fix: 重置松果重新部署后的S2攻击成长
This commit is contained in:
1 parent
1db8e51023
commit
8ebdcf5d17
5 files changed
+120
-3
No files matched your search
@@ -1561,3 +1561,12 @@ yet — enemy content must pick one); 抵抗 covers the control statuses of `RES
|
||||
“余音” owns a same-call pulse queue (DESIGN §28.24): reflected hits enqueue their earned pulses instead of recursively
|
||||
entering another pulse. This retains the lethal-hit pulse and prevents valid high hit-count chains from tripping the
|
||||
general hook guard. The queue is emptied in `finally`; it introduces no timer or simulation RNG.
|
||||
|
||||
|
||||
### Per-deployment skill counters
|
||||
|
||||
An operator-specific ramp that resets on deployment uses its own `unit.mem` counter, cleared by its `deploy` hook.
|
||||
Do not reset `SkillRuntime.activations`: it is the cumulative sequence used by skill lifecycle/reentry guards.
|
||||
松果 S2 uses `pinecnS2Uses` this way (PRTS 松果 S2 备注); each new deployment starts at the first blackboard step,
|
||||
while further casts within that deployment advance and cap normally. S1 and her deployment SP-recovery talent keep
|
||||
their existing behavior.
|
||||
@@ -224,3 +224,9 @@ the hook-depth guard on scaled hit-count HP; the guard remains unchanged for oth
|
||||
as well as by its aura tick, including the first hit and immediately after the last companion leaves.
|
||||
|
||||
Sources and the feedback items requiring no simulation change are recorded in [§28.25](../history/0.2.3.md#2825-community-feedback-of-2026-10-10).
|
||||
|
||||
|
||||
**Pinecone S2 (松果 电能过载).** The ATK ramp counts casts since the current deployment and resets whenever she
|
||||
is deployed again. Its counter belongs to that unit's kit; the engine's `SkillRuntime.activations` remains a
|
||||
cumulative activation sequence. Source: [PRTS 松果](https://prts.wiki/w/松果), S2 备注; rank-specific steps are read
|
||||
from `data/chess.json` blackboards. No mode rule carries this ramp through a new deployment.
|
||||
@@ -479,3 +479,16 @@ imports and typecheck pass. This round changes no production UI and does not rer
|
||||
the `roster`, `bonds`, `fields`, `matches`, `standins` families are unchanged.
|
||||
- **Contracts updated:** design/engine.md, SIM.md (cabin), design/client.md, design/network.md, META.md and PLAYING.md
|
||||
(local identity recovery and handshake).
|
||||
|
||||
|
||||
### 28.28 Pinecone S2 resets its ATK ramp on deployment — `kits/ops/chess_char_3_10-pinecn.js`
|
||||
|
||||
- Cause: the kit selected its ATK step from `SkillRuntime.activations`, which counts all activations in the battle;
|
||||
the first cast after automatic redeployment therefore used a later step.
|
||||
- Source: [PRTS 松果](https://prts.wiki/w/松果), S2 备注: the count starts from this deployment. Rank-4/rank-7 steps
|
||||
come from `data/chess.json` skill blackboards. The mode's redeployment/carry-state rules do not preserve this count.
|
||||
- A unit-local S2 counter is cleared in her existing deploy hook. Growth within one deployment and the fourth-step
|
||||
cap remain; S1, range, trigger, talent and the global cumulative activation sequence are unchanged. No new
|
||||
[ASSUMED] rule. This touches only Pinecone's kit, not 凛冬's separate deployment lifecycle (issue #315).
|
||||
- Regression: `test/content/kits_pinecn.test.js`, normal, elite without module and elite RPR-X; ramp/cap, later
|
||||
paid redeployment, natural enemy knock-out and two copies with independent counters. These are model tests.
|
||||
Reference in new issue
Block a user