- Shared card pool and item stock as in the official mode (the owner's request, from the 路标档案馆 video and public notes): solo low-tier exceptions, co-op operators x2, shared item stock with special-item exceptions, normal / advanced occupancy 1 / 2, re-check on buy, 机变 exhaustion, release on consume / destroy / elimination; 整备 never exceeds stock; invariant held <= cap.
- 突袭: one taunt-first reselect right after a revive (the owner's rule); ordinary polls and redeploys keep the 0.2.3 rule; a revive on an enemy's tile stays. #475: online boss / hidden-boss Raid may land on the empty half (owner-limited scope).
- 浊心斯卡蒂 S1 x 坚守 no longer bounce redirected damage, and the two shares conserve the hit (idea of PR #464 by @HappyTanuki).
- Operator fixes adapted from PRs by @Xinglan233 (#459 哈洛德 S2 elemental trigger, #460 涤火杰西卡 shield on the four adjacent tiles, #461 家族徽章 first hit after stealth, #462 松果 S2 regrow per deploy); 克莱门莎 S2 splash 1.1 / 100% ATK and air auto-trigger (#467); 白面 copying 歌蕾蒂娅 / 余 skips the three-member condition (#473, the owner's video).
- 香槟炸弹 is not chosen by friendly auras / heals (PR #468 / #476 by @Siornya); 琳琅诗怀雅 coin balance (PR #450 by @IceCodeNew); Touch prep display (PR #451 by @YinqiMargatroid); lobby header (PR #471 by @YuriRestia); Japanese UI terms (PR #448 by @scutkane).
- AI-only boss countdown grace, render-failure cache / recovery without restarting a battle, device units kept on a same-field re-entry, DIY detail / skill previews, large-text result buttons.
- golden: battle families equal 0.2.3 except diy-135 / diy-136 (克莱门莎); all 18 release matches move through the shared pool, every flip named in docs/history/0.2.4.md §29.12; two coverage scenarios added.
- The owner asked to verify the official rule before deciding (research: PRTS 失衡位移机制 / 推与拉 / 游戏数据基础 / 特殊机制,
the client's Rigidbody2D data); the PR's 0.6387·√tiles formula and fear exemption were not taken.
- Grok review: MERGE AFTER FIXES — two HIGH (static pulls, 弧光锋卫's bleed), skills during the hold, husk 重生, exact
frames — fixed on the branch (1db84d79).
- Golden union: bond-miraShip-low differs from HEAD (fewer enemy attacks: 169 → 166 events, same kills) though the
branch alone left it — HEAD's ×1.1 练度 path now displaces an enemy there. Broad node --test 5681 (5679 pass, 0 fail,
2 skipped); lint 0 errors.
Seen: the sim moved a displaced enemy at once and it walked / swung again on the next tick; players compared with the
official game (PR #392: 「被推/拉完的敌人会先站住一小会儿再走」). Research (scratchpad report r3): PRTS 失衡位移机制 — a force > 0
puts an enemy in the UNBALANCE state machine with a 0.1 s 失衡硬直 floor, ending when its speed is ≤ 0.1 m/s, on 浮空 or a
forced state change; 异常效果图鉴/失衡免疫 「失衡期间无法自主移动、发动攻击、使用技能」; 异常效果 — not an abnormal status (水遁忍者's
check still runs); 游戏数据基础 推力-位移近似对应表 位移时间 (whole frames at 30 fps); 推与拉 拉力 「作用时间默认为 1 s；若本次受力
等级 < −1，作用时间改为 0.5 s」, 急停 keeps the state to the window's end; 特殊机制 静态刚体 (0.1 s 保底); 术语释义 浮空.
Now: constants.js PUSH_UNBALANCE {−2: 0.2, −1: 0.4, 0: 0.8, 1: 0.9, 2: 32/30, 3: 35/30}, PULL_UNBALANCE 1 /
PULL_UNBALANCE_WEAK 0.5, UNBALANCE_MIN 0.1. battle/displacement.js: push() holds for its level's row (the −2 correction
included), kept when the 特效 column or a wall shortens the slide [ASSUMED for the wall], the 0.1 s floor when a wall
stops it at the first step [ASSUMED]; pull() holds for its window whatever the travel (≤ −3: none; an enemy the puller
blocks: still no pull [ASSUMED, unchanged]); a 静态刚体 (data staticBody, not 失衡免疫 noDisplace) hit by force level
≥ −2 gets 0.1 s and no movement (_staticForce); _unbalance sets enemy.unbalanceUntil (a longer running one kept) and ends
the attack clip's stand. ai.js updateEnemy: no walking (恐惧 / 诱导 too — PR #392's fear exemption is not taken), no new
normal attack, a swing short of its frame cut, 浮空 clears it; a 重生 clears it (archetypes.js reborn). Not a stun: no
status, no immunity, SP and the attack cooldown run. No global skill lock: enemy abilities keep ticking ([ASSUMED] scope).
The move stays instant (the unbalanced() talents keyed off the jump unchanged). The displace fx carries `dur` (game s);
render/units.js slideTo divides it by ctx.animRate (2× default: a 0.8 s push slides 0.4 real s); DISPLACE_SLIDE·√D is
only the fallback for an fx without it. Not taken from #392: UNBALANCE_TRAVEL 0.6387·√tiles / the client's 0.3194 (derived
from the page's μ = 0.5 单位假设, one frame high on every push row, the wrong shape for pulls) and the fear exemption.
Tests: test/sim/unbalance.test.js (8: the push rows and walking after them; no attack in the hold, a feared enemy held;
the −2 correction's 0.2 s; 特效 / wall-shortened rows and the first-step floor; pull windows incl. the 急停 and ≤ −3; 静态刚体
0.1 s without movement, 失衡免疫 never; 浮空 clears; the fx `dur` — the file fails on the parent); test/render/unitview.test.js
(+1: the slide at 2× for a push and a pull, the fallback — fails on the parent); feedback1b-displacement B5 (the drone is
now held 0.1 s per force, never moved). test/sim + test/content + test/render: 3433 pass. Docs: SIM.md (push / pull,
displace, assumptions), PLAYING.md 推开 / 拉拽, research 13 §0 / §5.
Golden — every moved battle has pushes / pulls (probe of the callers) and its enemies are now held; outcome moves:
roster-002 (野鬃: killed 29→33, leaks 15→11), roster-003 (野鬃: deaths 3→7), roster-007 (薄绿: killed 110→118, leaks 53→47),
roster-008 (薄绿: 45→47 / 42→40), roster-010 (歌蕾蒂娅: killed 45→44, leaks 57→58), roster-011 (莫斯提马: 78→79 / 4→3),
roster-031 (钩索师 pulls, tier4: killed 64→72, leaks 12→4, deaths 6→5), roster-041 (钩索师: deaths 5→4), roster-043 (山:
deaths 15→14), roster-044 (山: killed 107→103, leaks 12→16, deaths 15→14); bond-deputShip-low (薄绿: 18→21 killed, leaks
3→0, deaths 1→0), bond-kazimierzShip-high (野鬃: deaths 3→2); boss-boss_3-pair (锏: p2 deaths 3→2), boss-boss_4-pair
(莫斯提马 / 歌蕾蒂娅: killed 11→12, boss HP 7442429→7375628, p2 leaks 2→1, deaths 19→18), boss-boss_6-pair (boss HP
6932482→6865243, p2 deaths 19→17), hidden-boss_8-pair (boss HP 12585770→12585798), hidden-boss_9-pair (锏: boss HP
3630612→3636991); standin-05 / 08 / 10 (预备干员-特种 pulls: killed 5→8 / 42→44 / 28→29, leaks down by 3 / 2 / 1);
diy-101 (傀影: killed 22→21, leaks 92→93), diy-114 / 115 / 116 (黑键: 106→99 & 83→90 / 154→156 & 78→70 / 3→2 & 120→121),
diy-137 (预备干员-特种: 68→79 / 79→68); matches through those battles — coop2-HARD-4, coop2-NORMAL-14-standins,
coop2-NORMAL-3, coop2-NORMAL-70-diy, coop3-ABYSS-5, coop4-ABYSS-7, solo-ABYSS-2 (LP 2→4 at round 10), solo-HARD-2
(ai_0 now survives round 13: eliminated 13 → alive), solo-HARD-9-boosted (LP 399→398), solo-NORMAL-1. Other moves (damage,
heals, snapshots only): roster-009/012/017/019/032/039/040/042/046, 13 bonds, boss-boss_2-pair, hidden-boss_10-pair,
standin-03, diy-053/084/085/102/103/108/109, coop2-ABYSS-10-boosted, solo-ABYSS-1.
GitHub PR #386 by @leiming2333 (this commit is substantially the PR: its blizzard particles and flash). The official
blizzard blows bottom-up (the owner's confirmation of 2026-10-08; community report 「谢拉格的效果特性不明显，暴风雪应该
从下向上刮」): snowflakes rise off the bottom edge on a crosswind (48, low 22; were 26 drifting down), 10 gust streaks
(low 4) race up through the field, and the cold flash goes from 0.35 / 0.8 s to 0.45 / 1 s. Counts, speeds and lives
are marked [ASSUMED] (a visual choice).
Added on top of the PR: the same 'chill' fx is also the coldWind of 灵知 S3 (失温症, at its start) and of 麦哲伦 S1's
pulse (every 3 s — 248 in one golden battle); the PR turned every one of them into the full-screen blizzard. Only the
bond's field-wide wind (the device emits no anchor unit) gets it now; an operator's own coldWind (`id` = the unit)
keeps the light flurry and the softer flash it had.
Tests: test/render/fxchill.test.js (2; the gust case fails on the parent). Display only: golden unchanged.
- Conflict: docs/PLAYING.md 技能自动释放 bullet — both insertions kept (the 铁钳号 rule from fb7-owner-sim, the #298
pacing sentence).
- Grok review: MERGE AFTER FIXES — the HIGH #298 finding (the cast replaced the due attack) fixed on the branch
(db29acf9: per-skill nextCastAt; golden of #298 alone shrank from 84 scenarios to 2).
- Golden union: 4 matches differ from HEAD though the branch alone left them (coop2-NORMAL-3, coop4-ABYSS-7,
coop2-NORMAL-8-serverrun, solo-HARD-9-boosted): HEAD's side had moved all four (#333's shop RNG …) and on those new
bot paths the merged rules apply — the diffs are in-battle layer gains / damage (卡西米尔 50 → 45, 精准 80 → 76,
叙拉古 176 → 168, kills ±2, solo-HARD-9 LP 397 → 399), no outcome change.
- Broad node --test 5571 (5570 pass, 0 fail, 1 skipped); lint 0 errors; typecheck.
A push and a kill on the same sim step bring the displace fx and the 'die' together: the view died before its slide
could anchor, and render/app.js syncBattle gives a dying view only its position (v.x = s.x) — never sync(), the only
place that set slide.live — so update() held the slide's start until its 0.5 s timeout, past the 0.8 s DIE window,
and dropped it: the body stayed on its pre-push tile (8.852 instead of the sim's 10.5; a summoned enemy and a flyer
alike). The per-view step of syncBattle is now units.js syncView (unchanged for living, device and revived views); its
dying branch calls UnitView.followSample, the position part of sync(): the sample anchors a slide from `at` on, and is
the position only when no slide owns it. A slide whose destination snapshot never shows (the unit left the snapshots)
now goes on to the fx's destination after the 0.5 s hold instead of being dropped where it started.
The frame-order test fed the view with sync() after its death and stepped the battle through the harness (whose
step() drained the events away from the test's own queue): it now steps the battle itself and goes through syncView.
Tests: test/render/unitview.test.js 38/38 — the app.js frame order over real battles: a 1.7-tile push; a kill on the
push's step for a route enemy, a summoned one and a flyer (the three failed before: bodies at 8.852 / 7.262 / 8.852)
and a kill 6 ticks later — the body ends where the sim's lies, every frame step < 0.35 tiles, no step back for the
same-step kills; a never-shown slide goes to the destination. Render interp, attack recovery, champagne, eyemask,
enemy attack clips, sim displace-fx: 94/94 with unitview.
The contributor's code (render/spine.js syncFormPose, the render/units.js wiring and FORMS entry, the headless
test/render/forms.test.js block), applied from PR #365, so they are the author of this commit. Maintainer side: the
hard-coded Spine id became a STEALTH_FORMS table (form while the stealth bit is off), comments and docs/ASSETS.md
reworded. Not ported: the PR's sim targeting rewrite (content/enemies/invisible.js), its enemies_bosses tests, its
roster.json and the Chrome suite (forms.browser.test.js, not run here).
Seen (#296, FrogThai): a blocked 骨刺 kept the snake. The sim already sends UF.STEALTH only while its 隐匿 is effective
(not blocked, not revealed — targeting.js enemyStealthed); the view now keys the FORMS 'revealed' clip set (Idle_B /
Move_B / Die_B / Attack_B, all in the skeleton) off that bit in UnitView.sync — before a DIE of the same snapshot — and
re-poses at once: a running attack continues in the other attack clip at the same point relative to its strike frame
(Attack_A hit 0.533 s, Attack_B 0.733 s; a pending wind-up still strikes on time), a stun / idle restarts on the new set
without a crossfade, and a frozen skeleton is posed with update(0); a model loaded later is posed the same way.
Golden: unchanged (render only).
Applied unchanged after review (Grok static review: MERGE; the diff read in full: no network, no dependency).
UnitView._updateImpostor refreshed a unit when (frame + phase) % interval === 0 with a random phase per unit, so the
refreshes of a frame were binomial: reproduced on the parent with 7 pen figures at interval 3 — 5 1 1 5 1 1 … and
2 4 1 2 4 1 … refreshes per frame instead of 2 / 3. Now render/app.js hands every impostor unit the next slot of the frame
(ctx.impostorSlot, a counter reset in frameBody), the unit refreshes when (frame + slot) % interval === 0, and a unit
whose slot keeps moving with the frame (the units before it culled on and off in step) is still refreshed after 2
intervals (imp.last); a context without slots keeps the random phase. Each unit's rate and the total work are unchanged;
no sim or pixel-as-state change.
Step 2 of #327 (a time budget by priority) is not part of this and still needs a decision.
- Tests: test/render/unitview.test.js 'impostor refresh slots' (3: slots in update order — fails on the parent, the
starvation cap, the no-slots context); unit view / render suites unchanged.
- Docs: design/client.md. Golden: unchanged (client only).
Applied unchanged after review (Grok static review: MERGE; the diff read in full: no network, no new dependency, no
shipped eval). The leak was real on feedback7: mountFieldView raced createFieldView against a timeout and, on failure,
removed every child of the host while the factory kept running — Pixi finished later and appended its canvas over the
simplified view (taking its clicks); the host wipe could also delete a newer view mounted by the same hook.
- ui/fieldHost.js: mountFieldView(host, { signal }) owns an AbortController; the timeout, a failed import or an invalid
view aborts it, a view a factory returns after that is destroyed, an outer abort (useFieldView's cleanup) rejects
without a fallback; nothing clears the host's children any more.
- render/app.js createFieldView takes the signal: the Pixi load / manifest / font waits end on abort, the synchronous
build registers its disposables (setupCleanup) so a throw releases the app, tiles, fx, impostors, listeners and
observers it made, the board-art and 3D-board waits moved after the view exists (aborting them is view.destroy(),
idempotent), destroy() also removes the canvas. render/app/host.js withTimeout(p, ms, signal).
- Tests: test/ui/fieldHost-lifecycle.test.js (5, node vm, fake timers: all fail on the parent) and the opt-in real-browser
test/render/initialization.browser.test.js (6 scenarios in headless Chrome, 5 fail on the parent: a late Pixi load
covering the fallback — with a factory that honours the signal and one that ignores it —, art timeout, a setup
exception, an unmount racing a newer mount); test/ui/playtest2.test.js's source pin takes the new indentation.
fieldview / browser suites in Chrome unchanged (the two 4× throttled fps tests flake when two Chrome files run
at once, pass alone).
- The createFieldView body stays unindented inside its new try (the PR's choice, to keep the diff reviewable).
- Docs: design/client.md §9. Golden: unchanged (client lifecycle only).
The three PRs patched the same snapshot and HUD lines (Grok's review: do not merge three diffs), so this is one design
built from their code and tests; IceCodeNew is the commit author (#303, #378), Convey123 co-author (#383).
- b.snap gains three optional, display-only lists beside the unchanged 9-field tuples (snapshot.js ammoView / wolfView /
negView, called from battle/events.js snapshot(); a living, deployed, visible ally, only when it has one):
ammo [[id, rounds left, magazine]] (whole rounds; the magazine is the activation's real total that unitTuple's SP
fraction runs on, extras and refills included), wolves [[id, 狼影 left, talent maximum]] (mem.shadows of the new
mem.wolfCapacity, set by tokens.js wolfPack and the managed vigil kit), neg [[id, fill]] (unit.negFill(), a function
op-hsgma2.js sets: the pool is read when the snapshot is taken, no per-tick sync). UnitInfo.ammoSkill (an ally whose
skill is an ammo magazine) is known from the first appearance: fx events run before the frame's sample, so the aura
cannot wait for the first ammo row.
- render/interp.js keeps the lists per snapshot (malformed rows / unknown ids dropped) and sample() hands them out from the
older snapshot like flags (they step with the skill flag); render/units.js draws them: the ammo magazine replaces the
SP bar (a cell per round, thin cuts, exact fill, no cuts under 2 px a cell), a row of diamonds under the bars for the
wolves (the element gauge row moves below), the red bar over the drained HP bar for the pool. An ammo skill draws no
standing aura (rings.js _aura; the activation burst stays).
- PR #378's wolf-head avatars are NOT taken: the count is data (the talent text 至多3只, mem.shadows), the HUD art of the
official counter is in no game data we have, so the pips are plain diamonds [ASSUMED look]. The PRs' own tests are
folded in and extended (test/sim/hud-readouts.test.js, test/render/hud-readouts.test.js, op_hsgma2.test.js).
- Docs: network.md, SIM.md §9, client.md, PLAYING.md, types/core.js.
- Golden: unchanged (283 scenarios; the canon hashes units / down / elem only, no sim state moved).
- Tests: 4 + 14 new + 1 in op_hsgma2 (12), all fail on the parent; 344 related tests pass; lint 0 errors, typecheck,
check:imports, docs tests.
Co-authored-by: Convey123 <Convey123@users.noreply.github.com>
SpineActor.attack() starts an attack that was not wound up (no look-ahead: a batch that arrived late) on its strike
frame — right for a quick rhythm, but for a slow clip-per-attack enemy the wind-up was cut out: 重犯 (Attack_grey
2.333 s, strike at 1.0 s, attacks every 3 s) showed the slam without the ball ever rising, and only 1.33 s of the clip.
The owner's decision of 2026-10-07 on #246: re-implement on the current spine.js so that, when the attack interval
leaves room for the whole clip, it plays once completely from the start. A clip-per-attack actor (enemies) with
interval ≥ the clip now plays it from 0 at its own speed for the whole clip, the strike landing a little late; a
quicker rhythm keeps the strike-frame start, and operators' looping attacks are unchanged (the PR's condition had it
the other way round: it changed the looping attacks and left out the clip-per-attack enemies it was written for).
Tests: test/render/feedback2-enemy-attack-clip.test.js (+1: 重犯 plays the whole clip from its wind-up, a quicker rhythm
and an operator start on the strike frame) 5/5; render eyemask, unitview, texas skill clip, skill idle 51/51.
The eyelids of 佩佩, 仇白, 隐德来希, 琳琅诗怀雅 and other models are Spine clipping attachments: the polygon is the eye's
opening and the eyeball / eye-white meshes drawn after it are clipped to it, so a blink or a knock-down closes the eye
through the mask. render/app.js pickClipping keeps the masks only for a lone clipped skeleton at high quality, so in a
battle the eyeballs showed through the closed lids. SpineActor._eyeMaskFallback now runs after each update while the
masks are off: it walks the draw order like pixi-spine (from a slot whose attachment in force is a clip, to its
endSlot — any skin) and hides the slot container of a covered slot whose drawn centre (a mesh's mean vertex, a region's
quad centre) lies outside the clip polygon [ASSUMED: all or nothing per slot]; turning the masks on shows them again.
PR #384's version hid `slot.currentSprite`, which pixi-spine leaves null on a mesh slot — every covered slot of the
417 local operator skeletons' 73 clipped ones is a mesh (313) or a region (54; regions have no worldVerticesLength, so
it never measured them) — and its test stubbed a currentSprite onto every slot, so nothing was hidden in the game. Kept
from it: the diagnosis, the ray-casting pointInPolygon and the key-frame expectations.
Tests: test/render/eyemask.test.js (new, 7, on real runtime objects — the official skeletons in @pixi-spine/runtime-3.8's
real Spine, @pixi/core display objects with a 16 px canvas stand-in, the real SpineActor: the eyeballs are meshes in
the clip's container; 佩佩 / 仇白 Die and 隐德来希 / 琳琅诗怀雅 Idle hide exactly the eye slots in the closed frames and
nothing open; a 30 Hz sweep of six clips; masks on hide nothing, turning them on shows all) 7/7 — the five skeleton tests
skip without public/assets (not committed; npm run assets), and fail on PR #384's code; render unitview, enemy attack
clips, texas skill clip 45/45.
The owner reported 「琳琅诗怀雅的2技能的香槟会掉血，会被治疗」. Allies always draw an HP bar (render/units.js
_updateHud), so the bomb's every HP change showed. PRTS 香槟炸弹 (实体类型 装置, 1000 HP) says no HP bar is shown even
below full HP; nothing there makes it invulnerable or unhealable. Its bar is now hidden at any HP (NO_HP_BAR). Its HP,
the damage it takes (活性源石 hurts what stands on it — 70 true damage a second) and healing are left as they are: no
invulnerability added; whether it should be unhealable is the owner's decision (no written source says so).
Tests: test/render/champagne-hpbar.test.js (new, 2: no bar at full HP, hurt or healed; another summon keeps its bar)
2/2; render/unitview 35/35.
The sim displaces at once (battle/displacement.js), so the snapshots only show the destination: a ≤ 2.5-tile push was
lerped over one snapshot interval (3 frames, 0.58 tiles in one frame — frame-loop repro over a real battle), a longer
pull snapped. render/app.js now starts UnitView.slideTo as soon as the interval it samples ends on the snapshot that
carries the destination (a `displace` fx still queued up to SnapshotBuffer.nextSnapT); the view holds where it stood
until that snapshot is shown, then eases into the sampled position under a constant deceleration (v0 = 2D/T, a = v0/T,
T = 0.14·√D in 0.12–0.45 s — ours, [ASSUMED]: no source gives the official timing; docs/research/13 has the official
impulse-and-friction classes). A unit that dies on the way finishes the slide and lies where the push put it.
Death: DIE_CLIP_MAX 1.6 → 8 s (data/assets.json: 234 enemy Die clips, the longest 7.97 s, 25 over 1.6 s), and a
flyer holds FLY_HOVER until the fade tail (DIE_FADE_TAIL 0.55 s).
Substantially @xcdoge's code (PR #380: the slide, the death timing, research 13 / 14, most tests), authored as theirs.
Changed on review: the corpse no longer goes back to where it stood (the PR's test asserted that; the sim's body is at
the destination); the PR's pre-pass scanned (renderT − 0.5, renderT] after processEvents had taken those events, so
it never fired and the slide started a frame before landing — it now looks ahead to the next snapshot and holds the
view until then; the slide eases into the live sample instead of the fixed destination (no jump back onto an enemy
that walked on); each step uses its mean velocity (the end-of-step speed ran 9 % short, then jumped); the constants'
comments say the timing is ours; research 14 no longer cites a sweep script and a data guard that were not in the PR.
Golden: unchanged (client only).
Tests: test/render/unitview.test.js 35/35 (slide: deceleration, idempotence, facing, snapshots in flight, friction,
√D, corpse lies where pushed, hold-then-ease without a landing jump, a never-shown slide handed back; the app.js frame
order over a real battle: a 1.7-tile push over ≥ 8 frames with every step < 0.35 tiles, a push-kill's body staying at
the destination; death: a dying flyer's lift, every Die clip within DIE_CLIP_MAX and a 7.97 s clip scheduled whole);
test/sim/displace-fx.test.js 1/1; render/interp 16/16; docs tests pass.
Since 0.2.0 the sim stands an unblocked enemy for the rest of its attack clip (ai.js attackStand → atkStandUntil),
but render/interp.js lerped x/y over the whole snapshot gap, so the view slid through a recovery the sim held
(repro: a real 隐形弩手, snapshots 4.9 / 5.0, stand until 4.934 — sampled at 4.917 it slid at vx −0.2).
The snapshot now carries, only when non-empty, `stand: [[id, until]]` (atkStandUntil of enemies alive, deployed,
visible, neither feared nor stunned) and `standCut: [[id, at]]` (the latest time each listed enemy's recovery was cut
or ignored: a displacement, fear, a stun, hiding, leaving — Battle._cutAttackStand, display metadata the battle never
reads). sample() holds the older position until `until` and interpolates the rest (HP / SP keep the whole interval),
stays linear when the newer snapshot records a cut inside the interval, and extrapolates nothing for a newest snapshot
that shows a stand, a death, a stun or a block. Teleports and redeploys still snap first; old snapshots without the
fields interpolate as before.
Substantially @ZiyueW1116's code (PR #381), authored as theirs. Changed on review: the docs (SIM §9, design/network.md)
and code comments in English like the rest of them; the snapshot builds its listed-id set only when an enemy has a cut
to report.
Golden: unchanged — the digest canon hashes neither field and the sim's timing does not move (all 283 scenarios match).
Tests: test/render/attack-recovery-interp.test.js (new, 25: metadata validation, hold-then-slide, cuts, legacy
snapshots, extrapolation guards, real 隐形弩手 / Weedy push-kill / stun / freeze / sleep / levitate / block / fear
battles) 25/25; render/interp 16/16; sim battle, perf, robustness, downelem 69/69.
The 赤霄·天喟 wave is a kit-managed probe (it turns corners and lives as long as the skill), not a snapshot
projectile, and the sim emitted one `dash` at the cast and a `slash` per hit — nothing along the way, so the client
drew no wave. op-chen3.js now emits fx('chen3Wave', { x, y, fromX, fromY, dr, dc, id }) every WAVE_FX_EVERY (0.12 s;
4 ticks), the trail restarting at a corner; render/fx registers chen3Wave → 'qi' (pt: drawn at the event, never on
her) and draws a streak plus a crescent head turned the way it travels. The look is a procedural stand-in, not the
official particle prefab.
Substantially @Convey123's code (PR #382), authored as theirs. Changed on review: the head is projected after
streak() — streak() reuses the effect system's projection scratch, so the PR drew the crescent on the trail's start,
one event behind the wave — and the trail, head and heading share one height; an unused test variable dropped.
Golden: diy-073 (5_diy*=chen3 S3 in two slots) events.fx 508 → 1558, fx.chen3Wave → 1050; diy-074 (6_diy*=chen3 S3)
events.fx 272 → 872, fx.chen3Wave → 600. HP, kills, damage, hooks and snapshot hashes unchanged: battle.fx only
records an event.
Tests: test/content/op_chen3.test.js (+1: the events, their cadence, id, direction, trail start, silence after the
skill) 12/12; test/render/fxchen3wave.test.js (new, 6: the archetype, drawn at the wave not at her, the head on the
wave and turned with it, a sweep joining into one trail, junk input, nothing left after) 6/6; boardart 15/15.
Applied as-is. Captured real battles live in public/dev/perf (including the round-13 联防 on the wide round battlefield), with a browser speed page and a headless benchmark. public/dev stays out of the package. The Chrome benchmark was not run. The numbers in the original pull request are from 0.1.4 and are not quoted here.
Tests: node --test test/perfbench.test.js test/render/battle-perf.browser.test.js — 13 pass, 0 fail; the Chrome suite skipped (RENDER_E2E unset).
Seen (community report 「蕾缪安3技能范围里没人好像会自动结束」, relayed by the owner): with S3 locking, the locked enemy
killed or walking out of her range and nobody else in it, the skill ended at the next lock tick — the shells fell at
once and the bullets left were thrown away (repro on 0255afd: 3 of 5 bullets lost; a walker leaving her range, 1 lost).
Cause: chess_char_6_01-lemuen.js:169 `if (!e) { if (m.lemLocks.length) skill.end('ammo'); return; }`, there since the
initial import with no decision behind it.
Official: 「攻击装有5发弹药，打完后结束（可随时停止技能）」 — an ammo skill ends when its bullets are spent or the player
stops it; PRTS 卫戍协议/帮助 技能操作 「通常不会自动关闭技能」 (only 玛露希尔's 召唤使魔 is closed automatically); the client
(charpack char_4193_lemuen, buff_template_data lemuen_s_3): the S3 lock is an attack that needs a target and the skill
buff switches to the bombardment mode on ON_BUFF_FINISH only.
Now: with nothing to lock the skill waits (bullets and locks kept, the marks following their enemies or staying where
they left), its lock ready, so the next enemy in range (or 通缉) is locked on the next tick; it ends when the bullets are
spent and the shells fall on every mark in lock order as before. The S3 'lock' fx carry `hold`: render/fx/locks.js keeps
those reticles while the shooter's skill bit (UF.SKILL) is on and starts their LOCK_T at the skill's end (they faded 5 s
into a wait otherwise); S2's aim lock keeps its timeout.
Golden — traced on both trees (SkillRuntime start / end, the round's fields, the carried SP); in each the first
divergence is an S3 the parent ended with bullets left and nothing in her range:
- roster-018: the elite 蕾缪安's S3 of 7.63 s ended at 9.63 s with 2 of 6 bullets left; now the 6th goes at 12.00 s
(first differing snapshot window 0–10 s).
- boss-boss_3-pair: her S3 of 27.8 s ended at 30.3 s with 4 of 9 left; now all 9 by 34.43 s (window 30–40 s); the boss
now falls at 199.53 s instead of the 200 s cap.
- coop2-NORMAL-14-standins: ai_0's 蕾缪安 in round 11 ended S3 at 17.2 s (6 of 8 left) and 77.97 s (7 of 9); now she
spends them, enters the 联防 with 25.96 SP instead of 10.27, casts S3 there and kills one more of h_0's leaked enemies:
h_0 loses 1 LP less (rounds 11–12 LP 18 → 19, 8 → 9; lpLost 30 → 29; ai_0 kills +1).
bonds, standins, diy unchanged.
Tests: test/sim/feedback5-lemuen-s3-wait.test.js (3; the killed target and the walker who leaves fail on the parent),
test/render/fxproj.test.js +1 (a held lock waits while the skill bit is on; fails on the parent renderer).
Docs: CHANGELOG 0.2.0 规则（按官方）, docs/history/0.2.0.md §25.22.8 (also notes, unchanged: S2 spends a bullet on every
normal attack where PRTS says only the 通缉 snipe does — left to the owner).
Cause: 引星棘刺 S2's drifting alchemy unit is re-sent every second (position and radius change) and the client drew each
event as a new additive zone lasting the rest of the unit's life — up to twelve layers on one spot; overlapping units
(锡人's charges, several alchemists) added up the same way (render/fx/zones.js: disc alpha 0.28, edge 0.75, ADD blend).
Now: the sim fx carries `key` (kits/ops/chess_char_5_15-thorn2.js) and zones.js updates the zone under that key in place;
zoneAlpha draws a persistent area (> 3 real s) at 0.16 / 0.5 and ÷ √n among the n persistent areas overlapping it (centres
closer than ¾ of their radii summed); short bursts and telegraphs unchanged.
Golden: unchanged (fx events are counted by kind; the snapshots do not carry fx).
Tests: test/render/feedback5-zone-stack.test.js (3, fail on the parent).
- Cause, client: the prisoner skeletons carry three clip sets whose collar light is grey, blinking orange
and red. The official battle prefabs (local client battle/enm_pfb_6.ab: the Graphic component's anim key
-> clip table, the modes' replace pairs, the clip the Spine starts on) draw mode Default (【禁锢】) on the
grey set, mode R (the warning) on the orange set and mode L (【解放】) on the red one. The name-based role
resolver gave 普通囚犯 / 老练囚犯 their unnumbered set — the red one — so they came out of the gate freed;
no FORMS entry existed, so no prisoner ever changed (强壮囚犯 / 拳师囚犯 / 重犯 / 传奇重犯 stayed grey).
- Cause, sim: buff_template_data enemy_confinement[atk_cnt] spends a charge ON_BEFORE_ATTACK (not while
麻痹) and switches to mode R when one is left; enemy_liberty_listener[warning] switches to mode L before
the next attack (PRTS "进行第4次攻击前，切换至解放状态"). archetypes.js prisoner() counted on the 'attack'
hook, after the damage, so the 4th attack still hit confined, and it sent no form.
- Now: prisoner() counts in `before` (an attack 麻痹 interrupts never gets there, PRTS 拳师囚犯), setForm
'warning' (fx 'phase', drawn silently: kinds.js SILENT_PHASES) before attack times-1 and 'liberty' (fx
'liberate') before attack times; 重犯's free-all and 杰斯顿's KO switch the freed ones too.
render/units.js FORMS maps 'warning' / 'liberty' for the six skeletons (普通 / 老练囚犯 *2 / unnumbered,
强壮囚犯 *2 / *3, 拳师囚犯 / 重犯 / 传奇重犯 *_orange / *_red). tools/assets/spine.mjs PREFAB_SPINE_ROLES gives
普通囚犯 / 老练囚犯 their grey *3 set; data/assets.json regenerated (fetch-assets --offline --add-only, JSON
compare: only those two anims and the hash), data/enemies.json too (build-data --offline: only their
attackAnim.clip Attack -> Attack3, same length and strike frame). The 3D board draws the same Pixi views.
- Docs: ASSETS.md (PREFAB_SPINE_ROLES, the prisoner forms), SIM.md (the form fx), design/client.md (FORMS).
- Tests: test/content/feedback5-prisoners.test.js (the six: warning before attack 3, attack 4 hits with
ATK +50 %, the forms reach the client and UnitInfo; free-all; 麻痹), test/render/feedback5-prisoners.test.js
(manifest = the prefab's grey set, FORMS sets, view clips, PREFAB_SPINE_ROLES = the committed manifest) —
both fail on the parent but the 麻痹 pin; test/render/feedback5-prisoners.browser.test.js (RENDER_E2E=1:
three prisoners on the 2D and the 3D board, each model's set follows its sim form).
- Golden: 28 scenarios. Only the warning's new 'phase' fx (the prisoner fell before its 4th attack):
roster-001, roster-003, roster-007, roster-024, roster-033, bond-miraShip-low. The 4th attack now hits
freed, one attack earlier than before, so the fight after it moves: roster-004, roster-022, roster-026,
bond-kazimierzShip-low, bond-raidShip-low, bond-soloShip-low, bond-victoriaShip-low, bond-visiShip-low,
standin-01, diy-024, diy-061, diy-065, diy-066, diy-094, diy-097, diy-104, diy-105, diy-127; only the
per-second snapshots besides the fx: roster-021, diy-017; and the bot matches whose battles field
prisoners: solo-ABYSS-2, solo-HARD-9-boosted.
Ported by hand from PR #211 by @xcdoge:
- render/units.js: FLY_HOVER 0.32 → 1.3, flat (the official fly offset is one model-independent Vector3(0, 0.35, 0) in
the client's character space, whose unit is the standard battle-prefab scale 0.27; an official screenshot measures
1.2–1.4 tiles); enemyModelScaleY multiplies the skeleton's Y scale and the head (bar) height, also in the impostor
path; a mirrorX model flips on top of the direction flip. The body, its HP bar, damage numbers, projectile hits and
skill rings ride `hover`, the shadow stays on the ground; one camera drives the units over the 2D and the 3D board.
- tools/build-data.mjs MODEL_STRETCH_Y / MIRRORED_PREFABS → enemies.json modelScaleY / mirrorX; data/enemies.json
regenerated offline here (the PR's one-line blob was not taken) — JSON compare: only enemy_1112_emppnt /
enemy_1112_emppnt_2 modelScaleY 1.263 and enemy_1196_msfyin mirrorX true.
- tools/local-extract/enemy_model_offsets.py (optional, host-side prefab reader) and
docs/research/12-flying-visuals-official.md (without the claim that the owner said 「都要」 on 2026-10-05 — the owner's
decision of 2026-10-06 is cited; where the quirk enemies appear corrected against 0.2.0's data: 帝国炮火先兆者 is a FLY
special enemy of the mode, 帝国炮火中枢先兆者 comes with the bounty 战术特训·飞行III, 木制瑞印 is in no wave). No
screenshot or extracted art committed.
Tests (from PR #211): test/render/unitview.test.js (FLY_HOVER ≈ 0.35 / 0.27; the same lift for every flyer; a flying
view lifts by FLY_HOVER — added here: its body and HP bar ride the lift, its shadow stays on the ground),
test/render/playtest6-enemies.test.js (the data keys; the 1.263 aspect and bar height; the mirrored model's x sign);
both fail without the renderer change; the 425 render unit tests pass. Docs: DATA §9, ASSETS, design client.md,
research 00 / 12, DESIGN §25.13.5 and the §25.10 table.
Golden unchanged (drawing only; the sim reads neither field).
Ported by hand:
- public/js/ui/gameLogic.js is a re-export shim on feedback5: #185's TERRAIN_TIPS / terrainInfo went into the new
public/js/ui/gameLogic/terrain.js (the barrel gains one re-export line); closesOnFieldPress 'terrain' into
ui/gameLogic/shortcuts.js.
- i18n (docs/I18N.md): every new user-visible string goes through t() with an English entry in public/i18n/en.json
(42, after the 补位 strings): the terrain names / tags / fact table are N_() and translated where terrainInfo hands
them out, the mechanism lines are msgids with their numbers as params (a line with an optional half is two whole
msgids, the list joins use an array param), the card's 地形机制 / 这一格 titles in ui/detailPanel.js. Official EN
terms where the EN client has them (stage_table tileInfo: Active Originium, Mire, Exhaust Grille, Deep Water Zone,
Incursion Point, Protection Objective, Tunnel Entry / Exit, High Ground; Camouflage). The Chinese output is
byte-identical to feedback4's terrainInfo on all 5187 tiles of the 13 stages and on synthetic stages.
- public/js/screens/game.js (MatchScreen still lives there): the import keeps feedback5's names plus terrainInfo /
fieldTile; the terrain stage is the one view.setStage gets, so it includes feedback5's fieldStage — a 联防 field is
played on its own map (GitHub #41), and a tap there explains that map's tile, not the round stage's. The tileClick
handler and render/app.js emitTileClick / ui/fallbackField.js auto-merged in place. terrainInfo reads the stage only,
so a 补位 / 自选 board explains its tiles the same way (their pieces only change a piece's art and name).
- public/js/ui/detailPanel.js: the terrain branch of resolveDetail before feedback5's stand-in / 自选 helpers.
- public/js/screens/title.js (#183): feedback5's LangToggle / t() imports plus GIcon / SettingsModal; the gear
button's aria-label / title are t('设置') (en.json already has "Settings").
- tools/fetch-assets.mjs / tools/assets/downloader.mjs: --add-only (keepExisting, the font skip) beside #24's
--asset-source / SP_ASSET_SOURCE and the shared MirrorPolicy; feedback5's extraOperators / moduleTypes and the
voice plan unchanged.
- data/assets.json: feedback5's manifest plus #134's audio.sfx.battle.leak (after heal, the plan order); stats files
6993 → 6994 and bytes +23 240 as measured on feedback4; the hash recomputed with the generator's contentHash
(c01aa739a2d3 → 6a547fc5c937). A field-wise JSON diff against feedback5 shows only those four fields.
- docs: DESIGN §24 intro and §24.5 / §24.6 as on feedback4 (the rest of DESIGN.md is feedback5's); ASSETS.md,
DEPLOY.md, PLAYING.md auto-merged.
- Auto-merged and checked: audio.js (the sim still emits ['leak', id], server/sim/battle/deploy.js), render/style.js,
textures.js, tiles.js, board3d/materials.js, boardArt.js, tools/assets/*, tools/setup.mjs, extract.py, the tests.
Checks: the PRs' unit tests (audio 22, gameLogic 44, terrain-detail 1, originium-art 4, boardart 15,
board3d-extract 9, asset-network 21, fetch-assets-shrink 4), i18n 9 / i18n-data 8 / i18n-server 4, client-static 313,
docs-consistency 24, kits_layout 6 pass; lint 0 errors (no new warnings), check:imports, typecheck; golden: all 283
scenarios match, nothing regenerated.
Tapping a 活性源石 / 沼泽 / 深水区 / 排气格栅 / gate / teleport tile opens a read-only card with that stage's own terrain numbers (the boss-prep board picks the board tile once, through prepXf.toBoard); the 2D atlas cell and the 3D shader now share one 活性源石 palette and recipe and draw to the tile edge. Contributed by @Convey123.
Only the extraction (tools/local-extract) and the manifest pointer change; no official textures are committed and nothing in play changes. Contributed by @Cloudnyco.
Ported by hand:
- #109 kit hunks → kits/ops (the tier files stay re-export shims): 宴 chess_char_1_18-utage.js, 野鬃
chess_char_1_19-wildmn.js, 砾 chess_char_2_12-gravel.js, 斯卡蒂 chess_char_3_05-skadi.js, 伊内丝
chess_char_4_04-ines.js, 缄默德克萨斯 chess_char_4_16-texas2.js, 耀骑士临光 chess_char_6_17-nearl2.js. The +/- lines
are identical to #109's tier1/2/3/4/6 diffs.
- #109 generic.js against W5D's own-duration passives (一击即退): the deployment window is computed where #109 computes
it, with W5D's case included. Both forms run #109's duration lifecycle (activateOnDeploy, spType 'none', the
skill's own mods), and W5D's 冲锋号令 DP start effect is kept.
- #103 room.js through i18n phase 1: t('{name}邀请你加入卫戍协议：盟约【{difficulty}】') with the translated difficulty
name (the Chinese text is the PR's), plus its English line in public/i18n/en.json.
- data/assets.json: feedback5's manifest plus #72's audio.bgm.combatAlts (after boss, the plan order). files +3 and
bytes +5 037 595 as measured on feedback4; the hash is recomputed with the generator's contentHash. No other data/
file changed.
- docs: DESIGN §5.6 (#109's timed deployment paragraph before feedback5's trigger rules; "Permanent passive skills are
always on"), SIM §7.4 (both generic passive cases).
- Auto-merged and checked: #149 bond-bar fold (still in screens/game.js), #160 skills.js / spine.js, #109 skills.js /
snapshot.js, #66 addon/battle.js (restTile / isDown live in sim/battle/tiles.js), #150 units.js, #72 audio.js /
plan.mjs, #74 plan.mjs / docs.
Where the PRs met feedback5 work:
- 突袭 (W5L, GitHub #49): raidPoll counts a skill that needs no SP while it is on. That is a passive, as before, or a
#109 deploy-timed skill while its window runs. Once the window has ended only the idle trigger is left. Without this
change 缄默德克萨斯 / 宴 lost #49. research 02 §3.18 says so.
- Tests changed to #109's representation: bonds_addon #49 test (kind 'duration', activateOnDeploy) plus a new test
(window over → idle trigger only), and standin.test.js's 预备干员 PASSIVE branch (the skill's mods instead of the
generic:passive buff).
Golden regenerated on the merged code. Attribution: each PR's gameplay change was reverted alone in a scratch copy, and
all three together. With all three reverted the digests equal feedback5's exactly. Every moved scenario was replayed
and fields the unit that explains it.
- #160 only (0.5 s skill events of deploy passives; events.skill only): roster-001/006/025/049, bond-arcaneShip-low,
bond-emptyShip-high, bond-investShip-high, bond-kjeragShip-low, bond-siracusaShip-low/high, bond-victoriaShip-low,
bond-yanShip-high, boss-boss_2/4/6-solo, hidden-boss_8/10-solo, standin-03/05/08/10, diy-005/006 (普罗旺斯 S1,
琳琅诗怀雅 S1/S2, 星熊 S2, 凯瑟琳 S1, Misery S3 as stand-in / 自选 prototype).
- #109 only (deploy-timed skill state and lifecycle; 突袭 技能就绪 limited to the window; generic 一击即退):
roster-003/004/005/021/023/024/039/041/042/046/047, bond-egirShip-high, bond-indomShip-high,
bond-kazimierzShip-low/high, hidden-boss_8-pair, standin-01, coop4-FUNNY-6, coop2-NORMAL-8-serverrun.
- #66 only (灵巧's aura stays only for a killed member, from its body tile): roster-028, coop2-NORMAL-3,
coop3-ABYSS-5, coop4-ABYSS-7, solo-HARD-9-boosted.
- #109 + #160: roster-002, roster-029, bond-investShip-low, bond-raidShip-low/high, bond-suntShip-low,
bond-visiShip-high.
- #66 + #160: roster-008, roster-027.
- #66 + #109: roster-022, roster-040, bond-indomShip-low, boss-boss_2/4/6-pair, hidden-boss_10-pair, standin-06,
coop2-HARD-4, coop2-NORMAL-14-standins.
- all three: roster-010, roster-012, roster-031.
69 scenarios. Every other digest is unchanged.
sganggs