Quintine
af360c836b
Implement both rescue bands with legal physical casualty routing, owner controls, all four bot profiles and durable battle receipts. Preserve paid support, elite identity, killed/returned leader custody and replay rejection. New Advanced Atreides skill choices exclude Suk while the KH loss-count ruling is pending; legacy assigned combinations retain their preflight guard. Public gates remain. Independent review fixed stale dead trainer assignments and detached rescue receipts. Final focused Suk/Leader Skills union: 46 passed. Typecheck, lint, 4,452 offline tests, production build and 43 HTTP tests passed on unchanged source tree 3b76239a46d0ba874299495e64295a5eaaa8f2415c3e9a041bee944bbce50c5a. Actual browser rescue and refresh retained counters, private hand and pending winner card cleanup. Two four-profile Basic/Advanced smoke games completed with JSON restoration and no rejected actions; neither exercised a Suk rescue. After the outage report the server was healthy. All 360 pre-check rooms and original seat/recovery records are unchanged; checks added 33 rooms. SQLite integrity and foreign-key checks pass. No server restart or database reset.
101 lines
6.1 KiB
Markdown
101 lines
6.1 KiB
Markdown
# Suk Graduate rescue prototype
|
||
|
||
13 September 2026. **Prototyped / Partial**. This connects both Suk Graduate
|
||
bands to real battle resolution, player choices, AI and saved continuation.
|
||
Complete module, faction and publication gates remain closed.
|
||
|
||
## Source contract
|
||
|
||
Use the physical card and the [Leader Skills source archive](LEADER_SKILLS_RULES.md).
|
||
The [GF9 CHOAM & Richese rulebook](https://www.gf9games.com/dune/wp-content/uploads/2021/11/CHOAM-Rulebook-low-res.pdf)
|
||
page 9 governs visibility, survival and skill-before-faction ordering; its
|
||
page 12 Q&A gives a captor only the lower band of an actually used captive.
|
||
The [designer walkthrough at 11:10–11:59](https://www.youtube.com/watch?v=XT_azRVLq_0&t=670s)
|
||
expressly distinguishes the normal and selected-leader rescues and says they
|
||
do not accumulate. The archived card photograph supplies the precise quantities.
|
||
|
||
After an ordinary win with physical casualties, a face-up native Suk Graduate
|
||
returns exactly one chosen casualty to reserves instead of the Tanks. This band
|
||
is mandatory. When the actual skilled leader is used and survives, the player
|
||
may instead save zero through three casualties, limited by the raw casualties.
|
||
If any are saved, exactly one stays in the battle territory and the rest return
|
||
to reserves. Declining this lower band does not trigger the normal band.
|
||
|
||
A concealed skill with another selected leader grants neither band. A killed
|
||
selected skilled leader grants no rescue. A different selected leader may die
|
||
while the face-up native trainer remains alive and grants the normal rescue.
|
||
A traitor victory loses no winning forces; mutual traitors and explosions have
|
||
no winner. None creates a Suk rescue. The zero-dial case also creates no window.
|
||
|
||
## Physical settlement and continuation
|
||
|
||
The existing legal raw normal/elite casualty choices are retained. Choose that
|
||
allocation first; Suk then redirects only those actual counters. Paid support
|
||
is unchanged. Saved elites remain elites in reserves or on the board. The
|
||
normal band resolves automatically when only one counter type can be saved;
|
||
the lower band's optional choice remains with its owner.
|
||
|
||
The receipt freezes revealed eligibility before the live battle is cleared or
|
||
a used captive returns. It records the battle, raw commitment, source groups,
|
||
physical pool and remaining card cleanup. A separate obligation in the last
|
||
battle context prevents a missing receipt from silently bypassing rescue.
|
||
Actions, views and automatic continuation validate it. A stale event, altered
|
||
allocation, changed physical pool or missing choice rejects without mutation.
|
||
The server alone projects the legal rescue options; the receipt is not exposed.
|
||
|
||
The current engine assigns a raw casualty mix across sectors in existing
|
||
location order. A saved counter that remains stays in that original sector;
|
||
there is no free sector relocation. This is an explicit implementation inference:
|
||
the card specifies a territory, not how to select sectors. Saved counter types
|
||
and any eligible sector choice remain visible to the player.
|
||
|
||
Suk settles before winner card choices and faction aftermath, including capture
|
||
and Face Dance. Mandatory loser-card cleanup uses its existing continuation.
|
||
The exact ordering against future combined optional card interruptions still
|
||
needs integrated acceptance. Homeworld custody, Nexus and expansion-faction
|
||
combinations are explicitly rejected before resources or traitor votes commit;
|
||
the genuine Leader Skills initializer currently admits base factions only.
|
||
|
||
## Atreides loss-count question
|
||
|
||
The base Advanced rules activate Kwisatz Haderach after seven force losses.
|
||
Suk describes rescued counters as losses redirected away from the Tanks, and
|
||
the designer describes the retained counter as having been killed. Counting
|
||
the raw dialed casualties is therefore the stronger textual inference, but
|
||
the retrieved publisher/designer sources do not expressly settle this combination.
|
||
The user question is pending. New Advanced Atreides setup and revival offers
|
||
retain their physical cards but make Suk unavailable, explain why and allow the
|
||
other offered card. The server, private view, controls and bots share this
|
||
prototype restriction. Already-assigned saved combinations retain the battle
|
||
preflight guard before final traitor votes or resources commit. This availability
|
||
restriction is not an official rule. Other rescued counters are kept out of the
|
||
physical Tank and ordinary battle-loss totals in this bounded prototype.
|
||
|
||
## Evidence and remaining work
|
||
|
||
Focused regressions cover genuine module setup followed by conserved targeted
|
||
battles; both bands, no stacking, decline, killed leaders, captured-leader return,
|
||
normal/elite routing, unchanged support and all four bot profiles. JSON and
|
||
SQLite store restart retain the pending choice. Concurrent submissions commit
|
||
one rescue; the stale submission changes no counters. Component checks cover
|
||
clear destinations, required selection and disabled controls.
|
||
|
||
Fresh browser room `BD33WTTV` used genuine Emperor/Guild module setup and selected
|
||
Suk Graduate/Hasimir Fenring. A separately staged, conserved battle position used
|
||
one physical Baliset from the deck. Actual browser rescue saved three ordinary
|
||
casualties: one stayed in sector 10 and two returned to reserves. The result was
|
||
two board forces, seventeen reserves and one Tank. Refresh after another reported
|
||
outage retained that result, the private Crysknife/Baliset/Burseg and winner card
|
||
cleanup. This was a targeted battle, not a naturally reached full game.
|
||
|
||
Four-profile samples completed Basic seed 20260918 (92 actions, three JSON
|
||
restorations) and Advanced seed 20260919 (188 actions, six restorations), with no
|
||
rejected actions and conserved cards/forces after each action. Neither sample
|
||
encountered an eligible Suk rescue; effect evidence comes from the targeted
|
||
tests and browser journey. Checkpoint-wide results and final preservation are
|
||
recorded in the commit and source-bound private report.
|
||
|
||
These tests do not certify the full expansion, every combination,
|
||
AI strength ordering or complete human games. Seven other skill effects remain
|
||
missing; retain the [full scope and pending rulings](CURRENT_STATUS.md).
|