 1'000'000 && ticks > 0 && ticks <= 1'000'000 && founders > 0 &&
founders <= 16 && (rules == 8 || rules == 9),
"generation work bound");
run(seed, ticks, static_cast<std::uint32_t>(founders), rules == 9, rich, paired);
return 0;
} catch (const std::exception& error) {
std::cerr << error.what() << '\n';
return 1;
}