M11.9: record solo feedback and living-cell reconnect blocker
Foundation / checks (checks) (push) Successful in 5s
Foundation / checks (client) (push) Successful in 5s
Foundation / checks (ecs) (push) Successful in 4s
Foundation / checks (deploy) (push) Successful in 5s
Foundation / checks (postgres) (push) Successful in 4s
Foundation / checks (network) (push) Successful in 4s
Foundation / checks (tsan) (push) Successful in 4s
Foundation / postgres-tools (push) Successful in 5s
Foundation / server-image (push) Successful in 5s
Foundation / image (push) Successful in 5s

This commit is contained in:
dingo
2026-10-16 07:27:48 -04:00
parent 214991c9a3
commit 87c1edcd2c
7 changed files with 127 additions and 3 deletions
@@ -0,0 +1,77 @@
# M11.9 solo feedback recorded 2026-10-16
Source: the maintainer's answers to the eight questions in
[the predeclared playtest plan](M11_9_REPEATED_LIFE_PLAYTEST.md), supplied in chat.
This is a report date, not a verified session date. One solo participant is
reported; session duration, exact life count, region/server digest, client build,
assistance settings and start/end telemetry were not supplied. No additional
participants, full protocol completion or product acceptance are inferred.
## Findings against the original questions
| Question | Reported observation | Assessment and proposed follow-up |
|---|---|---|
| 1. Movement and food | The maintainer suggests smaller food pools with many more scattered spots, including clumps, like rice thrown onto the ground. Further testing was blocked by "Existing cell is already bound to this account" on login. | Time to food and first offer were not measured. First repair/diagnose returning-player access; then compare scattered small patches and clumps while explicitly controlling total finite resource stock. Distribution is a proposal, not accepted balance. |
| 2. Mutation tradeoffs | Choices/costs are visible. No meaningful play change has been established; too few mutations were obtained to assess a viable change. | Readability has positive solo feedback; affordability, accumulation and meaningful effects need further observation. Do not claim mutation variety is validated. |
| 3. Predation | Testing was unsuccessful/incomplete. More testing and actual UI/animation changes are needed, including toast-like feedback describing what happened and why. | Predation usability is unvalidated. Exercise successful and refused engulfment, owner-visible reasons, timing and presentation with more players. |
| 4. Descendants | The maintainer budded intentionally but questions its purpose. A possible future idea is creating similar cells and gathering them for multicellularity. | Intentional budding is observed; competition, ecological significance and player motivation remain unproven. Record multicellular gathering as an exploratory design idea for the later gate, not a new M11 rule or mandatory progression path. |
| 5. Death and restart | Cause of death seems clear. The message should appear in the middle with an option to start as a new cell. | Preserve terminal death and primitive restart. Improve prominence of the death/restart presentation; never describe the old organism as reviving. Repeat the flow with the same account. |
| 6. Offline | Offline behavior seems fine so far. | Positive limited feedback; it does not close the offline gate. The subsequently clarified return-to-living-cell failure below needs separate investigation. |
| 7. Variety | No variety was experienced in the solo test. Each new life rushed to the nearest food, then stayed feeding until it was drained. Offered mutations did not meaningfully change play; long waits dominate. A higher feeding rate might help, but the maintainer is unsure. | Strong solo evidence of a repetitive loop, not a measured cross-player result. Test food placement, feeding dwell time, mutation cadence and actual behavioral effects separately. Feeding-rate changes are a hypothesis, not an approved tune. |
| 8. Fun | The maintainer is unsure whether the loop is fun; more people and further confirmation are needed. | Product acceptance remains open. Do not substitute passing engineering tests or synthetic ecology for the missing product review. |
## Blocking incident: M11-PT-001
Reported message: **"Existing cell is already bound to this account"**.
The maintainer clarified that this occurred when **logging back into a living
cell**, and **nothing appeared**. This is a resume/reconnect incident, not a request
to replace a living organism or evidence that the organism died. Exact region,
build, logout interval, server tick/commit, session route and persisted presence
were not supplied. Root cause and a live fix are not yet established.
Read-only source triage found:
- `RemoteSession::cellular_results` displays this string for the server's
`PrimitiveLifeResult::existing_life` result. Each freshly opened client issues
a primitive-life request; that result alone cannot establish a resume failure.
- The outcome intentionally carries no existing life identity on the wire. The
owner-filtered perception supplies the actual controlled life when available.
- The client disables a new-life attempt for `existing_life` while waiting for
that perception. Allowing a second life or changing the account is not a repair
for an invisible living organism.
- ADR-0128 excludes suspended cells from replication/perception. The accepted
resume rule tries the retained position and bounded nearby positions, and can
remain suspended when all are blocked. The server's owner tick reports
`resume_blocked`, but the bridge skips suspended lives before building their
owner perception. This is a diagnostic candidate, not confirmation that the
maintainer's cell was blocked by placement.
- The server should tick when an established session holds a living account and
submit an unfrozen baseline to request resume. Verify that ownership routing,
account control, publication and subsequent perception actually complete on
the affected DEV instance; local owner resume tests alone do not prove this.
Investigation order: reproduce return to an isolated suspended cell; repeat after
wildlife occupies its old position; distinguish bounded placement refusal from
wrong/unavailable region routing, non-advancing commits and missing owner views.
Retain life/genome/inventory identities through recovery. Do not delete bindings,
create a new account, reset a region or expand the resume radius as a workaround.
No deployment, database mutation or runtime fix was performed for this note.
`kubectl config current-context` reports that no current context is set in this
workspace, so live DEV logs/state were unavailable for confirming these candidates.
## Retest backlog
| Local issue | Priority / scope | Acceptance evidence |
|---|---|---|
| M11-PT-001 | Blocking: living-cell reconnect/resume | Same account returns to the same funded life after logout/recovery; control and visible owner state arrive, or a specific bounded-resume failure is explained. No duplicate life or resource changes. |
| M11-PT-002 | Food distribution experiment | Compare scattered small spots/clumps with the current layout under matched total stock; record travel, competition, dwell time and time to first offer. No implicit refill. |
| M11-PT-003 | Feeding/mutation pacing and effects | Observe enough affordable choices to compare movement/intake/defense behavior; measure waits and test one declared versioned tuning hypothesis at a time. |
| M11-PT-004 | Predation feedback and animation | Multiple-player successful/refused attempts have timely, understandable reasons; owner feedback does not expose hidden opponent state. Retest presentation with participants. |
| M11-PT-005 | Central death/new-cell action | Prominent cause and primitive-new-life action; reconnect and repeated deaths never restore the dead organism. |
| M11-PT-006 | Budding purpose / later multicellular idea | Explore ecological reasons to bud during M11 testing. Review any gathering/multicellularity proposal at the later design gate without mandatory intelligence, collection ownership or a linear class ladder. |
The suggestions do not select new resource budgets, feeding rates, mutation sizes,
resume radii, new life stages or a multicellular progression mechanic. First unblock
the test, then recruit the predeclared sample and repeat sessions before accepting
the loop or using it to calibrate M12 cellular population fidelity.
@@ -1,5 +1,9 @@
# M11.9 repeated-life playtest plan
Recorded feedback: [2026-10-16 solo report](2026-10-16_M11_9_SOLO_FEEDBACK.md).
It preserves the original questions below, records a living-cell reconnect blocker
and six local retest issues, and does not grant product acceptance.
This plan predeclares what the M11.9 playtests try to learn, who plays, how sessions
run and what is recorded. It is written before any session so that results are not
read into after the fact. It is not a balance sign-off; every value it touches is a