3.5 KiB
P9.2 Step 3 Controls And Result Evidence
Source:
f4ced3f392e021d0d770318c5803eb35e641375cStep:P9.2-STEP-3 CONTROLS_AND_RESULTEvidence level:automated_verified,runtime_verifiedand supportingagent_visual_reviewed
Focused Runtime Marker
ABG_P9_2_CONTROLS_RESULT_RUNTIME_OK touch=10 fallback=0 direct=0 pause=true speed=1x,2x inspect=readonly retreat_tick=50 duplicate_receipt=true withdrawal_ticks=20 results=4 lines=5 layouts=12 viewports=844x390,1280x720,390x844 locales=zh_CN,en text_130=true casualty=false
The test proves that Pause freezes accepted ticks, 2x and 1x request two and one fixed ticks per 100 ms of presentation time, and speed/Inspect selection does not mutate battle authority. Retreat is disabled before tick 50, accepts one parsed ScreenTouch at tick 50, preserves its original tick-50/tick-70 receipt on a duplicate parsed touch, and resolves after 20 accepted ticks.
Victory, defeat, timeout and retreat each render exactly five ordered result lines: outcome, cause, survivors and HP, reward/cost, and next action. The final line explicitly states that P9.2 applies no permanent casualty or recovery. Home and Retry remain at least 44 px.
Strict Headless Set
Evidence root:
/private/tmp/abg-p92-step3-f4ced3f-headless.V3qqaZ/
| Log | SHA-256 | Marker |
|---|---|---|
p92-controls.log |
88582fa679feaacd2a48df05b840271757007e7f559ac2ca2f8d26d2da90d74c |
ABG_P9_2_CONTROLS_RESULT_RUNTIME_OK |
p92-domain.log |
cae7038217d4d4e15ceba435977152b5b910fe60e781402d7ea0a7bd8a252a17 |
ABG_P9_2_AUTHORITATIVE_BATTLE_DOMAIN_OK cases=93 actions=37 |
p92-horizontal.log |
f957b2d8fafc6c8f49ac212d00c601bfe728070e85555a9d703fb5b873ed0815 |
ABG_P9_2_HORIZONTAL_BATTLE_RUNTIME_OK |
p91-foundation.log |
c5aac0495cd734ca9eb7d68fc015088903b485da631f08a03116fe1085a9b5b9 |
ABG_P9_1_FOUNDATION_ATOMICITY_OK success=76 rejected=28 |
p91-session.log |
e66b7dd4b6b41090eef48a3fb04fc8823800b0d5f858acaf8af88cdfd4a1565b |
ABG_P9_1_FULL_FIRST_SESSION_OK paths=180 |
All five logs have one required marker and zero script/parse/compile/resource,
ObjectDB, cleanup, ERROR: or WARNING: markers.
Normal-Render Frames
Evidence root:
/private/tmp/abg-p92-step3-f4ced3f-normal.tTqH0b/
The normal Godot 4.6.1 GL Compatibility run exited zero. Its log SHA-256 is
a25081d94b3e1e4202f872d9b24720ef699db296224987abf4d023a59f923e9c
and has zero strict diagnostic markers.
| State | SHA-256 |
|---|---|
01-controls-locked.png |
c9a6e15b73ed3efc08ec4a0f94a0961627c52629847149ba09bd7a7080445675 |
02-speed-2x.png |
a0d7f3538f7e229d953a92f28c81870aac0b2e62c79484f001f59217cd7374c6 |
03-inspect.png |
ba5d92a1fbd363a55858caa2d81e57a20a0508480d1588a0d52f566f4d1eef0e |
04-retreat-pending.png |
948d3f239d25e070d0991b98d4665e6942ad87819143645085c6e9b6b6f2f0a6 |
05-retreat-result.png |
26796a37ec38581d8114467437e18d59364df32f97dd5aa6c69aa5470e21fc25 |
All five PNGs are native 844x390. Supporting agent visual review found the five controls readable without covering the party/enemy HP panels, the central event ribbon compact, Inspect intentionally modal but immediately reversible, withdrawal status explicit, and the five-line result plus Home/Retry readable.
Open Gates
- Step 4 integrated five-profile, full-path and H5 evidence is not yet complete.
- Owner understanding, fun, control feel, visual acceptance and listening are open. No human claim is inferred from these frames.
- Physical devices, persistence/casualty, packaging, stores and release remain frozen.