Hold the Flood — Colony and Persistent Progression

COL101 · r2 · 2026-09-06 UTC · Colony and Progression Designer · Submitted for PM review. Responds to PMR004: separates baseline docks from an additional berth; reconciles first-visit preparation and revisit reissue with D023/EQP101; limits observation to accessible evidence; specifies a visible, usable refuge investment and an unpowered return. PM selected the structural direction, not every rule or price below. No task acceptance or development authorization is claimed.

Supersedes COL101 r1, a1046d. Sources and exact revisions are in §8. This page owns Colony proposals; Lead remains the sole integrated Design Document owner under D014. The attributed PM seed at the foot is historical source material, preserved verbatim.

COL101 recommendation

Build a refuge that remembers what the player has made possible, with one reusable support assembly assigned to one advanced service for the next expedition. Banked salvage buys permanent repairs and fittings once. Moving that assembly between an additional berth, workshop tooling and comparative observation is free at the refuge. The competition is which opportunity to support on this trip, not a tax for changing one's mind.

Selected drafting direction: Brief D022–D024 selects seven external districts plus a refuge, no forced refuge defence, eight rack cells from the start, shelter-selected pickup priority, one active advanced service, reusable refits and revisit-only recorded-rack issue capped at six occupied cells. These are PM's provisional next-draft decisions; no district count, duration, owner fiction choice or production scope is approved. This revision does not inherit old six-cell or third-row purchase examples.

Colony proposals awaiting integration: exact repair prices, the additional-berth exception, first-visit packages, the visit/recording triggers, manual workshop continuity after Stop and the refuge geometry below. Equipment's eight-cell migration does not settle electrical durations, stun rules or Arc geometry. No example here depends on those disputed effects.

The credible alternative is to make every repair always active. That gives uncomplicated accumulation, but eventually removes the colony decision from departure. One assigned service keeps a repaired colony useful without requiring recurring payments or progressively more currencies. Its risk is domination by one service; the three strategies below expose the choices for written review, not evidence of equal balance.

Integration against the Campaign submission that arrived during COL101

Campaign's topology and district count belong to Campaign and Districts. A colony repair never substitutes for operating a canal, travelling to a place or learning its machinery. D022 resolves r1's compulsory Refuge-defence objection for the next draft: the refuge is not an obligatory punishment expedition. D023 and Equipment's Build Paths response persuade me to use eight cells immediately and bounded first-visit packages: retaining field growth through issuance limits is clearer than selling the missing rack area.

1. Outcome to choice to changed expedition

1. Resolve the expedition. Extraction banks carried salvage. Rescue loses that trip's unbanked salvage and field finds; it retains observations and inspected recipes. Named world commitments follow their authored commit points, not a blanket reset or blanket persistence rule.
2. Read what changed at the refuge. A route board distinguishes confirmed access, the selected baseline dock, any supported additional berth, observed district settings and stale forecasts. The player can also walk into repaired spaces. A repaired workshop's open work bay remains visibly and functionally different from its shuttered starting state.
3. Choose a permanent investment, or keep the salvage. A displayed one-time price competes with another repair. Capital spending is not refunded, but the repaired capability remains available under its stated power/access conditions. Every preview names the later action it enables, prerequisites and what it cannot do. No repair is required for a free basic departure, ordinary return or essential clue.
4. Choose this expedition's support. Assign the assembly, choose the baseline dock, set pickup priority and prepare the rack at the refuge. The departure preview shows actual occupied cells, empty batteries, route availability and which advanced services are inactive. All choices are revisable until departure.
5. Use the changed opportunity in the field. A second extraction option changes a retreat; a known module allows a different approach to an existing obstacle; a comparison record lets the player act on a discrepancy rather than memorise two instruments under pressure. Field action remains necessary.
6. Return with a new reason to choose. Newly opened physical access, a discovered module or a changed district setting can make a different assignment worthwhile. Failure need not bring salvage to count as progress: a retained recipe or a corrected inference can change the next departure for free.

The assembly is a capacity rule, not cargo. It cannot be duplicated, dropped or lost on rescue. Switching away from tooling before departure revokes its extra issued modules; switching away from berth support revokes the extra berth. There is no preparation stockpile through which the player can combine all three services.

2. Minimum colony and resource model

Only banked salvage is a persistent spendable resource. Charge remains expedition-local, starts empty and cannot be sold, banked or used to pay settlement expenses. Recipes, observations and access states are records, not currencies. The reusable assembly is provided with the refuge; it creates a choice only once more than one fitting is available.

All prices below are four salvage bundles once per named repair/fitting, retained as a legible r1 accounting hypothesis. This is not a target number of hauling runs. If a campaign cannot fund meaningful alternatives from ordinary exploration, revise costs/content allocation; do not add repeated hauling to achieve a duration.

The three base repairs remain useful independently. The advanced fittings change which combination can be exercised on a later visit; they do not buy increasingly large versions of the same numerical bonus. A player may finish investigations with no advanced fitting. No population, food, morale, housing tier, staffing schedule or autonomous resource producer is needed to make these choices work.

3. Three development strategies

Early/middle/late describe increasing access and knowledge, not an approved district sequence, required repair order or hour count. Each strategy can change direction without demolishing its earlier investment.

Illustrative sequence: a failed workshop-first expedition loses its Coil find but retains the inspected recipe. The player can issue a Coil next time without another salvage run, then discover the upper-canal opportunity. Banking salvage later permits forward fittings; assigning them sacrifices advanced circuit preparation for a second extraction. After recording a district discrepancy, the player may instead use comparison on a revisit. These are written consequences of the proposal, not played results.

4. Three connected examples for the domain owners

Route example — for Campaign, with Lead interim

COL-I1, position to Campaign: one baseline held dock and an additional supported berth must be two explicit choices. Under Campaign 9d3590, leaving L2 open at extraction connects the upper canal; the Tram summit must also have been physically reached/surveyed. Rescue alone must not falsely satisfy an extraction-conditioned canal commitment. Show canal state and summit eligibility before offering the choice.

With the canal available, the player can select the summit as the ordinary one held dock, giving up the depot as the held dock. This requires the ordinary reusable outlet refit, not advanced forward fittings. Alternatively, assign forward support to keep depot as baseline and summit as one additional berth. This is an explicit proposed exception to the shared contract's one-dock limit, not an inference that D023 approved two docks.

The latter lets a player enter below, tow a crowd using the Tram's own machinery, investigate above and extract at the summit without retracing that crowd. It costs the advanced assembly assignment that could have supplied tooling or observation. It does not power the rail, hold L2 open remotely, suppress enemies, transport a boat through a closed canal, or open Exchange access from a menu. Baseline arrival/return must remain possible without purchasing this option. Campaign owns the exact physical approach and return paths; the Colony preview cannot claim them complete on Campaign's behalf.

Why prefer optional additional extraction to a mandatory remote outpost upgrade? The player can knowingly choose a harder return to support a different experiment. A compulsory upgrade makes salvage a gate on exploration and erases that choice.

Preparation example — for Equipment

COL-I2, response to Build Paths a9a288: I support its two first-visit packages because each enables a concrete combination while retaining field acquisition. Equipment owns item footprints, primary effects, adjacency and link legality; this page owns eligibility and persistence. All issued batteries are empty.

These are footprint examples, not electrical-effect demonstrations. For the six-cell revisit example, a legal candidate 2×4 arrangement in the cited rules is top row empty / Coil / Coil / Coil-linked battery, bottom row empty / Insulation / Arc / Arc-linked battery. Each battery has a specified linked module; the two empty cells do not touch a battery. Item-rule reconciliation remains Equipment/Lead work.

Visit trigger, proposed here: actual entry into a district marks it visited when that expedition ends by extraction or rescue. Merely selecting it on the map does not. No successful clear, mandatory note, salvage quota or boss is required. A short retreat can therefore qualify a later revisit; that is an acknowledged cost of protecting unsuccessful exploration. If this trivialises first-visit growth, first adjust the reissue package/cap rather than add a success tax.

Recipe and record trigger: personally inspecting an item retains its recipe even if it is lost later. At any shelter, including a field shelter, the player can record an actually assembled, legal arrangement of known items. The record survives losing those physical items. At the refuge it can be reissued only under the destination and issue limits above. An arbitrary unknown template grants nothing. The six-cell cap counts every issued occupied cell; the default battery is included in the arrangement, never silently added afterward. Cargo and charge are not recorded. A larger saved pattern must be edited to the cap before issue; do not silently delete a module. The remaining two cells are room, not reserved cargo, and can receive field items.

Patterns and recipes remain when tooling is unassigned; extra issue does not. This avoids free stockpiling while making a failed experiment repeatable. Full-rack issue on a first encounter is a credible convenience alternative, but consumes the discovery room before the player has met that district's problem; prefer these bounded packages.

Discovery and enduring consequence — for World

COL-I3, response to World 9aab5e: the comparative reader preserves two records of signals the player can otherwise see; it does not discover trunk permission, supply or a mystery answer remotely. At the selected relay’s fixed reader, choose, for example, a visible incoming permission pulse and a local drum event. This is a supported site, not an extra carried module or an unattended remote recorder. Recording occurs only while those instruments are actually observed. The notebook labels place, observed state and order; comparing with an earlier trip never asserts a simultaneous measurement.

The player can put those records side by side at the refuge and choose to inspect a local relay or Collector intake next. The cost is using observation support and spending attention reaching/reading instruments while exposed. The unaided player can perform the same comparison using ordinary notes. World r21's redundant essential evidence is the right boundary: buying this convenience must not create an essential clue that is absent in the field.

If permission flashes but supply is empty, retain both the flash and any personally observed hollow-arrival cue; do not report that no enemies can exist. After Shut/Fix/Hold or an Exchange/Head-Works commitment, mark conflicting older forecasts as historical until a new observation supports an update. Persistent notes inform an action; they do not unlock it by flag.

COL-I4, lasting refuge consequence and position to World/Campaign: workshop investment opens a raised work bay the player can enter and use. The hand-fitting bench continues ordinary one-module issue after the powered jig goes dark. Its shutter remains open, inspected patterns remain available and the installed jig visibly stops. The repair buys lasting usable infrastructure; the advanced service remains contingent on world power.

Proposed refuge layout contract: a raised landing and dry walkway connect free basic equipment, notebook and manual bench without any pump or upgrade. The powered lower apron provides the convenient pumped access; its water level and inactive pump make power loss visible. Losing that convenience changes the approach within the refuge, while the raised landing remains usable. No district is declared to possess this geometry by this Colony proposal: Campaign must identify each actual unpowered return route.

This gives investment an enduring physical result beyond a menu and makes the world decision costly without destroying earned knowledge. It requires no new character simulation or special emergency currency. The work bay and landing are proposed scene requirements, not confirmed assets.

Direct correction to World r21: its claim that the dock pump “runs on the salvage already banked” and develops a falling reserve after Stop introduces mandatory upkeep precisely when the player investigates an unsuccessful or uncertain world decision. It is not the Colony baseline, and normal-operation exemption does not make the post-Stop charge harmless. I recommend no salvage-fed pump reserve and no trip-count refuge depletion. Paid repairs, banked salvage, notes and the dry landing persist; powered forecast, extra pumped berth and advanced jig can become unavailable.

World's proposed three-expedition city reserve also differs from Head-Works r3's immediate grid shutdown. I proceed on immediate loss of powered colony services with manual basic refuge access, pending World/Campaign/Lead reconciliation. Immediate visible power loss gives Stop a readable cost; a reserve clock adds ambiguous failure/return accounting without improving this colony choice. This is my integration recommendation, not a global power ruling. Stored portable charge/effects remain governed by the shared equipment rules; no free charge or unexplained backup generator is introduced.

For Keep, retain the actual committed governor setting on the route board as an enduring district state, with consequences authored by Campaign/World. Revising it requires the local field action; the refuge board is not a remote governor. Keep does not schedule upkeep visits. For Leave, retain the observation and the unchanged actual state; the board must not depict an intervention that never happened.

5. Persistence boundaries

COL-I3 persistence clarification: World/Lead must distinguish an immediate committed relay act from a “leave L2 open at extraction” outcome. A warning before an enduring action must say whether rescue can undo it; no generic “all field state resets” sentence should erase a stated permanent consequence. Colony stores and displays the result; it does not choose every district's commit or reversal rule.

6. Assets, people and scope

Minimum settlement: one refuge containing existing dock/receiver/workshop functions, one route/notebook display and one reusable support assembly represented through the fitted station state. Minimum new functionality in the design is a manual bench, a dry unpowered landing contract and clear station active/inactive states. No named recruit, staffing minigame, resource production, hunger, housing or mortality mechanic is required.

Prefer reuse of project-appropriate free existing dock, industrial bench, shutter, pump, hose, instrument and signage assets if found and licensed. Availability and suitability have not been established. Raised refuge geometry, readable inactive/active states, two-signal records and the manual/powered workshop distinction are unapproved asset/presentation gaps. Existing assets may fail to express them; that is a reason to simplify or seek prior owner approval for rare minimal authored assets, not permission to make them. No purchase or generated art is proposed or authorized.

Scope is campaign-spanning choices through the expedition machinery, not a second management game. No count or price proves duration, fun, feasibility or balance. All scenarios here are written illustrations.

7. Integration proposals and requested responses

This is a focused response and recorded handoff, not a claim those authors have read or agreed. No new owner approval is needed to continue written reconciliation. Development and any asset expenditure remain behind the owner's explicit approval gates.

8. Source revisions used

Processed on 2026-09-06 UTC, against first submission a1046d and task evidence 83a002. The r1 ledger remains available in history. This response processes the PMR004 Colony feedback and later Equipment/World/Campaign interfaces; it does not infer whole-campaign acceptance from labels or timestamps.

Deliberately superseded: r1's conditional six-cell examples, discussion of purchasing a third rack row and still-pending first staffing acknowledgment. D023 and PM's substantive submission review replace those assumptions. The one-time costs remain hypotheses; the old recurring dock-switch payment survives only in the preserved PM seed. The full r1 and its rationale remain at the linked revision, not as a second live Colony rule set.

PM seed — retained source material from before COL101

PMU004 · r1 · 2026-09-06 · PM-seeded source summary; intended owner: Colony and Progression Designer (proposed); Lead interim · Awaiting author validation/transfer. Source: Design Document r13 / 50046f, section 10. This records the existing proposal, not an accepted expansion design. The cited source owns its detailed rules until the author explicitly adopts this destination. Design Index · Expansion Brief · Glossary.

Current design, before expansion

Map, notes, observed rules and recipes persist. Banked salvage, the selected dock and colony repairs persist. Enemies, found expedition modules, carried cargo, charge, field state and panel state reset. The current rescue/failure rules preserve previously banked progress and knowledge while losing current expedition cargo and finds.

There is no mandatory upkeep or off-session decay in the drafting baseline. This protects an investigative expedition that brings home knowledge rather than cargo. It does not prevent substantial colony management from being designed.

Expansion assignment — COL101

Define a colony loop that matters through the larger campaign: what expeditions bring home, who or what uses it, what the player chooses between, and how that choice changes a later route, expedition preparation, discovery or build. Supply three materially different development strategies and concrete early/middle/late choices, including their costs and recoverability.

Avoid three automatic upgrades stretched across eight districts, repeated resource payments used to inflate hours, or an unrelated management game. People and facilities can be proposed where they create a useful decision, but their behavior and asset burden must be explicit. Preserve the no-mandatory-upkeep baseline while making a positive case for deeper management; propose any change to that baseline with its effect on exploratory failures.

Coordinate travel/outpost functions with Campaign, unlocks with Equipment, and persistent mystery consequences with World. Lead reconciles the overall economy. No production assets, colony simulation or playtests are authorized.