Allow Sapho order behind a started turn

This commit is contained in:
Quintine committed 2026-10-25 23:33:04 +11:00
1 parent 3bf7d26e80
commit 8e6602633d
17 files changed
+1121 -62

No files matched your search

+73 -1
View File
@@ -13,6 +13,72 @@ contract supersede the earlier absence of that mode below. The
and a fixed public final opportunity in fresh Richese battles; old battles keep
their early protocol.
## Order after a started combined turn — 25 October 2026
**Prototyped / Partial.** The printed card permits first/last in a turn-ordered
phase/action. The [source review](JUICE_OF_SAPHO_SOURCE_UPDATE.md#movement-firstlast)
finds no entitlement to split an already-started combined shipment/movement
turn. Under the 14 October choose-recommendation authority, adopt its bounded
recommendation: retain that current actor and reorder only later unstarted
turns. The current actor finishes normally; the changed tail then takes effect.
This is visibly provisional online timing, not a new publisher clarification.
An unstarted holder at a clean boundary after another player's actual shipment
or movement may go first among the tail, or last including after Guild. It
cannot split its own begun turn, displace an active Guild, replay a completed
turn, reset shipment/movement counters or intervene during unresolved card,
arrival, payment or movement preparations. Existing last protection and no-op
rules remain. Advanced Guild receives its native next-boundary choice; moving
first in the tail does not consume or suppress that privilege.
The pure [movement-order quote](../game/sapho-movement-order.ts) binds the
original phase event, current actor and whole remaining public order in the
queued action event. Its returned last protection retains the original
`movement:turn` event. The owner-only option adds display metadata `afterCurrent`;
human and minimal bot actions explicitly send only card/scope/event/mode.
Persistent controls name the actor who must finish. Existing clean options and
physical storm order remain. Semuta's new optional continuation field keeps
that same current actor through the original discard claim/pass; absent-field
legacy continuation/signature bytes retain their original path.
This supersedes only the earlier blanket partly-completed-turn exclusion for
an **unstarted different holder** behind the current actor. No split-turn mode
or new ordered phase is invented. Wider combinations, full modes and final
assurance/deployed acceptance remain open.
Frozen **11 files, 133/133**, types/lint/build and eleven actual engine/
GameTable SSR/isolated button-callback journeys pass on unchanged source. New
helper8, controls6 and engine14 cases cover native Basic/Advanced first/last,
actual paid current shipment, unchanged current resources/counters, native
Guild Take/Wait, old-event/no-op/active-holder rejection, exact metadata-free
actions, all four minimal profiles and Semuta claim/pass.
Seven native journeys genuinely earn the cache1 purchase, then ship one actual
counter before intervention. Observed source prefixes are Basic46/Advanced53;
the original granted Guild source55 pays one spice for its native shipment.
All original four turns finish once in8–13 continuation actions and retain45
physical cards. Four separate Basic/Advanced Homeworld supplements inherit
explicitly **controlled conserved Sapho/Semuta custody**, rather than earning
those cards. They then make a real paid shipment, stage the used Sapho and
claim/pass through the exact current-turn continuation. Their source counts
are recorded per journey; no native acquisition is claimed for that supplement.
First related proof125/133 reveals a real preparation-only guard omission and
legacy blanket-started expectations. A Hajr/preparation without actual shipment
or movement remains blocked. Revised old tests retain stale original-quote
rejection, actor/counter/grant identity and cost preservation. A cyclic test
finishes the existing public discard response before its next bid. Test typing,
an unused import and a redundant copy are repaired without relaxing legality.
Private frozen report: `/home/quintine/.local/share/dune/checkpoints/20261025-sapho-started-lint-fixed/report.json`;
journeys/HTML: adjacent `20261025-sapho-started-runtime/`. No saved rooms,
databases, server, resets, restarts, deployment or recovery campaign is touched.
Broader combinations, complete modes, hydration/browser/mobile/deployed and
final assurance remain due. The engine/SSR evidence does not certify them.
Evidence: [native and labelled supplemental sources](../tests/fixture-sapho-started-movement.ts),
[engine cases](../tests/sapho-started-movement-engine.test.ts),
[quote cases](../tests/sapho-movement-order.test.ts),
[controls/callback cases](../tests/sapho-started-movement-controls.test.tsx).
## Extortion payer order — 18 October 2026
First/last now has a provisional adapter for native Moritani Extortion's unanswered
@@ -196,7 +262,13 @@ The [battle-order follow-up](SAPHO_BATTLE_ORDER.md) connects first/last at a cle
boundary between actual battles. It does not implement the separate aggressor
mode discussed below.
Later battle aggressor intervention, other ordered actions and partly completed movement remain unfinished. The current-lot cyclic prototype above supersedes the earlier normal/Black Market absence claim. First/Guild competing priority is guarded. Reacquiring and replaying Sapho while an existing movement protection remains is also guarded. Silent submission is simultaneous and receives no fabricated order or tie change.
Later sections supersede the earlier early-only aggressor, cyclic-auction and
blanket started-movement exclusions. Queued movement now supports an unstarted
different holder behind the active turn under the selected provisional policy.
Other ordered actions, splitting the holder's own started turn and competing
first/Guild precedence beyond the retained native choice remain unfinished.
Reacquiring and replaying Sapho while existing last protection remains is still
guarded. Silent submission stays simultaneous with its original storm tie order.
The battle audit in [the battle review](JUICE_OF_SAPHO_BATTLE_REVIEW.md) found a real Stone Burner counterexample: a plan admitted with original aggressor tie priority can become allocation-ambiguous after late Sapho. A complete implementation needs the actual allocation continuation (and a usable, possession-independent late intervention boundary), not rejection based on an opponent's hidden plan. No aggressor handler has been enabled around that defect.