docs: open P9.0R4 second cycle

2026-08-13 01:23:31 +08:00
parent 232e3a0a45
commit 05cc64279d
5 changed files with 236 additions and 6 deletions
@@ -10,12 +10,18 @@ Runtime source belongs under `runtime/`; accepted HTML design authority remains
 under `prototype/` and may not be silently rewritten to fit implementation.
 P9.0R3 is the Owner-accepted baseline for tutorial-battle pacing and
 event-bound combat feedback inside the preserved Title -> Guild Yard ->
-tutorial battle -> victory flow. Its source is frozen.
+tutorial battle -> victory flow. Its source is frozen. P9.0R4 is the only
+active runtime slice: diagnose that first victory, apply the existing 20-Coin
+Steady payout, choose one of two run-local loaned adaptations, and see the
+consequence in one second short battle. It uses a fast H5 preview before full
+candidate evidence so Owner direction can be reviewed without waiting for a
+complete matrix on every iteration.
 It shortens dead holds between accepted authored animation states and adds
 bounded damage/heal number pops, impact nudge and existing VFX timing without
 changing HP results, roles, victory, animation assets or the opening flow.
-Contract, difficulty, Market, equipment, Party order, authoritative combat,
-persistence, chapters, device packaging, stores and release remain frozen.
+Contract, difficulty, Market, permanent equipment inventory, Party order,
+authoritative combat, persistence, chapters, device packaging, stores and
+release remain frozen.

 ## Originality Boundary

@@ -1,13 +1,30 @@
 # Current Checkpoint

-Active Goal: `docs/goals/PHASE9_0R3_BATTLE_FEEDBACK_GOAL.md`
-Active contract: `docs/runtime/PHASE9_0R3_BATTLE_FEEDBACK_CONTRACT.md`
+Active Goal: `docs/goals/PHASE9_0R4_SECOND_CYCLE_ADAPTATION_GOAL.md`
+Active contract: `docs/runtime/PHASE9_0R4_SECOND_CYCLE_ADAPTATION_CONTRACT.md`
 Active ledger: `docs/reviews/PHASE9_GODOT_VERTICAL_SLICE_LEDGER.md`
+P9.0R4 baseline: `232e3a0a45ec2437ddd528502d487d5264f1c6ae`
 Animation repair baseline: `db75a17ce8b0a3bc702ad535fa16491debf582e4`
 Fixed product source: `425fae82d3d88a002bdaadeabb498d8bf0d069cd`
 Battle-feedback behavior commit: `3b26b3d671ae7d043ab4bfcbd8b31e77df6d6d0d`
 Battle-feedback baseline: `be17debf4ed3ebe5c3e6de2701d5e3c374de1dcb`
-Stage: `P9_0R3_OWNER_ACCEPTED`
+Stage: `P9_0R4_SECOND_CYCLE_ADAPTATION_ACTIVE`
+
+P9.0R4 owns only the first-result diagnosis, the existing `B0/R1/Q0` Steady
+`+20 Coin` payout, one unselected choice between run-local loaned `EQ-002` and
+`EQ-061`, and one second tutorial battle whose survival/time consequence is
+visible. The accepted first battle remains unchanged. Market, permanent
+inventory, Party order, casualty/recovery, persistence, chapters, authoritative
+combat, packaging and release stay frozen.
+
+This milestone uses a two-stage review cadence. First publish a fast H5 preview
+after its focused test, one `844x390` normal-render route and strict scan. Only
+after Owner direction acceptance run the complete candidate regressions,
+matrices, both route recordings, immutable inventory, PCK/browser and final
+Gallery gates. Shortest next action: implement the bounded second-cycle loop
+without reopening P9.0R3.
+
+## P9.0R3 Owner-Accepted Baseline

 Fixed clean/upstream source `425fae8` tightens the six-phase automatic battle
 from the prior `13.50` seconds to `7.25` declared / `7.33` rendered seconds.
@@ -0,0 +1,117 @@
+# Phase 9.0R4 Second-Cycle Adaptation Goal
+
+## Owner Decision And Role
+
+On 2026-08-13 the Owner accepted the P9.0R3 battle cadence, authored character
+motion and event-bound combat feedback, asked development to continue, and
+reported that waiting for each playable delivery is too slow.
+
+- Role: development task.
+- Repository: `aetherbound_guild` only.
+- Baseline: clean/upstream P9.0R3 Owner-acceptance closure `232e3a0`.
+- Fixed accepted product source:
+  `425fae82d3d88a002bdaadeabb498d8bf0d069cd`.
+- Active contract:
+  `docs/runtime/PHASE9_0R4_SECOND_CYCLE_ADAPTATION_CONTRACT.md`.
+- Active ledger: `docs/reviews/PHASE9_GODOT_VERTICAL_SLICE_LEDGER.md`.
+
+## Player Hypothesis
+
+After the first automatic battle, a new player can identify the causal lesson,
+understand the earned reward, choose between safety and tempo, and see that
+choice change a second battle without opening a form-like Market or inventory.
+
+The bounded loop is:
+
+`first battle -> diagnosis -> +20 Coin -> two-option adaptation -> second battle -> changed result`
+
+## Owned Slice
+
+1. Replace the first-victory terminal with a readable diagnosis that names the
+   selected front/rear roles and their committed contribution before totals.
+2. Bind the tutorial fight to the existing `B=0/R=1/Q=0` Steady reward formula:
+   `battle_coin = floor(18 + 2) = 20`. Apply it once on first victory.
+3. Show the next known threat before preparation: the veteran Riftblade's first
+   strike pushes and heavily damages the exposed front.
+4. Offer exactly two stable, unselected, run-local loan choices from the
+   accepted equipment catalog:
+   - `EQ-002 Frontpin Greaves`: resist the first push and reduce its damage, but
+     delay the next front basic;
+   - `EQ-061 Early Bell`: resolve the first front basic before the enemy strike,
+     but delay the second front basic and accept the full strike.
+5. Commit exactly one loan, start one second automatic battle using the accepted
+   actors, animation states, feedback language and controls, then show the exact
+   surviving front HP, elapsed battle time and first decisive difference.
+6. Keep Home and confirmed Restart available throughout. Restart clears Coin,
+   adaptation and both battles; repeated input cannot grant Coin or commit the
+   loan twice.
+
+## Fixed Second-Cycle Fixture
+
+The second encounter is a bounded tutorial veteran using the existing
+Riftblade identity and assets. It is not a new catalog enemy or authoritative
+full-combat implementation.
+
+| Loan | First decisive change | Front result | Cadence consequence |
+|---|---|---|---|
+| `EQ-002` | First push is resisted; committed hit is `100 -> 90` | `90`, or `100` after Kindhand's capped heal | Front follow-up resolves later; safer route is slower |
+| `EQ-061` | Front basic commits before the first enemy strike | Full strike commits `100 -> 70`, then Kindhand may heal to `82` | Delayed second front basic still produces the faster route |
+
+Farstring keeps one committed rear hit; Kindhand keeps one committed heal. Both
+adaptations win this tutorial, but one preserves more front HP and the other
+finishes sooner. The result must compare only actual committed facts.
+
+## Preview-First Delivery Cadence
+
+P9.0R4 uses two explicit gates so the Owner can judge direction quickly:
+
+1. `FAST_PREVIEW`: run the new focused domain/UI test, one normal-render
+   landscape route and strict diagnostic scan; export and publish a clearly
+   labeled H5 preview. This is not a candidate and cannot upgrade final evidence.
+2. `FORMAL_CANDIDATE`: only after the Owner accepts the preview direction, bind
+   a clean pushed source and run affected regressions, required viewport/
+   language/accessibility matrices, both adaptation routes, MovieWriter,
+   immutable inventory, PCK scope, browser verification and final Gallery
+   publication.
+
+No formal evidence is silently omitted. Expensive evidence is postponed until
+the product direction is accepted instead of being regenerated for every
+preview revision.
+
+## Frozen
+
+- P9.0R3 first-battle HP values, roles, targets, event order, victory and timing.
+- Existing character, enemy, background, VFX and animation source assets.
+- Title, Yard selection, Home and confirmed Restart semantics.
+- Market, recruit hiring, permanent equipment inventory, Party order, risk
+  selection, casualty, recovery, persistence, chapters and content expansion.
+- Authoritative full combat, paid generation, audio expansion, device packaging,
+  TestFlight, stores and release.
+
+## Acceptance
+
+- The diagnosis precedes the reward and adaptation choice and uses no raw IDs as
+  its primary player explanation.
+- Coin changes `0 -> 20` exactly once on first victory and never on view/replay.
+- Neither adaptation is preselected. Each names benefit, tradeoff and expected
+  next-threat interaction before commit.
+- The two routes produce different event order, front HP and elapsed time while
+  preserving one victory and exact feedback deltas.
+- The second result names what changed because of the selected loan and offers
+  Replay and Home without opening the frozen Market.
+- The fast preview is playable in phone landscape and is clearly reported as
+  direction-review evidence rather than a fixed candidate.
+
+## Legal Terminals
+
+Preview:
+
+`Aetherbound Guild P9.0R4 fast H5 preview ready`
+
+Formal candidate after Owner direction acceptance:
+
+`Aetherbound Guild P9.0R4 second-cycle adaptation candidate ready`
+
+or
+
+`Aetherbound Guild P9.0R4 redesign required: <earliest failed player boundary>`
@@ -24,3 +24,4 @@ separate.
 | 2026-08-12 | source `425fae8`; `p9-0r3-425fae8-attempt03` | `INVALID_WEB_EXPORT_TEMPLATE_HOME` | Product, capture, recordings and normal audio-lifecycle test passed, but isolated Web export stopped before packaging because its HOME did not expose the installed Godot 4.6.1 templates. This was an export-environment failure, not a product failure, and the attempt was not reused. |
 | 2026-08-12 | fixed source `425fae8`; `p9-0r3-425fae8-attempt04`; Gallery `_N8bdzMkdn1c` | `P9_0R3_BATTLE_FEEDBACK_CANDIDATE_READY` | Fresh isolated archive passes feedback `duration=7.25 values=-18,-12,-36,+12,-30,-34 once=true`, default audio lifecycle, P9.0R2 animation and P9.0R1 domain/touch/layout regressions. Eighteen normal-render frames cover three landscape viewport/locale/text profiles. Two `1280x720`, 30 fps H.264 recordings run `15.233333` seconds with battle victory at `7.33` seconds, all four route payloads, visible victory holds and no teardown gray. Strict diagnostics are zero. The 65-row inventory SHA-256 is `d1040cec77d3cc63e8322c63d2df5dcc6ee88c35ab04329d52f6687c5529720a`. Runtime-only PCK excludes tests/evidence, and the nine-file Web manifest SHA-256 is `d48922ed4892e15057284275e237d6ef35dd77201a2c67efd170e870c98225bb`. Local Chrome reaches Title -> Yard -> battle -> victory, portrait rotate guidance and browser diagnostics zero; remote HTML/JS/WASM/PCK identity, MIME and bytes pass. Human phone feel/readability, fun, listening, device, packaging, store and release remain open. |
 | 2026-08-13 | Owner play review of Gallery `_N8bdzMkdn1c` | `P9_0R3_OWNER_ACCEPTED` | Owner reports that this version is acceptable and authorizes continued development. The accepted P9.0R3 battle cadence, character motion and event-bound damage/heal feedback become the frozen `human_reviewed` direction. Listening, broader fun/continuation, physical device, packaging, store and release remain open. The next slice must start from the first victory and prove one readable adaptation before a strategically different second battle; it may not reopen accepted battle values or presentation. |
+| 2026-08-13 | P9.0R4 task opened from clean/upstream `232e3a0` | `P9_0R4_SECOND_CYCLE_ADAPTATION_ACTIVE` | Owner authorizes continued development and requests faster playable review. The bounded slice is first-result diagnosis, existing Steady `B0/R1/Q0` `+20 Coin`, unselected loan choice `EQ-002` safety versus `EQ-061` tempo, one changed second battle and exact consequence. A fast H5 preview follows focused plus one-route runtime validation; complete immutable candidate evidence waits for Owner direction acceptance. P9.0R3 and all frozen Market/persistence/content/release scope remain unchanged. |
@@ -0,0 +1,89 @@
+# Phase 9.0R4 Second-Cycle Adaptation Runtime Contract
+
+## State Machine
+
+```text
+first_battle
+  -> first_result_diagnosis_and_coin
+  -> adaptation_unselected
+  -> adaptation_selected
+  -> second_battle
+  -> second_result
+```
+
+`first_result_diagnosis_and_coin` commits the existing Steady `B0/R1/Q0`
+victory Coin exactly once. Viewing or leaving any page does not mutate Coin.
+The adaptation is reversible until `Begin second battle`; after that it is
+locked for the encounter.
+
+## First Result
+
+The cause line is derived only from committed P9.0R3 facts:
+
+- Hearthguard: first strike `-18`, explicitly blocked;
+- Wayrunner: first strike `-12`, explicitly evaded/reduced;
+- Farstring: one rear hit `-36` opened the finish;
+- Kindhand: one heal `+12`, capped by current front HP.
+
+The player sees the cause before `+20 Coin`, current front HP and the next
+action. The result does not claim casualty, reward draft, Market or persistence.
+
+## Adaptation Choice
+
+| Stable ID | Player-facing promise | Tradeoff | Recommended when |
+|---|---|---|---|
+| `EQ-002` | Anchor the front against the veteran's first push; first hit becomes `-10` | The next front basic resolves after the rear action | No rear healer is present or survival is preferred |
+| `EQ-061` | Resolve the first front basic before the veteran strikes | Full `-30` strike lands and the second front basic is delayed | Kindhand can recover the hit or speed is preferred |
+
+Recommendation is explanatory, never selection. Both buttons remain equally
+reachable and the selected state can be changed before commit.
+
+## Second Battle Event Map
+
+### `EQ-002` safer route
+
+1. both lines enter;
+2. veteran telegraphs a pushing strike;
+3. Frontpin resists displacement and the front commits `100 -> 90` (`-10`);
+4. Farstring commits enemy `100 -> 64` (`-36`) or Kindhand commits front
+   `90 -> 100` (`+10`, exact capped heal);
+5. delayed front/rear coordination commits enemy to `34`;
+6. finisher commits enemy `34 -> 0`.
+
+### `EQ-061` faster route
+
+1. both lines enter;
+2. Early Bell advances the first front basic and commits enemy `100 -> 72`
+   (`-28`) before the strike;
+3. the full pushing strike commits front `100 -> 70` (`-30`);
+4. Farstring commits enemy `72 -> 36` (`-36`) or Kindhand commits front
+   `70 -> 82` (`+12`);
+5. the delayed second front basic commits enemy to `24`;
+6. finisher commits enemy to `0`.
+
+Every HP mutation emits one matching P9.0R3 feedback payload. Non-mutating
+telegraph/recovery states emit no number. Existing animations may be retimed or
+reordered but their source frames cannot change.
+
+## Presentation
+
+- Result and preparation remain battlefield scenes, not receipt tables or
+  nested cards.
+- The two loan choices read as physical tactical plinths with a familiar shield
+  and bell/tempo symbol, exact effect, cost `借用/Loan`, and one tradeoff.
+- The known threat and current two-person squad stay visible with the choice.
+- Second result leads with `更稳/Safer` or `更快/Faster`, then shows exact front
+  HP and elapsed time. It never says one route is globally best.
+- `844x390` is the primary fast-preview viewport; content must not require
+  scrolling. Portrait retains the rotate notice.
+
+## Preview Evidence
+
+- Focused test covers once-only Coin, no preselection, selection revision,
+  commit lock, both route event maps, exact HP, distinct elapsed time, Restart
+  and Home cleanup.
+- One normal-render `844x390` Simplified-Chinese route captures first result,
+  unselected adaptation, selected adaptation, second battle and second result.
+- Logs are scanned for script, parse/resource/import, assertion, ObjectDB and
+  cleanup markers.
+- The exported H5 is labeled preview. It is not immutable candidate evidence.