commit d71b9b

Commit `d71b9b`

2026-09-07 09:20:19 Anonymous: The Exchange r2: implement PMR007 section 1 and REV104-16/17/18. The r1 expedition was not legal and is REPLACED rather than patched - it lit a lure from batteries that start empty, held it for five minutes when a full bank buys a hundred lit seconds, then had a crowd spread while a reachable lit lure existed, and drew a rack five over three. r2 supplies a legal 2x4 rack; earned charge from cutting the hall's standing stragglers; an explicit per-minute ledger with both banks tracked and an earned 14 / spent 10 / lost 4 / carried out 0 total; the lure's real hundred seconds and what actually retargets when it expires (nothing - b1 is still lit beside it); the Grounder moved to the second surge so the spread it causes is legal precisely because the lure has run out; and two free cells for the retrieval, because the beacon and its battery left the rack when they were planted. Fixes the flash-location convention: this district has one way in, so an intake marker at the landing flashes and the three beacons are lures only, which is a better statement of the two-cable distinction than r1 had. Sets the bank's found state at three engaged clutches on Pumping, Lock Flight and Tram, which is what REV104-18 needs to explain three already-announced districts, and explicitly declines the seized-Pumping-clutch candidate because it would remove the campaign's best worked example of a later discovery changing an earlier return. States two dispatches for a ten-minute district and asks Lead to confirm. Corrects the cable reel to Equipment's U6, a bridge between exposed power terminals.

`hold the flood/districts/the exchange.md` ..
@@ 1,6 1,6 @@
	# Hold the Flood — The Exchange

-	CAM102 · r1 · 2026-09-06 · Campaign and District Designer · In review; proposal. Nothing here is an owner approval, a validated number, or tested behaviour.
+	CAM102 · r2 · 2026-09-07 · Campaign and District Designer · In review; proposal. Nothing here is an owner approval, a validated number, or tested behaviour. r2 implements [PMR007 §1](/Hold%20the%20Flood/PM%20Reviews/PMR007) and REV104-16/17/18. The r1 expedition was not legal and is replaced, not patched: it lit a lure from batteries that start empty, held that lure for five minutes when a full bank buys a hundred lit seconds, then had a crowd spread while a reachable lit lure existed, and drew a rack five cells over three. r2 supplies a legal 2×4 rack, earned charge with an explicit spend-and-loss ledger, the lure's real hundred seconds and what retargets when it expires, two free cells for the retrieval, an intake-local dispatch marker so the flash-location convention holds, and the bank's found state of three engaged clutches that REV104-18 requires to explain the districts already announced. §11 corrects the cable reel to Equipment's U6. r1 is in [page history](/Hold%20the%20Flood/Districts/The%20Exchange/history).

	District 4 of the seven on [Districts](/Hold%20the%20Flood/Districts). [PMR006](/Hold%20the%20Flood/PM%20Reviews/PMR006) assigns CAM102: "deepen the Exchange to the same standard as the three detailed districts." This page does that. The district is the World designer's example ([World and Mystery](/Hold%20the%20Flood/World%20And%20Mystery)), and the lead's outline at [DES104 §3.4](/Hold%20the%20Flood/Campaign#34-the-exchange-in-outline) supplies the hall, the drums, the bank, the shut-regime meaning of a dark position, the bank's coupling and the Carrier's part in it. Everything in that outline is kept; this page adds the ground, the panel, the interaction that takes time, two loadouts, a written expedition, the failure case and the asset list. World owns the clue logic; Lead owns any contradiction with the campaign page or the contract.

@@ 45,20 45,22 @@
	- The cable gallery runs the east wall at head height. Lines conduct when dry (contract r4 §1), so this is a live conduction network with no water in it at all, and it is what a Grounder climbs.
	- Shelters: the dock only. There is no platform on the floor and none at the bank. That is deliberate: the bank is the thing you walk to, and a shelter beside it would delete the district's only pressure.

-	### The panel
+	### The panel, and why nothing here flashes at a beacon

	Format as [DES104 §3](/Hold%20the%20Flood/Campaign#3-three-districts-in-depth) tables each district's.

-	\| Panel position \| Powered branch \| Intake selected \| Beacon that flashes \|
+	\| Panel position \| Powered branch \| Intake selected \| What flashes at a dispatch \|
	\| --- \| --- \| --- \| --- \|
-	\| Dock \| The dock end: b1 lit, the gallery's dock run live, the tail race's lower reach energised \| The upper landing (arrivals walk the length of the floor to reach b1) \| b1 \|
-	\| Floor \| The mid run: b2 lit, the gallery's centre run live \| The upper landing \| b2 \|
-	\| Head \| The head end: b3 lit by the bank, the gallery's head run live, the race's upper reach energised \| The upper landing \| b3 \|
-	\| Off \| Nothing \| No dispatch: the shut regime, trickles down from the upper landing \| None \|
+	\| Dock \| The dock end: b1 lit, the gallery's dock run live, the tail race's lower reach energised \| The upper landing — always; the panel cannot change it \| The landing's own intake marker \|
+	\| Floor \| The mid run: b2 lit, the gallery's centre run live \| The upper landing \| The landing's intake marker \|
+	\| Head \| The head end: b3 lit by the bank, the gallery's head run live, the race's upper reach energised \| The upper landing \| The landing's intake marker \|
+	\| Off \| Nothing \| The upper landing; the shut regime if the trunk does not permit \| The landing's intake marker \|

-	Note what the intake column does not do. The Exchange has one way in for the Flood — the upper landing at the head end — so the panel here never chooses where a surge enters, only where it goes once it has. This is the only district in the campaign whose panel is purely a lure and not an intake selector, and it is a deliberate contrast: the district that lets you route the whole city cannot route its own front door. Pursuit range here is about ten metres (contract r4 §3), and the floor is wide enough that a body at the gallery wall does not chase a surveyor at the race.
+	r1 had the panel flash a beacon at the far end of the hall from the only way in, and [PMR007](/Hold%20the%20Flood/PM%20Reviews/PMR007) is right that this breaks the convention the whole campaign has taught — that the light which flashes is where the bodies appear. The fix costs nothing and improves the district: the Exchange has one way in, so it has an intake marker at the landing, and that is what flashes. The three beacons are lures and never intake selectors.

-	---
+	So this district splits in two what every other district joins. Elsewhere the panel chooses where the Flood enters and where it then goes; here it chooses only where it goes. The player meets, for one district, the half of the panel that is pure lure — which is the cleanest possible statement of the distinction the campaign has been building since the pumping district's two cables, and it is why this district is worth visiting even though it holds nothing. Pursuit range is about ten metres (contract §3), and the floor is wide enough that a body at the gallery wall does not chase a surveyor at the race.
+
+	Two dispatches, not three. This is a ten-minute district: dispatches at about 3:00 and 7:00, with the evacuation window from 10:00. That is a district property, stated so the baseline on [Districts](/Hold%20the%20Flood/Districts) and the scene in §7 agree — [PMR007](/Hold%20the%20Flood/PM%20Reviews/PMR007) is right that r1 wrote a twelve-minute scene against a ten-minute allowance. Lead should confirm that a district may carry two dispatches; nothing in the contract requires three, and the pumping district's three are its own.

	## 3. Shared-rule diff

@@ 75,21 77,19 @@

	## 4. Two loadouts that do different things

-	Eight cells (D023). C coil (2), R cable reel (2), B battery, I insulation, K kit, L portable beacon, S salvage bundle.
+	Eight cells, two rows of four* (D023). C coil (2 cells), B battery, I insulation, K kit, L portable beacon, S salvage bundle. r1 drew five over three, which is not a legal rack.*

	```
	The cacher The dry hand
-	B B L C C I K S S
-	I K S S S S S
+	C C B L I K S S
+	B I K S S S S S
	link B1 → C, B2 → L no battery, no link
	```

-	- The cacher carries charge into the one district where carrying charge is hunted. The Carrier walks to the nearest reachable loose charge in preference to a lit beacon (contract r4 §3, R5), so a surveyor with two batteries is a moving lure. The cacher's play is to make that deliberate: plant the lit beacon and its battery at the dock end, then walk the floor, so every Carrier in the hall is going the other way. It costs two cells and a battery to buy an uninterrupted walk, and the coil is there because the tail race is the one place a crowd can be stunned.
-	- The dry hand carries no charge at all and therefore cannot be hunted for it — the Carrier has nothing here to want, and reverts to line 3, the nearest lit beacon. It has four bundles' worth of space for whatever the hall does hold, and it has no coil, no beacon and no way to power anything: it can set clutches, work the panel, and cut. It is the cheapest correct answer to this district and it is also the one with no recovery — if the panel decision was wrong, the dry hand walks back and does it again.
+	- The cacher carries charge into the one district where carrying charge is hunted. A Carrier's target is the nearest reachable loose charge within its acquisition range (contract §3, R5) — so it is not that any Carrier anywhere comes for you, it is that one close enough will, and the counterplay is distance as much as quantity. The cacher's play is to make that deliberate: plant the beacon and a battery, and let the loose charge be somewhere you are not. The coil is aboard because the tail race is the one water in the district.
+	- The dry hand carries no charge at all, so no Carrier has anything here to want and every one of them reverts to line 3, the nearest lit beacon. It has six cells for cargo and the cutter, and it can set clutches and work the panel and nothing else. It is the cheapest correct answer to this district and it has no recovery — if the panel decision was wrong, it walks back and does it again.

-	Neither is better. The cacher buys safety in advance and pays four cells for it; the dry hand pays nothing and has no second chance.
-
-	---
+	The constraint that makes both interesting, and which r1 got wrong. Batteries start empty every expedition; charge resets ([Design Document §10](/Hold%20the%20Flood/Design%20Document#10-progression-and-persistence)). A lit portable beacon drains one unit every ten seconds, so a full bank of ten buys a hundred lit seconds and no more ([Campaign](/Hold%20the%20Flood/Campaign), the lure's number). And a coil discharge spends a full bank. So one filled battery buys either a hundred seconds of lure or one discharge, never both — and in a ten-minute district there is rarely time to fill two. That is the decision this district actually poses, and §7 is built on it.

	## 5. Discovery

@@ 110,34 110,48 @@

	## 6. Outward consequence: the bank

-	The bank's pricing is not restated here. It is stated once, as a rule, in [Districts §6](/Hold%20the%20Flood/Districts#6-the-routing-bank-priced) and in [DES104 §3.4](/Hold%20the%20Flood/Campaign#34-the-exchange-in-outline), and duplicating it is how a wiki grows two versions of a rule. What belongs here is the local half — what setting it actually costs you inside this district:
+	The bank's pricing is not restated here. It is stated once, as a rule, in [Districts §6](/Hold%20the%20Flood/Districts#6-the-routing-bank-priced) and in [DES104 §3.4](/Hold%20the%20Flood/Campaign#34-the-exchange-in-outline). What belongs here is the found state and the local cost of changing it.

-	- Two of seven, and the third clutch is seized. The player is always choosing what to be blind to. The seized clutch is visible from the first visit, so the constraint is legible before it bites, and whether it can ever be freed is the head-works' business, not this district's.
-	- A swap is disengage, walk, engage. The positions sit in the districts' own order along the shaft, so swapping two adjacent districts is a short walk and swapping the pumping district for the head-works is the length of the floor. The map's shape becomes the cost of the plan, which is the neatest thing this district does.
-	- The walk is the exposure. There is no shelter on the floor. Whatever the panel was set to when the player left the dock end is what the crowds are doing while they work.
-	- Nothing is spent. No salvage, no charge, no module. The bank costs a trip and a walk, and that is the whole price — which is why the trip has to be worth planning rather than worth farming.
+	The found state: three clutches engaged, on Pumping, Lock Flight and Tram Incline. This is REV104-18, which [PMR007](/Hold%20the%20Flood/PM%20Reviews/PMR007) correctly calls a campaign dependency rather than a footnote: the player has by now had announced dispatches in three districts, and two engaged clutches cannot explain three announced districts. The answer is that the bank drives three, and the player finds it already set to exactly the three districts they have been visiting — which is not a coincidence to be explained away but the district's best single moment, because it tells the player that the route they thought they chose was the route something upstream had already permitted.

-	---
+	I have not adopted the critic's seized-Pumping-clutch candidate, and [PMR007](/Hold%20the%20Flood/PM%20Reviews/PMR007) is right to make that an explicit choice rather than a preserved ambiguity: a seized clutch on Pumping would mean the player can never darken their opening district, and darkening Pumping is the campaign's best worked example of a later discovery changing an earlier return ([Districts §4](/Hold%20the%20Flood/Districts#4-the-progression-map-main-route-an-optional-discovery-and-a-later-discovery-that-changes-an-earlier-return)). Three live clutches keeps that and costs nothing. The seized fourth position stays as it was: visibly seized, never usable, and its repair is the head-works' business rather than this district's.
+
+	Three of seven, and the scarcity argument is unchanged. Most of the city is unread at any time, and every district added costs one given up. What changes from r1's two is only the arithmetic, and it changes in the direction the campaign's own history demands.
+
+	The local cost of a change.
+
+	- A swap is disengage, walk, engage. Positions sit along the shaft in the districts' own order, so swapping neighbours is a short walk and swapping the ends is the length of the floor. The map's shape becomes the cost of the plan.
+	- The walk is the exposure. There is no shelter on the floor. Whatever the panel was set to when you left the dock end is what the crowds are doing while you work.
+	- Nothing is spent. No salvage, no charge, no module. The bank costs a trip and a walk — which is why the trip has to be worth planning rather than worth farming.

	## 7. A written expedition — illustrative, not a playtest

-	Declared setup. Middle campaign. The player wants the Collector permitted (they are going there next, and they want its dispatches announced so a mass can be held deliberately) and is willing to darken the Tram Incline for it. Currently permitted: Tram and Pumping. Rack: the cacher. Panel found on Head — b3 lit, by the bank. Three dispatches at roughly 3, 8 and 12 minutes; the intake is the upper landing whatever the panel says.
+	r1's scene was not legal and is replaced. The corrections [PMR007](/Hold%20the%20Flood/PM%20Reviews/PMR007) and REV104-16/17 ask for are all in the ledger below rather than in prose around it.

-	- 0:00. The dock. The board reads first dispatch in three minutes. b3 is lit at the far end, which means the first surge will walk the whole floor away from the player — good for the walk, bad for the bank, because the bank is where they are going.
-	- 0:30. The player throws the panel to Dock, lighting b1 beside themselves, and plants the portable beacon and its battery at the dock end too, twenty metres off. Two lures at the dock end, one of them loose charge.
-	- 3:00. The dispatch. b1 flashes; the surge enters at the upper landing and walks the length of the floor toward b1, past the bank the player has not left for yet. The player waits at the dock end, at the edge of pursuit range, and cuts the leading edge for the first motes.
-	- 4:30. The crowd is gathered at b1 and the two Carriers in it are on the planted battery, not on the player's rack. The floor is now empty in the direction the player wants to walk. They start north.
-	- 6:00. The bank. Disengage Tram — its drum stops, and the board's lamp for it goes out. Walk four positions to Collector, engage — its drum was already turning, and what changes is the clutch. Two engaged again: Pumping and Collector.
-	- 8:00. Second dispatch while the player is still at the head end. b1 is still lit and still nearer to the landing than anything else, so the new surge also walks south. The player is behind it the whole way.
-	- 9:30. A Grounder that came in with the second surge climbs the gallery's dock run and grounds it. b1 goes dark. Under R1 the crowd re-evaluates: b2 and b3 are unlit, so there is no other reachable lit beacon, and it spreads along the floor — toward both ends, including the player's. The planted beacon is on its own battery and is not on the grounded run, so it stays lit and holds the Carriers, which is the difference between an inconvenience and a disaster.
-	- 10:00. The player cannot reach the panel: it is at the dock end, behind the spreading crowd. They discharge the coil into the tail race where the spread has reached the water, stun what is in it, and walk the west wall past the gap.
-	- 12:00. Third dispatch. The player is near the dock, retrieves the beacon and its battery, and takes the landing route out at the evacuation window with two bundles and a set bank.
+	Declared setup. Middle campaign, ten-minute expedition, dispatches at 3:00 and 7:00, evacuation from 10:00. The player wants the Collector permitted and will give up the Tram Incline for it. Bank found engaged on Pumping, Lock Flight, Tram. Rack: the cacher — C C B L / B I K S, B1 linked to the coil, B2 to the beacon. Both batteries empty, as they are at the start of every expedition. Panel found on Head, so b3 is lit at the bank end. The hall's standing stragglers are its own: this room is never quiet, which is what makes charge earnable here before a surge lands.

-	What the expedition was about. Not the bank — that was thirty seconds of it. It was about the decision at 0:30, made before anything happened, to spend two cells and a battery on putting the crowd at the end of the hall the player was leaving. A player who walks north with the panel still on Head arrives at the bank with the surge, and the same trip is a fight. The Grounder is what makes the plan expire rather than simply succeed, and the ratified rule — spread only when no other lit beacon is reachable — is exactly what made the planted beacon worth its cells.
+	\| Time \| What happens \| B1 \| B2 \| Ledger \|
+	\| --- \| ---: \| ---: \| ---: \| --- \|
+	\| 0:00 \| Dock. Board reads dispatch 3:00, intake the landing. b3 lit at the far end — the surge would walk toward the bank the player is going to \| 0 \| 0 \| — \|
+	\| 0:20 \| Panel thrown to Dock. b1 lights beside the player, b3 goes dark. Now the surge will cross the whole floor away from the bank \| 0 \| 0 \| — \|
+	\| 0:20–3:00 \| The player cuts stragglers along the floor at one unit a kill. Collection goes to the player's set priority (D023): B1 first \| 8 \| 0 \| +8 earned \|
+	\| 3:00 \| Dispatch. The landing's marker flashes; the surge enters at the landing, head end, and walks the length of the floor to b1, the only lit beacon \| 8 \| 0 \| — \|
+	\| 3:00–4:30 \| It arrives and holds at b1. The player works its edge inside pursuit range, cutting \| 10 \| 4 \| +6 earned; B1 full \|
+	\| 4:30 \| The decision. One full bank. It buys either one coil discharge or a hundred lit seconds of planted lure. The player links B1 to the beacon, plants it twenty metres up the floor from b1, and starts north. The coil is now unusable — B2 holds 4, and a discharge needs a full bank \| 10 → planted \| 4 \| −10 committed to the lure \|
+	\| 4:40–6:20 \| The lure is lit for 100 seconds and no longer. Two Carriers in the crowd take it as the nearest reachable loose charge and gather on it, not on the player walking away \| — \| 4 \| — \|
+	\| 6:00–6:45 \| The bank. Disengage Tram — its drum stops, its lamp goes out. Walk four positions. Engage Collector. Three engaged again: Pumping, Lock Flight, Collector \| — \| 4 \| — \|
+	\| 6:20 \| The lure expires, battery empty, beacon dark. Retargeting, stated: b1 is still lit and is right beside them, so the crowd stays and the two Carriers revert to line 3 and take b1. Nothing moves. The lure bought a walk, not a rearrangement \| — \| 4 \| lure spent \|
+	\| 7:00 \| Second dispatch, and the player is at the head end. The landing's marker flashes beside them and the surge enters there. b1 is the only lit beacon, so it sets off south — away from the player, and along the road the player has to take home \| — \| 4 \| — \|
+	\| 7:00–9:00 \| The player follows it down the west wall, behind its press. A Grounder that came in with it climbs the gallery's dock run and grounds it: b1 goes dark \| — \| 4 \| — \|
+	\| 9:00 \| Retargeting again, and this time it is legal to spread: b2 and b3 are dark, the planted beacon died at 6:20, so no reachable lit beacon exists and the crowd spreads both ways along the floor, including back toward the player. The dock run is grounded, so the race's lower reach is dead — safe to wade, and no help \| — \| 4 \| — \|
+	\| 9:00–9:40 \| No coil (B2 holds 4, not a full bank). The player walks the spreading edge on the cutter and the insulation, retrieves the planted beacon and its empty battery — two cells are free, because those two left the rack when they were planted — and reaches the dock \| — \| 4 \| retrieval fits: 6 cells used, 2 free \|
+	\| 10:00 \| Evacuation. The 4 units in B2 are lost at the dock; charge never leaves a district \| — \| 0 \| earned 14 · spent 10 · lost 4 · carried out 0 \|

-	Not a playtest. No timing, difficulty or enjoyment claim is made or could be.
+	What the expedition was about. Not the bank — that was forty-five seconds of it. It was the decision at 4:30, which is the district's whole design in one choice: a single full bank buys either the walk or the way home, and the player who spends it on the walk comes back through a spreading crowd with no coil. A dry hand makes the same trip with no decision, no lure and no risk of that particular failure, and has to time the walk against the dispatch instead.

-	---
+	Why the Grounder is placed there rather than earlier. It makes a plan expire at the moment the plan has already been paid for. Had the lure still been lit at 9:00 the crowd could not have spread at all — a reachable lit beacon forbids it — which is exactly the illegal move r1 made and PMR007 caught. The scene now works because the lure ran out, not despite it.
+
+	Not a playtest. No claim is made about timing, difficulty, survivability or enjoyment, and none could be. Every number above is a proposal checked against the written rules, which is a consistency check and nothing more.

	## 8. Entry, revisit, and one consequence of failure

@@ 161,7 175,7 @@
	Roles are Combat's, in [CMB101 r6](/Hold%20the%20Flood/Enemies) and [Combat Rules r5](/Hold%20the%20Flood/Mechanics/Combat); Campaign owns placement. Everything below works with the Walker alone.

	- This is the Carrier's district, and the lead's outline says so. It is the one place the player's own stored charge is a liability with a designed answer — cache it deliberately — rather than a hazard to be endured. Placement proposal for Combat: Carriers present in every surge here, at low density, so the cacher's plant is worth its cells on the first visit rather than the third.
-	- The Grounder is what makes a plan expire, and this district is where that reads best: the gallery is at head height along one wall, so a Grounder is visible climbing it, and the player watches their arrangement come apart with time to answer. Placement proposal: at most one per surge, and never in the first — the first visit should teach the hall, not the exception.
+	- The Grounder is what makes a plan expire, and this district is where that reads best: the gallery is at head height along one wall, so a Grounder is visible climbing it, and the player watches their arrangement come apart with time to answer. Placement proposal: at most one per surge, never in the first surge of a first visit, and best in the second — where §7 puts it, because a plan that expires after it has been paid for teaches more than one that never forms. Never place one where it would darken the only lit beacon while the player has no charge and no route: that is not a plan expiring, it is a trap.
	- Anchors belong in the tail race and nowhere else here: one standing hazard in the only water, so the coil's own working ground is contested.
	- What this district does not want: Crusts and Sparks. Both are about water and conduction, this hall has almost none, and putting them here would blur the contrast with the [Collector](/Hold%20the%20Flood/Districts/The%20Collector). A district should decline roles as deliberately as it asks for them, and this is the one place in the campaign I would decline two.

@@ 186,5 200,5 @@
	- World (WLD103). The campaign clue in §5 is the lead's and the World designer's; the "board as the campaign's index of districts" convention in §5 is mine and is a proposal in World's domain, because it decides how a player learns a district exists.
	- Lead (DES104). Whether the panel table in §2 — a district whose panel selects no intake — is acceptable under the contract. I believe it is: the contract requires the intake column to be true, not to have more than one value, and one-way-in is a topology fact, not a rule change. If Lead disagrees, the fix is a second landing and the district loses a good contrast.
	- [CMB101](/Hold%20the%20Flood/Enemies). The three placements and the two declines in §9.
-	- [EQP101](/Hold%20the%20Flood/Equipment). The cable reel is listed in the contract as joining two wet points; this hall has one narrow race and a dry gallery, so the reel has little to do here and I have kept it out of both loadouts. If Equipment intends the reel to be a general dry-ground conductor, this district would want it — that is a question for them, not a request.
+	- [EQP101](/Hold%20the%20Flood/Equipment) — corrected at r2. r1 described the cable reel as joining two wet points, which was the r4 contract's wording; [contract r5](/Hold%20the%20Flood/Mechanics/Power%20And%20Attraction) restates it as Equipment's U6, a bridge between exposed power terminals. That makes it a far better fit for this district than r1 allowed — a dry hall whose gallery carries live runs is exactly where bridging terminals is worth two cells. I have still kept it out of both loadouts in §4, because the two the district needs to contrast are the one that carries charge and the one that carries none, and a third rack would blur that. A reel loadout here is a real third answer and Equipment is better placed to draw it.
	- Unresolved by me, honestly. Whether a district with no salvage worth carrying and one setting to change can hold an expedition's worth of interest on its third visit. My answer is the walk and the panel decision in §7, and I believe it; but this is the district in the campaign whose premise is thinnest, and if any of the seven fails the lead's written-expedition test on contact, I would expect it to be this one. That is why §10 records what it would become instead.

Commit d71b9b

Commit `d71b9b`