Keep the filesystem object store on a node-local Retain-class PVC per
environment (ADR-0070), with the database on the external PostgreSQL VM.
Because local-path does not enforce sizes, the object store now refuses
new objects below a free-space reserve with a distinct full error, so a
full volume publishes and acknowledges nothing. evo-world-clone gains
check-store to qualify volume retention, reserve, database version/TLS
and the runtime privilege boundary from inside the cluster; it now ships
in the server image. Add the storage class, object claims, hand-applied
store Jobs, policy rules forbidding password env vars, and the runbook.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The maintainer applied the platform layer, World0 admission policy and
projects; verify-isolation.sh passed every guardrail, RBAC-negative,
admission dry-run and cleanup check. Its three project failures were a
script defect (destinations and the empty cluster whitelist were printed
together as "[]"); check them separately. Ignore local result files, add
docs/BACKLOG.md with break-glass membership and the storage re-probe, and
record leaving Argo's default project unrestricted as an accepted risk.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Per-environment platform layers (dev, stage, world0) with restricted pod
security, quotas, limit ranges, default-deny networking and token-less
service accounts, applied by an administrator rather than Argo. A native
ValidatingAdmissionPolicy denies deleting World0 data, identity and Argo
objects to every requester without the break-glass group. Stage and
World0 Argo projects are scoped to one namespace, World0 blocks
automated sync, and a cleanup tool refuses World0. An offline policy
check runs as a new CI suite; a live verification script uses only
can-i, reads and server dry runs.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Record the maintainer-run local-path probe on all three workers
(1.9-2.5 ms synchronous 4 KiB writes, 29-36 ms mean object publication,
110-255 MB/s sequential), well inside the lease deadlines, and flag
worker-3's 87% full root filesystem. Apply the maintainer's corrections:
about 500 GB more host RAM, AEvolution PostgreSQL on a separate VM
outside Kubernetes, and manual-only inventory collection. Storage
figures stay provisional until the planned RAID change.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Summarize the maintainer-collected read-only inventory: RKE2 v1.32.5,
an untainted control plane and three 32 GB workers on one hypervisor,
no kubelet reservations or quotas, local-path as the only (node-pinned,
reclaim Delete) storage, Canal/ingress-nginx/MetalLB/cert-manager, a
manually synced aevolution-dev application and no PostgreSQL 18.6.
Propose a measured budget and dev placement, and add a ready-to-apply
local-path probe with one PVC and Job pinned to each worker.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Record the maintainer-stated host facts (R730xd ~128 GB, 32 GB workers,
automatic Harbor publish and Argo CD deploy, existing aevolution-db
Secret and its conflicts with the pinned version and role split), the
measured evo-server profile (peak RSS ~9.7 MiB, ~0.3% of a core) and a
local storage baseline. Add a read-only cluster collector that never
fetches Secret data and a by-hand storage latency probe outside the
Argo path. Cluster facts remain uninspected pending read-only access.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Review fixes for M7.6-M7.9. Schema v9 applies one lock order: the
barrier takes a database-wide upgrade lock exclusively, while
acquisitions, lease checks, provisioning and the retired-build trigger
take it shared first, so a queued acquisition can no longer regain
authority for a just-retired build and concurrent barriers cannot
retire the last current build. Provisioning takes a per-world lock
before classifying a world as original or a clone of one source. The
operations listener makes at most four accept attempts per serve().
Race regressions prove each contender is blocked before commit.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Schema v8 binds authority to builds (ADR-0068): acquisition names the
host build and only retained, unretired builds may hold a lease; no
commit may name a retired build; upgrade_barrier retires a build and
fences every holder in one transaction, while build-less pre-v8 hosts
keep working until the first barrier. Tests cover a writer surviving a
failed and then a real online migration, old-writer fencing on every
path, whole-history recovery by the new build, and the M7 qualification
fixture (19.6 s SIGKILL-to-ready under the production lease, no lost
acknowledgements).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Clone a verified committed head into an isolated world and database
(ADR-0067): world re-identification with a proven byte round trip, a
schema v7 clone genesis naming the source commit and digest with digested
inherited outcomes, a read-only source session, an owner-only atomic
target transaction and the evo-world-clone tool. audit_region_history
re-verifies every retained link. Real-process tests cover trajectory
equality after the cut, source isolation, killed clones and retention.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
ADR-0066: evo-server --ops-listen serves /livez (loop responding, never
authority), /readyz (200 only while the host holds the lease with a
verified recovered head) and /metrics (Prometheus text: 30 fixed
series for readiness, phase, lease epoch/remaining, simulated/committed
ticks, commit lag, checkpoint age, tick debt, queue use and bounds, and
counters; the only labels are a 6-value phase and 4-value drop reason).
The dependency-free OpsServer runs on the host loop with no thread:
GET on exact paths, 405/404/400/431 responses, 8 connections with excess
refused on accept, 4 KiB requests, 1 s deadlines, 16 KiB responses.
HostStatus/phase types moved to the portable evo::ops library with a
Windows-tested renderer. evo-server now refuses to start without its
required identity and connection settings.
Adds the server-runtime Docker target (evo-server + pinned libpq,
UID 10001, no credentials), a CI server-image job with smoke script and
optional publishing, and a Kubernetes example outside Argo.
Validated: Windows CTest 59/59; Linux Debug/Release/clang-tidy/ASan
CTest 59/59; PostgreSQL suites under Release, clang-tidy and ASan with
real process probes; two ops mutations caught. Docker build/smoke and
Gitea CI not executed locally.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
ADR-0065: before any readiness, recovery now verifies the head's tail.
The trace object must exist with its manifest digest, its header must
bind profile, region, predecessor and range; the predecessor (schema v6
load_region_commit) must hash to the digest the head names; replaying
the trace from the predecessor's verified payload must reproduce every
report and the exact head bytes; and trace-derived decisions must equal
the head's durable decisions. Missing, corrupt or non-reproducing tails
quarantine (RecoveryError::tail); nothing older or empty is served.
Test-only crash seams (HostConfig::crash_hook, PublishHooks) and
evo-server --crash-at deliver real SIGKILL at before-upload,
before-transaction, after-commit and after-ack. The kill matrix,
a SIGSTOP/SIGCONT stale owner and damaged-head startups all behave as
the durability contract requires; successors verify before readiness.
Validated: PostgreSQL suites under GCC Release, clang-tidy and
ASan/UBSan; tail-skip and replay-equality mutations each caught. No
Docker/Gitea CI, production-lease kill test or deployment.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
ADR-0064: RegionHost<State> (environmental and lineage profiles) is a
single-owner, single-threaded state machine advanced by step(now):
acquire the fenced lease and recover the verified head, run, resolve
ambiguous commits, drain and stop, or quarantine. Admission is
provisional and bounded (64 inputs, 64 KiB) with backpressure; FinalAck
and FinalOutcome are emitted only after a confirmed commit, and inputs
that will never be acknowledged are reported with a reason. Ticks follow
the caller's scheduling clock with bounded catch-up and measured tick
debt, never invented inputs. A death seals the batch and commits in the
same step, so no dependent tick uses an uncommitted outcome. Storage
outages pause and retry, unknown outcomes resolve by exact bytes,
fencing discards unconfirmed work and re-acquires; readiness fails
closed. Drain commits pending ticks and releases the lease.
evo-server hosts one region with a bounded stdin input feed, stdout
event lines, a readiness file and SIGTERM drain (exit 0/3/4).
Validated: Windows CTest 58/58; PostgreSQL suites under GCC Release,
clang-tidy and ASan/UBSan including real-process SIGTERM drain and
SIGKILL with successor recovery; four host mutations each caught. No
gateway, metrics, Docker/Gitea CI or deployment.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Review P1 (object durability): a name installed by a put whose directory
fsync failed was later reused as success without syncing again, so a
publication could acknowledge an object that might vanish on power
loss. Reuse (existing and lost-link-race paths) now fsyncs the verified
file and the directory before succeeding, and namespace opens always
sync their parent. A test-only ObjectStoreFaults hook injects directory
sync failures; tests cover failed sync then retries, namespace creation
syncs, and publication refusing to proceed until durability holds.
Review P1 (pre-v5 history): v5 recovery requires one durable decision
per terminal actor, but v3/v4 published checkpoints carry none, so an
accepted upgrade could make regions unrecoverable. pre_v5_guard.sql now
runs before 0005 in every upgrade path and refuses (EVL10) when any
published pre-v5 checkpoint exists; genesis-only regions pass. 0005 is
unchanged. CI publishes a real v4 checkpoint and proves the upgrade is
refused with nothing changed; a committed death without a decision is
shown to recover as quarantined history.
Validated: PostgreSQL suites Release x3, clang-tidy, ASan/UBSan x2;
reverting each fix is caught by its test.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
ADR-0063: manifest v2 (360 bytes) binds a per-tick validation trace
object and a canonical decision block by SHA-256. Publication replays the
trace from the committed predecessor's verified payload and requires
every report and the candidate bytes, derives deaths deterministically,
uploads payload and trace with verified read-back, then sends manifest
and decisions to one fenced statement. RegionBatch seals at a terminal
transition or 64 ticks and refuses ticks until durable confirmation.
Schema v5 records the checkpoint, its decisions and outbox rows
atomically; a unique index keeps one terminal outcome per actor
(EVL08), triggers make commits/decisions immutable (EVL09), an exact
retry of a committed ID returns the original receipt, changed bytes
conflict, and read-only resolve_region_commit answers committed/absent/
conflict. Outbox pending/acknowledge give idempotent at-least-once
delivery. Recovery reconciles durable deaths with restored lifecycle
(and archive deaths for the lineage profile).
Validated: Windows CTest 58/58; Linux GCC Debug/Release, clang-tidy and
ASan/UBSan CTest 58/58 each; PostgreSQL suites Release, analysis, ASan
(lost response, committed-after-timeout, terminated orphan retry,
changed bytes, repeated/partial/reordered decisions, gap, outbox
redelivery, history quarantine); five mutations each caught. No
Docker/Gitea CI, process kill, broker or deployment.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
ADR-0062 adds a canonical LineageArchive codec (parents before children,
organisms by id, FNV) that decodes by replaying admit/record_birth/
record_death and requires byte-identical re-encoding; it never overwrites
an existing archive. LineageRegion binds the environmental producer and
consumer accounts to archived organisms with identity high-waters,
enforces closure (lifecycle vs deaths, exact living set, tick bounds)
and records deaths in the tick that causes them.
Checkpoint profile 2 (environmental-lineage-v1, history profile 1)
publishes that container; schema v4 admits it with a 4,268,144-byte cap
and requires each commit to keep its predecessor's profile, RNG, history
and durability declarations. Linux adds provision/publish/recover for
the profile on the shared publication path.
Qualification found WSL2 stepping CLOCK_REALTIME back ~0.5 s every ~32 s,
which made two 1 s-lease tests flaky under ASan. Tests now wait on
database-time expiry; ADR-0060 records the slewed-clock requirement.
Validated: Windows CTest 57/57; Linux GCC Debug/Release, clang-tidy and
ASan/UBSan CTest 57/57 each; PostgreSQL suites 2x Release, 2x analysis,
5x ASan; SQL and C++ mutations each caught. No Docker/Gitea CI, remote
store, process kill or deployment.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
ADR-0061 publishes environmental-v1 full checkpoints. Manifest v1 is a
fixed 280-byte record (identity, chain, lease, versions, build, zero
RNG/history declarations, durability policy, payload length/SHA-256)
that PostgreSQL parses and digests itself. Schema v3 adds region_commits
with a digest CHECK and a composite head FK, owner-managed retained
builds, fenced publish/load functions and owner-only genesis; it drops
the bare v2 head kernel and fails closed while v2 heads exist.
Linux evo::region_store adds OpenSSL SHA-256, an immutable
content-addressed filesystem object store (temp+fsync+link, no
overwrite/delete, O_NOFOLLOW exact-length verified reads), and
publication (validate -> upload -> read back -> fenced transaction ->
receipt check) and recovery (re-digest, allowlist, verify, full restore).
Portable evo::region_checkpoint holds the codec and payload validation.
Validated: Windows CTest 56/56; WSL Release, clang-tidy and ASan/UBSan
private-PostgreSQL suites; four SQL and one host mutation each caught.
M7.4 stays open for 7.4b (lineage archive codec and history binding).
No Docker/Gitea CI, remote store, process kill or deployment.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Review P1: the v2 attribute-only runtime role check accepted a login that
inherits pg_write_all_data, which can UPDATE lease rows directly and
bypass fencing. apply.sql now audits, on every migrate run, the runtime
group and all transitive members: privileged attributes, any membership
outside the group (inherited or SET-able), table privileges in evolution
(including PUBLIC), CREATE rights and EXECUTE beyond the five fenced
functions. 0002 and its checksum are unchanged. CI reproduces the bypass
and proves each configuration is rejected and rolled back.
Review P2: valid(LeasePolicy) bounds every term by the duration before
arithmetic; LeaseAuthority validates durations before deadline math and
renewal_unknown ignores impossible (>60s) and fences non-positive requests.
Extreme min/max tests reproduce the old UBSan overflow and now pass.
Validated: Windows CTest 55/55; WSL Release, clang-tidy and ASan/UBSan
private-PostgreSQL suites.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
ADR-0060 and schema v2 add database-time region leases and a fenced head
advance. Acquire/renew/release/advance lock the lease row, check holder,
epoch and expiry with clock_timestamp() after the lock wait, and the
advance compares the exact predecessor. Holders are DB-generated tokens;
epochs never wrap or repeat. The runtime group role gets EXECUTE on five
SECURITY DEFINER functions only - no table, provisioning or CREATE rights.
Migration history admits clean->v2, v1->v2 and a v2 no-op, and refuses
privileged runtime roles.
Host side: portable LeaseAuthority fences from send-time deadlines on
CLOCK_BOOTTIME, keeps the earlier deadline on unknown renewal and stays
fenced until a newer epoch. Selected timings: 20s lease, 5s renew, 2s
margin, 2s renew and 5s commit deadlines. Linux evo::postgres wraps the
functions with typed LeaseStoreError results.
Validated: Windows CTest 55/55; WSL Release, clang-tidy and ASan/UBSan
private-PostgreSQL suites (contenders, expiry, paused/queued stale
writers, unknown renewal, exhaustion, bypass denial, upgrade path).
Three SQL mutations were each caught. Docker/Gitea CI, remote TLS and
cluster runs were not executed. No director, host or deployment added.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The two runs allocated contested breeding slots in different orders: the
individual run by member insertion order, the aggregate run by stored class order,
which is youngest first. When eligible parents outnumber free slots that decides
which parents reproduce, for a reason that has nothing to do with averaging. The
control never reached the capacity boundary and so could not see it, which left
ADR-0040's central claim unsupported rather than merely imprecise.
The allocation order is part of the stated rule now, oldest first, honoured at both
resolutions: classes partition ages, so the oldest class holds the oldest members,
and ordering inside a class stays something only the individual run can see, which
is a resolution effect rather than an implementation difference.
A second control sits at the capacity boundary, twelve eligible parents contending
for one free slot, and both controls compare the whole distribution of members
rather than censuses. That change matters by itself: whoever breeds hands the same
investment to a child either way, so every total agrees while the distribution does
not, and a census-only control cannot see a different parent having bred. The
probe diverging only by tick 44 is that effect.
Attribution was verified by mutation rather than argued. Restoring the individual
run's insertion order fails the new control; restoring the aggregate run's
youngest-first order fails it; the earlier one-unit drift in aggregate upkeep still
fails the first. Since either allocation order alone fails, the reported divergence
is attributable to averaging.
My first attempt at the control was wrong and the same sweep caught it. I gave the
contested fixture ages descending with the index, reasoning that this made
insertion order differ from age order; it does the opposite, and the
individual-side mutation passed. Ages ascend now, so oldest-first disagrees with
both the insertion order and the stored class order. Without the sweep the new
control would have looked like it established attribution while covering half of
it.
Also added the depletion check the ADR's attribution argument relies on: a shared
store that ran out would make intake order-dependent too. I had written that the
fixture asserts this before it did, which is now true.
The measurements were regenerated and are unchanged. Saying so plainly rather than
as confirmation: the finding was about what the numbers can be attributed to, not
about their values.
Release 38/38, analysis, ASan 38/38 and clang-format all ran and passed here.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
M4.5 closes M4 by measuring what aggregation costs, and adds no library code
except one defect fix the measurement found.
One rule is written once and applied at both resolutions. The control is what
makes the rest mean anything: with one member per class a class average is that
member, so the two runs must agree exactly, and they do, tick by tick, in every
channel. A deliberate one-unit drift in the aggregate rule is caught by it, so the
instrument is verified rather than assumed. Without it a measured gap would be
indistinguishable from a transcription mistake in one of the two runs.
The result is reported, not smoothed into a tolerance. At class widths ten and
fifty the aggregate population survives ten thousand ticks within 375 parts per
thousand on population and 43 on energy. At twenty-five, one hundred and two
hundred it dies out entirely while the individual run sustains seventy-five
members: a false extinction caused by resolution alone, because a whole class
starves the moment its average cannot pay upkeep, where individuals starve a few
at a time and the survivors keep the shared per-capita share up.
The collapse is not monotone in class width. Twenty-five destroys the population
and fifty sustains it. An earlier version of the test asserted that coarser is
never better; the measurement refuted it, so the assertion went rather than the
widths being chosen to make it hold. The non-monotonicity is asserted now, because
it is surprising and because the warning written around it could otherwise become
wrong without failing. A tier policy cannot assume degradation worsens smoothly as
classes widen, and M4.6 inherits that.
Conservation was exact at every tick on both sides of all five runs, including the
three that died out. That is the clearest available demonstration of the thing
ECOLOGY_MODEL keeps saying: conservation is not ecological evidence.
Two flaws in my own experiment were found before any number was believed. The
aggregate model had been producing overlapping classes, so the model measured was
not one a cohort could represent; newborns now extend the youngest class while it
is narrower than the width, and every step is checked to construct a valid Cohort.
The first figures were discarded.
The comparison also found a real M4.3 defect, which is what a validation milestone
is for. plan_advance checked affordability with arrivals added before losses were
taken, but emitted the ledger legs outflows-first, so an interval affordable as a
whole underflowed leg by leg and was refused at commit. A fifty-tick interval
shedding 10,480 against 6,950 held while drawing 13,080 in reproduced it. Arrivals
are the first leg now, matching the order the arithmetic was checked in, with a
regression and a corrected escrow expectation in an existing test.
Release 38/38, analysis, ASan 38/38 and clang-format all ran and passed here.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Two defects from review of 380d609.
Materialization treated an empty reservoir as available. A vacant cohort, or one
holding nothing, still owns its store, so a draw could credit it while the owning
record went on recording nothing: the same inventory in two places, reached from
the opposite direction to the destination defect in 1a59b63. Destinations are
checked against the registry's ownership now, not just their balance, and the
registry answers whether a reservoir is used rather than callers inferring it from
what the reservoir contains.
The same hole existed on the collapse side in a worse form, found by asking
whether the reported defect had a sibling. Nothing verified the destination was
unowned, so the shared-reservoir rule added in M4.3 would have refused the install
only after the transfer had run, leaving the members emptied and their biomass in
a reservoir nothing accounted for. That is exactly the stranding the M4.2 ordering
was chosen to prevent, defeated by a check that ran too late. Whichever registry
step follows a transfer is now proved to succeed before it, through may_install
and may_replace queries that install and replace delegate to, so there is one rule
and not a restatement that can drift.
The second finding was a functional gap the oscillation test hid by discarding its
registry each cycle. A full draw leaves a vacant record, which ADR-0036 retains
because absence is not extinction, and collapse then refused that habitat and
genome as occupied: a full materialization was a one-way door out of a habitat and
the lineage could never return. Collapse repopulates a vacant record in place now,
continuing its ownership and revision series rather than restarting them, and a
record still holding a population is refused, because merging two silently is not
a conversion. The test keeps one registry for all twelve cycles and asserts the
revision continues.
Four of the five new guards are load-bearing. The fifth is not and is recorded as
such: the may_replace pre-check in commit_materialize is unreachable given the
checks before it, since the installed record is already proved equal to the plan's
origin and the result differs only by a later revision. It is kept because it is
the right place for the guard and its collapse-side sibling is genuinely
reachable, not because it is exercised.
Release 37/37, analysis, ASan 37/37 and clang-format all ran and passed here.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
M4.4 turns whole members of a cohort back into identified individuals, and does
nothing else: no habitat transfer, no persistence, no resolution-equivalence
claim. Materialization is instantaneous at the cohort's current tick, because
advancing time is M4.3's interval.
The named risk is that re-materialization becomes free healing, and it has two
separable sources. The arithmetic one is closed outright. A drawn member gets its
class share by integer division rounded down, because rounding up would create
biomass, and a class drawn empty hands over everything it held because a class
with no members may keep no reserve. A residual is never drawable, so an expected
density cannot become an individual.
Age is taken as the oldest the class admits. The exact age is unrecoverable, so
the question is only which direction to be wrong in, and every other answer makes
oscillation profitable: a member re-materialized at the bottom of its class is
younger than it was, and repeating that is immortality. Taking the top means a
round trip can only age someone. Both of these are the risk rather than ordinary
validation, and both mutations are detected: rounding the share up, and drawing
the age from the class floor.
The second source is the declared resolution itself and is stated rather than
mitigated. Members inside a class are indistinguishable by construction, so
collapsing a starving member with a well-provisioned one and materializing both
returns two average members, healing the first at the second's expense. The tests
demonstrate both halves instead of only the flattering one: a round trip is exact
when each class holds one member, and a coarse class conserves the total while
averaging the individuals. Conservation of the total is not evidence that any
individual is right, and the mitigation is the caller's choice of classes when it
collapses, not anything this code can recover.
Materialization produces exactly the record a collapse consumes, so the two
compose and are tested against each other rather than described as inverses. The
oscillation test runs twelve cycles and asserts every cycle that the conserved
total never moves, nobody holds more than the richest starting member held, and
nobody is younger than last cycle; it ends by asserting everyone has been pushed
to a class ceiling, which records the cost of the age rule rather than claiming
it away.
Release 37/37, analysis, ASan 37/37 and clang-format all ran and passed here.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
M4.3 advances a cohort over one declared interval and deliberately does not model
one. ECOLOGY_MODEL keeps population maths an experiment and asks for several
candidates to be compared, so choosing one here would turn a recommendation into
a locked rule. A model proposes fluxes; this checks the declared result is
exactly what they imply, for counts and for every inventory channel.
Nothing is clamped. A loss larger than what is there refuses the interval and
says whether it was the population or the biomass, so a caller can subdivide;
subtraction is checked against what is present rather than performed and
repaired, so a negative population is never constructed. That is the answer to
the named risk: a clamp is how an interval a model could not integrate becomes a
plausible result, and how a false extinction arrives. Both clamp mutations are
detected by the tests.
Counts carry whole members and owed fractions as one quantity, so a population
below one member is a number rather than a rounding decision. Biomass has no
birth term at all, so a birth cannot bring inventory in and newborns are
necessarily funded from what the cohort holds. Immigration draws from a reservoir
that already held it, so a late transfer is not new production. Conservation runs
through the ledger as one atomic transaction; an interval where nothing moves
submits none.
Two amendments to M4.1 were forced, not chosen. A cohort records the reservoir
backing it, because an advance cannot otherwise know where the inventory is and
passing it alongside would let a caller advance one cohort against another's
store, which comparing balances cannot detect between equal reserves; the
registry now refuses two cohorts sharing one. And a cohort may be vacant, through
a separate factory so vacancy stays deliberate, because a decline to exactly zero
was otherwise unrepresentable and the only way through would have been the clamp
this milestone exists to refuse. A vacant cohort is not an extinction: the
record, its genome, its habitat and its provenance remain and it stays installed.
Fourteen guards confirmed load-bearing by removing each and rerunning. Two of my
own expectations were wrong rather than the code: escrow arithmetic in the
migration test, and a dwindling case that shed all but one femtogram while
declaring a result holding none. The second surfaced a real tension now recorded
in ADR-0038 rather than worked around: a class reduced to a fraction of a member
may hold no reserve, because reserve belongs to members and a fraction is not one
yet, so that biomass has to go somewhere the same interval.
Closure is checked and stability is not, so an interval whose arithmetic balances
and whose ecology is nonsense is accepted. The declared age structure is not
verified to be what ageing would produce, because that needs the within-class
distribution the record discards.
Release 36/36, analysis, ASan 36/36 and clang-format all ran and passed here.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Commit proved the destination reservoir existed and that the registry slot was
free, but never that the reservoir was empty. Gathering into an occupied
reservoir therefore succeeded and mixed inventory the conversion had never
sampled into the cohort's backing store: the cohort's recorded reserve
understated what backed it, and the surplus belonged to nobody.
Reproduced against the unfixed code rather than taken on trust. Gathering into
reservoir 5, which the fixture already seeds with seven picojoules, commits and
leaves the ledger holding 67 against a cohort recording 60, exactly as reported.
A destination holding anything is refused now. The success test also asserts the
destination balance equals the installed cohort's reserve, so the equality is
checked rather than assumed, and the regression asserts the refusal leaves the
balances, the action counter and the registry as they were, then that emptying
the reservoir first makes the same plan committable.
The check runs after the staleness loop rather than beside the other destination
checks. Both run before anything moves, so it is a reporting choice: replaying a
committed plan into a fresh registry would be refused by either, and that the
members no longer hold what was sampled says more than that the reservoir is
occupied.
The guard was confirmed load-bearing by removing it alone. An earlier attempt at
that reported it as not load-bearing; that result was void and discarded, because
a heredoc had mangled the mutation script and nothing was ever mutated. Second
time this quirk has produced a false negative in a sweep, and both times the
printed output rather than the exit status is what caught it.
Release 35/35, analysis, ASan 35/35 and clang-format all ran and passed here.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>