TODO.md had stopped tracking its own closures

Nothing was wrong with the ordering. What had gone wrong is that eight
items were done or superseded in one place and still open in another,
which reads as live work — worse than no entry at all. Found by
reviewing the file against the Gitea tracker and the git log, not by
reading it.

NEXT WAS 43 CHRONOLOGICAL ENTRIES, 26 OF THEM STRUCK THROUGH, and the
live work had stopped being visible in it. It is grouped by what each
item is WAITING ON now — blocked on a table, the screen, waiting on
something outside the code, unscheduled, standing practice — with the
closed entries moved to a Settled block at the end of the section.

EVERY NUMBER IS KEPT AS A STABLE ID. The rest of the file refers to
items by number ("decide with item 15", "same drawing pass as 19-21"),
so renumbering would have silently broken every cross-reference.
Numbers are ids, not positions, and the file now says so. The one
reference that pointed at a settled item (item 22, the dead centre of
the Division map) is rewritten to name the principle that outlived it
rather than the item.

EIGHT STALE ITEMS CLOSED. The Heavy Grade orientation was fixed in
cf018b4 and never checked off. The Fedora is TODO #29, done in this
same unshipped v0.7.9. The remaining six were the Division-map drawing
pass, all answered by Gitea#18 (closed 2026-08-26) replacing the layout
rather than fixing the drawing — 167 lines describing a wrapped map
that no longer exists, collapsed to 32 that keep the two ideas which
outlived their items: "SHARED in the middle, YOURS on the right", now
Gitea#20's territory, and why rotating the map to seat a viewer at the
bottom was refused (it moves the open gap between the buffer stops out
of the eye's path, and that gap is the only thing saying the Division
is a line and not a loop).

65 open items to 57; 2,400 lines to 2,179.

TWO PIECES OF DEAD CODE GITEA#18 LEFT BEHIND, found while confirming
the map really is one row before closing the items that say it is.
SIDE_GAP was declared and never read. And divisionSvg's header comment
still described "one row alone, two rows facing, a horseshoe of three,
a square of four" — contradicting the layout comment 200 lines below
it, which explains why that layout was dropped. A comment that
survives the code it describes is how the next reader gets it wrong.

No behaviour change. 878 tests pass.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01YTaNBL1jVxNqgFdjHkHoo3

CHANGELOG.md

@@ -19,6 +19,406 @@ page as `v0.1.0 · <sha> · <date>`, so what is deployed can always be identifie
 
 ---
 
+## 0.7.9 — 2026-08-30
+
+Four pieces of feedback on the solitaire setup screen, plus the rules bug the second one exposed.
+
+### The collision limits were dead settings in solitaire — now they are not
+
+Asked to reword the collision entries to "the game ends immediately and results in a loss", which
+turned out to be unwriteable: `advance.ts` gated the whole §3.4 check on
+`mode === 'competitive' || mode === 'coop'`, and a solitaire game's mode is `'solitaire'`. So both
+limits were offered on the New Game dialog as live settings, rode into the config, and never fired.
+A solitaire player could set a limit of 1 and crash all game. The *existing* text beside them ("the
+game ends in a loss") was already false; the new wording would only have made it more so.
+
+The exclusion was never a stated rule — §3.4 does not carve solitaire out — and nothing recorded a
+reason for it. **Jesse's ruling: the settings do what they say**, so the mode gate is gone rather
+than the controls.
+
+**A solitaire game can therefore now end early, and that is measured rather than asserted.** Over
+200 standard games with the developer bot: `loss/collisionFloor` **1 game in 200**, with Days played
+falling 5.00 → 4.98 mean and a *minimum of 1* — a bad opening Day can now end a game outright.
+Collisions per game are unchanged (0.14 mean, max 3), which is the point: the ending is rare because
+crashes are, not because the check is lenient. `0` still switches either limit off, at one seat
+exactly as at four, and that path has its own test now.
+
+Every figure in `TODO.md` quoted from a full-length solitaire run predates this.
+
+### Where an Extra may start defaults to your own Control Point
+
+Was "any player's". At one seat the two are the *same rule* — `apply.ts` only rejects `ownOffice`
+when `start.seat !== seatOf(s, player)`, which cannot happen — so this is a labelling fix with no
+gameplay effect and nothing to re-measure. The permissive label described a permission a lone player
+was never being granted, and named an "any player" they have no contrast with.
+
+### The extension question was hidden behind the results screen
+
+Jesse, 2026-08-30: "Solitaire game ended. I did not have an option to extend the game by a day."
+
+**The engine and the Frame were right the whole time** — checked before changing anything: a
+solitaire game at the end of its timetable reaches `awaitingExtension` with `extensionVotes: [null]`,
+and `isExtendable` covers both `revenueFloor` and `daysElapsed`. `renderActions` reaches
+`renderEnding` for any non-active status, and `renderEnding` writes "play one more Day" and "end the
+game here" into `#actions`.
+
+Then it opens `#resultsdlg`, **which is modal**. So the two buttons were rendered directly underneath
+a dialog whose only control was Close, and the player read a results screen that offered nothing but
+Close and concluded the game was over — which is precisely what it looked like.
+
+The results dialog now carries the question itself, hidden unless a vote is actually pending:
+**Play One More Day** and **End the Game Here**, casting the same `game.extend` intent. The
+`#actions` buttons stay, because they are what remains once the dialog is closed and what a player
+who reopens it with "see the full results" comes back to.
+
+**Why this survived being "verified".** `TODO.md` #35 recorded extended play as checked on
+`phoenix.local` — over the HTTP API, which renders no dialog. The browser path had never been run to
+the end of a timetable. Same shape as the three attempts before it: the thing that was verified was
+not the thing the player uses.
+
+### `configWith` let the day count and the Revenue floor disagree
+
+`minCombinedRevenue` fell back to `SOLO_CONFIG`'s constant — the floor for a *five*-Day game —
+whatever `days` said. So `configWith({ days: 1 })` asked a one-Day game to clear **15**, a figure a
+full five-Day game averages barely half of, and `configWith({ days: 10 })` asked for the same 15 a
+five-Day game does. It derives from the days it was actually given now.
+
+Not a live fault: `createLocalSession` is the only caller, and the page always writes the floor
+itself, so no dealt game was ever wrong. It was found by a throwaway probe written to reproduce the
+extension bug above, which passed only `days` — which is exactly how the next caller would reach for
+it. At the default day count the answer is unchanged, since `SOLO_CONFIG`'s own floor is this same
+formula at `DEFAULT_DAYS`; the three new tests pin that as well as the derivation.
+
+### A Heavy Grade shows which way it climbs
+
+Jesse, 2026-08-30: "heavy grade mainline card tooltip states climbs east, but card doesn't show it."
+The Frame has carried `gradeUp` since the Division map was rebuilt and the tip has read "climbs
+east" all along — but the one Mainline card whose orientation is set per game, and the one where
+Helpers and Brakeman mean opposite things at opposite ends, drew nothing to tell the two apart.
+
+A brown wedge in the card's lower right, rising toward the climb, with a bone arrow lying along its
+slope. **East is right on this map** (Gitea#18), which is what lets a wedge be read without a
+compass — and is the layout decision paying for itself again.
+
+**Four passes, and the first three are worth recording because each failed differently.** An arrow
+up the hypotenuse had to start at the wedge's thin corner, where there is no height to draw in, so
+its head sat over the edge and read as clipped. Level, it was contained but did not read as
+climbing. At 30° — steeper than the wedge's own 22.5° — it had to be tucked into the fat half to
+survive. Lying **along** the slope is the shape that fits: the perpendicular gap to the hypotenuse
+is then constant down the whole arrow instead of closing at one end, so it can sit on the triangle's
+centroid. The wedge grew 44×19 → 58×24 to pay for that, because a centred arrow has *less* room than
+an off-centre one — the centroid is about 7px from the hypotenuse.
+
+The final sizes are a search result rather than an eyeballed nudge: the roomiest arrow keeping all
+seven vertices clear of both edges, at 2.88px.
+
+**A test that passed a visibly broken glyph is the reason this is pinned properly.** The first
+containment check bounded the arrow against the CARD, which it never left — while the wedge clipped
+it. The check is against the *triangle* now, with a 2px floor, plus the arrow being parallel to the
+wedge (derived from the wedge, so resizing it cannot leave a stale angle) and centred on the
+centroid.
+
+**Orientation is always set — measured, not assumed**, since a wedge that invented a direction would
+be worse than no wedge. Across 400 seeds × 4 player counts: **535 Heavy Grades placed, 0 without an
+orientation**, 276 east / 259 west. `setup.ts` is the only place a Mainline node is created and it
+rolls the direction from the seed for every grade, so the `?? 'east'` fallbacks in `view.ts` and
+`advance.ts` are unreachable.
+
+### Two setup screens, not three — the in-game dialog is deleted
+
+Jesse, 2026-08-30: "If you are in the Solitaire game and you click the new game dialog, it should not
+go to a separate screen. We should reuse the Solitaire New Game Screen… in general we should reuse
+what we already have."
+
+`#newgamedlg` was a third copy of the same questions and the one that drifted. It was shown ONLY to
+a solitaire player, and it asked "Everyone loses if **combined** Revenue at the end is under" — a
+table's question, put to one person — and explained Employee Rotation in full, in multiplayer terms,
+beside a control it had itself disabled. Both were on the list to re-word. Deleting the screen
+removes the drift instead of restating it, which is the cheaper fix and the one that cannot drift
+again.
+
+**New game now opens the setup screen in place.** Not a navigation: the live session stays in memory,
+so the fields open on the rules actually being played — which is what the dialog was good for, and
+losing "change one dial, redeal, compare" would have been a real loss — and **Continue Existing Saved
+Game** puts the board straight back with no reload and no replay. Nothing is at risk either way:
+`render()` calls `save()` every frame, so the game in progress is always on disk.
+
+**And the two remaining screens now match below their headers.** Each keeps its own opening — the
+lobby's join/secret section and "Create a new game" are multiplayer's alone — but from the parameters
+down they are one form: the same three fields in the same order (seed, players at the table, days),
+the same note about a seed and settings dealing the same railroad, and the same note about every
+chair being taken. Solitaire shows **Players at the table** too, locked at one, rather than omitting
+it — a control that is present and fixed says "this is the same form, at a table of one", where a
+missing one just made it a different form that happened to share a rules block.
+
+The drift guard that had been checking three prefixes now checks two, and a new assertion fails if
+any `ng-` id ever reappears.
+
+### The two screens are worded the same, section by section
+
+Jesse, 2026-08-30: "let's get the text between the Multiplayer and solitaire as close as possible to
+the same." Walked section by section; where the two said the same thing differently, the lobby's
+wording won, because it was written for the harder case.
+
+- **The chair note is one paragraph on both**, replacing two that each described only their own
+  case: "Every chair must be filled before the game can start. For solitaire, there's only one
+  player. For multiplayer, that must be filled by a person or a bot. The number of players cannot be
+  changed once the game is created."
+- **The Game type heading names which game the screen deals** — "Game type (multi-player)" and
+  "Game type (solitaire)". That heading is now the whole explanation for the dimmed rows.
+- **No reason is printed beside a dimmed type any more.** The lobby appended "— dealt with the New
+  game button, not here" to Solitaire, and the solitaire screen appended "— use the Multiplayer
+  button; a table needs a server" to each of three. Jesse: "grayed out with no additional
+  explanation. The explanation above… is sufficient." One heading says it once; three rows were
+  saying it three times. `markUnavailable` is down to dimming, and `.lb-why` is gone with them.
+- **The sentence under the radios is deleted on both.** It restated the type just chosen to the
+  person who had just chosen it. Its one non-obvious case — how many settings a Custom game differs
+  by — is not lost: `form.mark` already puts "Co-op default: …" on each row that differs, which is
+  where a reader can act on it rather than a count they would have to go and find. The tests that
+  read that sentence now read the row hints instead.
+- **The Game settings, Starting hand and Victory conditions notes are the lobby's on both**, with
+  one edit Jesse asked for: "clicking a type again resets" → "changing the game type resets", which
+  names the control that does it rather than a gesture. Victory conditions was not on his list —
+  applied under the same rule, and the lobby's wording fits now that solitaire also shows the table
+  size.
+
+What is deliberately still different: each screen's own opening paragraph, and the heading above.
+
+### The status line said nobody was holding the game up
+
+Jesse, 2026-08-30: "waiting on shows 'nobody — the Division is running itself' BUT the system is
+actually waiting on the Superintendent."
+
+`Frame.actor` carried `clock.currentActor`, which is null for the whole Mainline Phase — so the three
+interruptions that stop the game and put a question to a named person (§8.1's clearance ruling,
+Gitea#5's Yard Office offer, Gitea#19's Red Flag prompt) all reported that the Division was running
+itself. `actingPlayer` has had the answer since the Gitea#5 refactor; the Frame threw it away.
+
+It carries `actingPlayer` now, **and what the question is**: "waiting on **Bob** · a clearance ruling
+· Train 4". Naming the person is not enough on its own — three different things can be pending, and
+"waiting on Bob" with nothing after it is a game that looks stuck to everyone except Bob.
+
+### The Fedora moved to the end of the phase row
+
+`TODO.md` #29. It sat on a line of its own between the phase chips and everything above them, which
+put a thing that moves every third Stage in among the things that move every Stage. It rides at the
+right-hand end of the phase row now — the row whose last chip is Supervisor Shift, the phase that
+passes it — and wraps underneath rather than squeezing the chips on a narrow screen.
+
+### The developer replay stopped printing a raw enum
+
+`TODO.md` #34. Its heading read `loss — revenueFloor`: the exact defect Gitea#16 was filed about on
+the playable page, still alive here a release after that was fixed, because nothing player-facing
+pointed at it. It reads "lost — The Division closed short: 3 Revenue between everyone, against a
+floor of 15…" now.
+
+`panels.ts`'s `reasonSentence` is exported and shared rather than reimplemented, so the replay and
+the results screen cannot explain one ending two different ways. It is fed the last recorded frame —
+the state the outcome was decided in — and stripped of markup, since it lands in an `<h1>` and a
+console line. The drift test that compared this string against the engine now maps `win`/`loss` to
+`won`/`lost`, so it still checks the two agree rather than that they are spelled alike.
+
+### The save warning, the buttons, and a claim that was simply false
+
+Three more from Jesse reading the two screens side by side.
+
+- **The save warning is a warning.** "That's a warning, not an 'oh by the way'." It had already moved
+  off `.ng-note` earlier in this release; it is larger and heavier again now — 15px, weight 500,
+  bright amber on a deeper ground with a 5px rule down the side. Worth noting what he was actually
+  looking at: `phoenix.local` runs v0.7.8, where this line is still the small grey `.ng-note`. None
+  of this release has been deployed.
+- **The buttons read the same on both screens**: **Continue saved game** and **Create new game**.
+  Solitaire said "Continue Existing Saved Game" and "Deal New Game"; the lobby said "Create game".
+  Three phrasings for two actions.
+- **"Off in every game type" is deleted from the Optional rules note, because it was not true.**
+  Checked against the presets rather than taken on trust: `discardTimetabled` — §6.2's "a Timetabled
+  train may be discarded" — ships **on** in all four types, not just Co-op. The note now says only
+  what is true of all of them: "Each one changes how the game plays."
+
+### A game you come back to has not begun
+
+Jesse, 2026-08-30: "when you are continuing the saved game out of that screen, do not post a message
+that says 'The game has begun.' … it needs to say 'The game has resumed.'"
+
+A restored game draws exactly like a dealt one — mid-Day, mid-phase, with a log already several turns
+deep — and **solitaire said nothing at all** to tell them apart. It flashes "The game has resumed —
+Day 3, Stage 7" now, on both ways back: the setup screen's **Continue saved game**, and a bare reload
+that restores the save.
+
+**The same line was wrong on the multiplayer side, in the other direction.** `noteFirstFrame` guards
+on `firstFrameSeen`, which is per page-load — so re-entering a game this browser already held a seat
+in, by reloading mid-game or picking it out of the lobby's list, announced that the game had **begun**
+to somebody who had been playing it for an hour. `beginRemote` carries whether this is a rejoin now,
+and the line reads "resumed" when it is.
+
+Both halves are pinned, including that a brand-new game does *not* claim to be a resume — an
+announcement that fires either way says nothing.
+
+### Three wording and layout fixes
+
+- **The collision entries** on all three screens now read "The game ends immediately and results in
+  a loss…". The Revenue floor keeps its own wording, since it is settled at the end rather than on
+  the spot.
+- **Employee Rotation** on the solitaire screen says "not applicable for solitaire" instead of
+  "meaningless at a table of one, shown here so this screen and the lobby read as one list" — which
+  explained the page's own construction to somebody who had not asked.
+- **The save warning is legible.** It sat in `.ng-note`, the same dim 11px grey as the twenty
+  explanatory notes above it, while being the only thing on the screen describing something
+  irreversible. It is 14px on an amber panel now — amber and not red because losing a save is a real
+  cost, not a danger, and red would outrank the actual rules above it. The buttons say **Continue
+  Existing Saved Game** and **Deal New Game** rather than "Continue saved game" and "Deal".
+
+878 tests pass, eleven of them new; one existing test asserted the opposite of the collision ruling
+above and says so where it was reversed.
+
+---
+
+## 0.7.8 — 2026-08-30
+
+### The setup screen was unreachable for anyone who had ever played
+
+Third report of the same symptom, and this time the build was confirmed current on screen
+(`0.7.7-mtf7hyxc`), which ruled out the caching fault v0.7.7 had just fixed and left the actual
+cause with nowhere to hide.
+
+**v0.7.5 skipped the setup screen whenever `load()` found a save**, reasoned in its own comment as
+"a saved game is a game to resume". The consequence went unnoticed: a browser that has ever played
+solitaire *always* has a save, so the door could never reach the screen again. Only a browser that
+had never played would see it — which is exactly why a fresh private window appeared to prove
+v0.7.7's cache fix. The private window had no save. Two genuine faults were stacked, the caching one
+was real and is fixed, and it masked this one.
+
+**The door now outranks a saved game.** `?solitaire` is an explicit request to set a game up;
+clicking "Play solitaire" is not a request to resume. A BARE reload still resumes, which is the
+zero-friction case D11 is about and is pinned by its own test.
+
+**Dealing from the door would have destroyed a game in progress**, since `commitNewGame` calls
+`clearSave()` — so the screen now carries `#ss-resume` ("Continue saved game") and states plainly
+that dealing replaces the save. Resuming navigates to the bare URL rather than building a session
+on the spot, so `start()` stays the only place that turns a URL into a game.
+
+**Recorded because the failure was diagnostic, not technical.** The first two attempts each fixed
+something real that was not this, and both were reported as verified. The routing fix in v0.7.6 was
+verified by reading what the server served; v0.7.7's by the same. Neither ever exercised the actual
+path with the actual state a returning player has. The reproduction here is a failing test asserting
+the door with a save present — written before the fix, and it failed with "a saved game swallowed
+the door".
+
+### The splash footer names both ways to play
+
+Was "solitaire runs entirely in your browser — no server code required", written when solitaire was
+the only door. Now: "Multiplayer runs on StartOS server. Solitaire runs entirely in your browser."
+(Jesse, 2026-08-30, asked to ride along with the next change rather than take a release of its own.)
+
+868 tests pass, four of them new.
+
+---
+
+## 0.7.7 — 2026-08-30
+
+### Two releases shipped to a browser that never received them
+
+Jesse installed v0.7.5, clicked **Play solitaire**, and landed in a dealt game instead of the new
+setup screen. v0.7.6 diagnosed that as a routing bug, fixed it, installed, verified — and it happened
+again, identically. The second report is what made the real cause findable: the fix was correct both
+times and neither one ever reached the browser.
+
+**`buildStamp()`'s no-git fallback was the literal `nogit`, and the `.s9pk` build has no git.** The
+Dockerfile copies the working tree in without `.git`, so `git rev-parse` fails there on every
+packaged build — and that string is not only the visible stamp, it is the cache-bust key every module
+URL carries. So v0.7.4, v0.7.5 and v0.7.6 all published `./web/main.js?v=nogit`, byte-identical, and
+a returning player's browser correctly concluded it had the file already. The fallback is now the
+package version plus the build's own timestamp, which is always distinct and needs nothing from the
+environment. Proven rather than assumed: two builds of an identical git-less tree now stamp
+`0.7.6-mtf6l8rm` and `0.7.6-mtf6lant`.
+
+**And the server sent no `Cache-Control` at all**, which is the other half — the pages are the one
+thing that cannot be versioned in their own URL, since a player types the address or follows a
+bookmark, so a cached `play.html` pins that player to the whole build it names including every `?v=`
+inside it. Fixed the exact way round that matters: a request carrying `?v=` may be stored for a year
+and marked `immutable`, and anything else is `no-cache`. `?v=` rather than "not HTML" because
+`build-web.ts` tags the modules and nothing else — a year of `immutable` on an untagged image or on
+the replay manifest would outlive several releases of it.
+
+Neither half is sufficient alone: without the varying tag there is nothing for a fresh page to point
+at, and without the header the fresh page is itself served from cache.
+
+**What this says about the two releases before it.** v0.7.5's setup screen and v0.7.6's door fix were
+both real, both correct, and both verified on `phoenix.local` by reading what the server served —
+which was true, and was never the thing in doubt. What went unverified was the browser, and a
+hard-reload would have told us on the first report. Worth remembering the next time a fix "has had no
+effect": check that it arrived before re-diagnosing it.
+
+864 tests pass, two of them new — one pinning the no-git fallback as something that varies per build,
+one pinning the header rule and that the `?v=` flag actually reaches `serveStatic`.
+
+---
+
+## 0.7.6 — 2026-08-29
+
+### The solitaire door could not reach solitaire
+
+Found by Jesse verifying v0.7.5 on `phoenix.local`: from a browser that had ever held a multiplayer
+seat, clicking **Play solitaire** on the splash landed straight in a Co-op, four-seat lobby left over
+from unrelated earlier testing — not the new setup screen v0.7.5 just shipped.
+
+`start()` checks a browser-remembered multiplayer session (`station-master.remote.v1`) before it ever
+looks at solitaire's own state, and there was nothing distinguishing "clicked Play solitaire" from
+"reloaded mid multiplayer game" — a bare `./play.html` load means both. `?lobby` already solved the
+identical problem for the door on the other side (D11); the solitaire door had no equivalent marker.
+
+The door now links to `./play.html?solitaire`, and `start()` treats that — along with an explicit
+`?seed=` or a `?hand=` the setup screen's own Deal button just wrote — as unambiguous proof this
+navigation means solitaire, checked ahead of the remembered-session lookup rather than only below it.
+The `hand` check matters on its own: without it, pressing Deal would work once and then bounce the
+very next load into the remembered game, since `commitNewGame`'s URL carries `hand=` but not
+`solitaire=`.
+
+862 tests pass, three of them new: the door reaching solitaire past a remembered game, a bare reload
+still correctly resuming one (unchanged behaviour, pinned so the fix does not overreach), and Deal's
+own URL surviving the same bounce.
+
+---
+
+## 0.7.5 — 2026-08-29
+
+### Solitaire asks first, the same way multiplayer already does
+
+Jesse: "let the user choose their options like the start of a multiplayer game"; "asking first is
+the only path." A bare visit to `play.html` used to deal a game on the spot, at whatever defaults
+`gameOptionsFromUrl` fell back to, and the only way to see or change a setting was to open the
+in-game "New game" dialog after the fact — compare a hand you already have, not one you are about
+to be dealt. The lobby has asked this question for every multiplayer game since v0.6.0; solitaire
+never did.
+
+A genuinely fresh visit now lands on a new `#solitairesetup` screen first: game type, starting hand,
+where an Extra may start, the three revenue rates, victory conditions, and the three optional rules
+— then a Deal button. A saved game, an explicit `?seed=`, or a URL a Deal already wrote (`hand` is
+the field every write always sets, so its presence is what tells the difference) all skip straight
+past it, the same way `?lobby` already skips the front doors on an invite link — those are not "no
+plan yet", they are a choice already made, elsewhere.
+
+**One shared block instead of two copies drifting apart.** The in-game dialog, the lobby, and now
+this screen all drive the identical `settings-form.ts` block through one new function,
+`wireGameTypeBlock()` — factored out of what used to be dialog-only code. Only Solitaire can be
+dealt outside the lobby, so the setup screen shows the other four types exactly as the dialog always
+has: present, disabled, with a note pointing at the Multiplayer door. Committing an answer — from
+either the dialog or the setup screen — goes through one `commitNewGame()`, which builds the URL and
+navigates; `start()` is still the only place that turns a URL into a game.
+
+Prefilling is deliberately left to the caller rather than folded into `wireGameTypeBlock` itself:
+the dialog opens on the game CURRENTLY IN PLAY, so redealing to compare keeps comparing against it,
+while the setup screen opens on the plain Solitaire defaults, since there is no live game yet to
+read.
+
+`index.html`'s door copy changed to match: "Start a game" reads "Set up a game" now, and the blurb
+states the floor (15, not "20 Revenue") since that is what a player is agreeing to before they deal.
+
+859 tests pass. **Not yet played in a browser** — verified by `tsc --noEmit`, the full suite, and
+reading the diff, not by loading `play.html` fresh and clicking through it.
+
+---
+
 ## 0.7.4 — 2026-08-29
 
 Three rules issues off the tracker, in the order Jesse asked for them: #13, #5, #19. All three are

TODO.md

@@ -5,280 +5,302 @@ not to — the point is that nothing quietly evaporates.
 
 Grouped by what kind of work it is — Next, Replay/Save Games, Bot Performance, Play Balance,
 Multiplayer, Display, Rules Questions, Other — and ordered within each by how much it is currently
-costing us. **Display** was split out of Other on 2026-08-22, when a session at the board produced
-seven items about the screen rather than the rules; the "most recent action" entry moved with it.
-Reorganized 2026-08-20 from a flat list; nothing below changed, only where it lives. Two duplicate
-entries (Heavy Grade orientation, the Local's coach) were merged into one each, and the industry-table
-item that had been sitting in a "these are all done" section without actually being done was moved out
-to Rules Questions.
+costing us. **Next is the queue**; the other sections are where the reasoning lives, and Next points
+into them rather than repeating them.
+
+**Reorganized 2026-08-30**, and the reason is worth recording because it will happen again. Nothing
+was wrong with the ordering — what had gone wrong is that the file had stopped tracking its own
+closures. Next was a chronological list of 43 entries with 26 struck through, and eight items in
+Display were still open `- [ ]` although Next already recorded them as superseded by Gitea#18 or
+shipped in v0.7.9. **An item closed in one place and left open in another is worse than no entry at
+all**, because it reads as live work. So: Next is now grouped by what each item is WAITING ON, the
+closed entries moved to a "Settled" block at the end of it, and the five dead Division-map drawing
+items collapsed into one that keeps only the two ideas which outlived them. 65 open items became 57.
+
+Earlier passes, kept: **Display** was split out of Other on 2026-08-22, when a session at the board
+produced seven items about the screen rather than the rules. Reorganized 2026-08-20 from a flat list.
+Two duplicate entries (Heavy Grade orientation, the Local's coach) were merged into one each, and the
+industry-table item that had been sitting in a "these are all done" section without actually being
+done was moved out to Rules Questions.
 
 ---
 
 ## Next
 
-Queued from the 2026-08-20 multiplayer planning session (reasoning in Multiplayer below), in order:
+**The live queue.** Every entry keeps the number it was given when it was raised — these are ids,
+not positions, and the rest of this file refers to them by number ("decide with item 15", "same
+drawing pass as 19-21"). Nothing is renumbered when something closes; settled entries move to
+[Settled, kept for the reasoning](#settled-kept-for-the-reasoning) at the end of this section instead.
 
-1. ~~**Fix the New Train phase car-placement round**~~ — done, see Multiplayer below.
-2. ~~**Unify victory conditions across solitaire, competitive and coop**~~ — done, see Multiplayer
-   below.
-3. ~~**Phase 2 of `docs/architecture/multiplayer.md` — server core**~~ — done, see Multiplayer below.
+Regrouped 2026-08-30 by **what each is waiting on**, which is the question actually asked of this
+list. It was a chronological queue of 43 entries, 26 of them struck through, and the live work had
+stopped being visible in it.
 
-Queued 2026-08-21, from playing the StartOS build:
+### Blocked on a table, not on code
 
-4. ~~**The lobby must offer every game parameter the solitaire New Game dialog does**~~ — done in
-   v0.6.0.
-5. ~~**Decide what the four `optionalRules` are**~~ — done in v0.6.0: `sisterTrains` deleted,
-   `employeeRotation` implemented, the other two were already live.
-6. ~~**Stop every release destroying every game in progress**~~ — done in v0.6.0, by replaying the
-   save rather than comparing version strings.
+The largest gap in the project, and it is not a coding gap. Four features shipped without a person
+ever meeting them.
 
-Queued 2026-08-22, from playing on StartOS:
+39. **NONE OF v0.7.4 HAS BEEN PLAYED BY A HUMAN.** The Yard Office offer, the Red Flag hold and its
+    out-of-phase prompt, and the loaded-Extra make-up rules are all tested end to end, packed, and
+    running on `phoenix.local` — and no person has met any of them at a board. Two are interruptions
+    that stop the Mainline Phase and put a question in front of somebody mid-thought, which is
+    exactly the kind of thing only play reveals.
+
+35. **Extended play has never been played at a real table.** **Verified live on phoenix.local,
+    2026-08-29**, against the installed v0.7.3:0 rather than in tests: a two-seat competitive game
+    (one human client, one bot) was dealt over the HTTP API with `days: 1`, played to the end of its
+    timetable, and reached `awaitingExtension` on Day 2 with `official = { win, winner 0,
+    daysElapsed }` frozen at Day 1 and votes `[null, null]`. Voting yes as seat 0 was accepted, the
+    bot followed as designed, and the game returned to `active` with `extraDays: 1` and the official
+    outcome **unchanged**. Both test games were deleted afterwards.
+
+    **The save carry-over claim was checked rather than asserted**: phoenix held five saves before
+    the update, of which `WHISTLE-4086` resumed and three were already refused by the 0.7.2 deck
+    change. After updating to 0.7.3 the log is identical — same game resumed with the same 7 intents,
+    same three refusals at the same move with the same code.
+
+    **What is still untested is what the item is named for: humans, at a table.** The multiplayer
+    vote has never been driven through two browsers — what a second player sees while waiting on a
+    first, and whether "waiting on Carol" is legible once Carol has closed her laptop, are still
+    unanswered.
+
+    **A HUMAN DID REACH IT ON 2026-08-30, AND IT WAS UNUSABLE — fixed in v0.7.9.** Jesse played a
+    solitaire game to the end and was never offered the extra Day. Nothing was wrong with the engine
+    or the Frame: `renderEnding` wrote the two buttons into `#actions` and then opened `#resultsdlg`,
+    which is **modal**, so the question sat underneath a dialog whose only control was Close. The
+    dialog asks it now. **The verification recorded above is exactly why this survived** — it was
+    driven over the HTTP API, which renders no dialog, so what was proven was that the SERVER
+    supports extended play, not that a player can reach it. Read that distinction into every
+    "verified on phoenix.local" line in this file.
+
+42a. **Nobody has clicked through the solitaire setup screen's own fields** and confirmed the dealt
+    game matches what was chosen. The screen took three attempts to become reachable at all (item 42
+    below); reachable is not the same as correct. Early item for the next play session, with #39.
+
+40. **A save from before v0.7.4 may not replay, and nobody has been told.** The Red Flags intent
+    changed shape, a make-up that was legal may now be refused, and a Yard Office arrival asks a
+    question no older history has an answer for. It fails safe — the server declines the save, names
+    the move and leaves the file untouched — and `WHISTLE-4086` did survive on `phoenix.local`, so it
+    is "may not" rather than "will not". Worth a line wherever the build is announced, and worth
+    knowing when a bug report arrives with a save that will not load.
+
+32. **Tell the 0.4.9 playtesters their saves are dead, before they find out.** The same shape as #40
+    but for the playtest line: the deck change shipped as v0.4.9h, so every save filed before it —
+    including the ones attached to Gitea#15 and #17 — stops replaying at its first `card.play`. They
+    fail safe and the files are kept, but nobody has been told. `PLAYTEST-0.7.4.md` (untracked, at
+    the repo root) is the note drafted for this and leads with it.
+
+### The screen — v0.8.0 candidates
+
+All six were raised by Jesse from play and all six live in **Display** below, where the reasoning is.
+Listed here in rough increasing cost. Two of them entangle with Gitea#20; see the note under each.
+
+28. **The game's settings belong in a card, not along the top line** — and show ALL of them, not the
+    four that fit. **Cheapest of the six**: nothing new has to be sent (`Frame` gained `mode` and
+    `optionalRules` in v0.7.0) and the renderer already exists — `rulesListHtml` (`settings-form.ts`)
+    draws exactly this list for the join preview and the seating screen.
+
+16. **Three explicit display options for the Office map** — always hidden, always on, auto-hide. All
+    three modes already exist (`districtMode`, persisted); only the BUTTON is a cycle, and it cannot
+    reach every one of them.
+
+17. **The same three options for the Division map**, which today cannot be hidden at all. `auto`
+    there means something different and useful — hide it now, bring it back at the end of the phase —
+    so it is a one-shot with an expiry rather than a standing rule, and probably should not share an
+    implementation with #16.
+
+23. **History: newest at the top?** Reversing is nearly free; the cap (`slice(-60)`) and the forward-
+    reading phase headings are not. Carries the timestamps question from #14. **Weaker if #15 lands.**
+
+15. **INVESTIGATE: a "most recent action" line under the status block.** The text already exists and
+    is already correct — this is placement, not content. **Overlaps #13 and should be decided with
+    it**: in solitaire "most recent" is the right unit, but in multiplayer what you missed is
+    everything that happened while you were WAITING, and one line showing only the last of them may
+    be the least useful line on the page.
+
+18. **Give every phase a visible beat.** The automatic phases are not too fast — they are never drawn
+    at all, because `pump` runs them all before the page renders once. **So this is not a `sleep`: it
+    needs an async pump that steps `advance()` one call at a time and renders between — which is the
+    same machinery Gitea#20 step 4 specifies.** Whichever is built first should be built for both.
+
+### Waiting on something outside the code
 
 7. **Make "Games in Progress" readable** — **ON HOLD, 2026-08-29 (Jesse).** StartOS 0.4.0.2 is
    expected to improve how action results are displayed, which is most of what makes this unreadable
-   — so wait and see what the platform fixes before rewriting the action around a limitation that
-   may be gone. Re-open it against 0.4.0.2 and re-read the output before designing anything.
-   Reasoning in Multiplayer below.
+   — so wait and see what the platform fixes before rewriting the action around a limitation that may
+   be gone. Re-open it against 0.4.0.2 and re-read the output before designing anything.
+
 8. **Give a player a way back into a game after losing their browser** — a fresh browser is still
-   locked out of a RUNNING game, even though the server knows who they are. Reasoning in Multiplayer
-   below; needs Jesse's call on whether a token in a URL is acceptable. **The lobby half of this was
-   fixed 2026-08-23** — a reload while seated no longer orphans the chair, and a seat can now be
-   given up rather than wedging the table.
+   locked out of a RUNNING game, even though the server knows who they are. **Needs Jesse's call on
+   whether a token in a URL is acceptable.** The lobby half was fixed 2026-08-23: a reload while
+   seated no longer orphans the chair, and a seat can be given up rather than wedging the table.
 
-Queued 2026-08-22, from a v0.4.9d gameplay-testing report (six bugs, forwarded by Jesse):
+13. **Watch the other players and the bots actually make their moves.** Settled 2026-08-29 as the
+    harder of the two readings — Jesse: "It's not fun to do my turn and have magic happen in the
+    background and then have to figure out what others did." So it is not the log-legibility fix. It
+    is the ordered, per-action presentation of everyone else's turns — **the same mechanism Gitea#20
+    step 4 specifies for the common board**, routed to a player's own screen as well. Jesse: "this
+    relates to issue #20 and will require a lot more thinking." **Do not start it standalone.**
 
-9. ~~**Two trains in one station answer to one button**~~ — done in v0.4.9e / the release below.
-10. ~~**A Freight House can unload the boxcar it just loaded; a platform can detrain the passengers
-    it just boarded**~~ — done in v0.4.9e / the release below.
-11. ~~**The Grocer's Warehouse ships and the Refinery receives**~~ — done in v0.4.9e / the release
-    below: both are one-way again.
-12. **PARTLY REPRODUCED: cars left behind when backing up over them** — see Rules Questions below.
-    Half of it turned out to be the second half of Gitea#17 and is fixed (2026-08-26): a train
-    pulling out through a 45° leg left its own cut standing. The "I can later drive right through
-    them" half is still unexplained and still needs a board from whoever filed it.
+33. **The second pass on the results screen — badges, and the brainstorm Gitea#16 asks for.**
+    **Needs Jesse and a conversation, not code, to start.** The first pass is in and reports
+    everything the Frame and the event tally know; what it deliberately does not have is the
+    interesting half. The raw material is already kept — `tally.trainsCompletedWithWork` is the
+    switching-master join, `longestStand` is the engine that sat on a siding — and because the
+    statistics are DERIVED from the event stream rather than recorded, a second pass can add any of
+    them retroactively to games already played and saved.
 
-Queued 2026-08-22, from the v0.4.9e gameplay-testing report filed as Gitea issues.
+36. **There is no per-Stage "this train did not move" signal, so "longest an engine sat on a siding"
+    cannot be answered.** The comment on Gitea#16 said `trainStoodStill` would supply it. That is
+    wrong, and was found by reading `advance.ts`: the event fires only for a train whose profile sets
+    `stopEarnsPoint` — the X18 Circus and nothing else — and `tray.stopPointClaimed` guarantees it
+    fires at most once per train per game, so a streak folded from it reads "1 Stage" for ever.
+    **What it would take:** a new event per Stage per stationary tray (cheap to emit, a lot of events
+    for a statistic nothing scores), or sampling live state on the Stage boundary the way `stats.ts`'s
+    funnel probe does — which the tally cannot do today, because it folds a batch of events AFTER
+    `advance` has already mutated past the moment they describe. **Settle with #33**, its only
+    consumer.
 
-12a. ~~**Gitea#4 — confirm that extra trains start properly**~~ — done in the release below. Either
-    Division Point, the Interchange with a direction chosen there, a Control Point gated by a new
-    `extraStart` house rule, and the Superintendent's hold on the way out of the Interchange yard.
-    Reasoning in `docs/rules/implications.md` §7.
+### Unscheduled
 
-12d. ~~**Gitea#7 — coach counts on four train cards**~~ — done in the release below. 1/2 Crack
-    Limited 3 coaches → 2, 5/6 The Sparrow 2 → 3. A change to the cards, so `Trains3.pdf` and the
-    transcription in `docs/rules/implications.md` §5 keep the original numbers with a footnote;
-    `src/engine/content.ts` and `docs/StationMaster-Home-Deck-v0.4.5.md` carry what the game plays.
+12. **PARTLY REPRODUCED: cars left behind when backing up over them.** Half of it turned out to be
+    the second half of Gitea#17 and is fixed (2026-08-26): a train pulling out through a 45° leg left
+    its own cut standing. **The "I can later drive right through them" half is still unexplained and
+    still needs a board from whoever filed it.** See Rules Questions below.
 
-12c. ~~**Gitea#6 — players may not discard train cards**~~ — done in the release below. Timetabled
-    and Extra alike; the forced play falls out of the hand limit rather than needing a mechanism of
-    its own. Reasoning in `docs/rules/implications.md` §6.2.
+14. **INVESTIGATE: stamp the history with wall-clock time** — even if nothing displays it yet.
+    Reasoning in Replay / Save Games below. **Half of it already exists server-side and is read by
+    nothing**, so the first job is to look at a real game's timings file rather than to build.
 
-12b. ~~**Gitea#2 — four porters, two passengers on the platform, and only one may be worked**~~ —
-    RULED AND FIXED in the release below, though not the way the report implies. The engine is
+15a. **Build documentation FROM the implementation, starting with a card reference.** Raised by Jesse
+    2026-08-22; reasoning in Other below. The prompt was finding train card data spread across five
+    documents of three different vintages, one of them superseded.
+
+41. **The bot cannot use the half of Red Flags a human would.** It takes the danger prompt
+    unconditionally and still plays zero flags in 200 games, because the prompt needs a colliding
+    arrival to coincide with holding the card. What it never does is plant a flag ON PURPOSE to buy a
+    Stage for switching, which needs it to know it wants time — a notion it does not have. Reasoning
+    and the measurement are under Bot Performance.
+
+### Standing practice — not items, rules for every release
+
+30. **Close each Gitea issue by hand, with a comment naming the commit that fixed it** and, where the
+    fix was not what the report implied, the ruling that decided it. Auto-closing leaves an issue with
+    no record of which commit or which release answered it. Done for every issue closed so far: #2,
+    #6, #8, #9 and #10 from v0.7.1; #3, #14, #15, #17 and #18 from v0.7.2 — including #15's REVERSAL
+    on review (the placement is legal; what was confirmed is that no train crosses the gap) and #14's
+    list of the ten unbuilt cards, so they do not vanish with the issue. The token and the API calls
+    are in the workspace's `AGENTS.local.md`.
+
+37. **The wrapper release sequence, unchanged since 0.7.2 and worth following exactly.** Tag the app,
+    fetch the tag into the wrapper's submodule, bump `current.ts` **in place** (the outgoing `up` has
+    been empty every time, `versions.md`'s common case — no new version file, no migration), rewrite
+    the release notes in all five locales, update `README.md` and `instructions.md`, then
+    `npm run check` / prettier / `make x86` / `make install`. `UPDATING.md` in the wrapper is the
+    authority and has not needed changing.
+
+    **Say which way games in progress go, in every locale, every time.** 0.7.2 broke them (206 cards
+    to 121, so a card id recorded under 0.7.1 refers to a different card or to none) and led with it;
+    0.7.3 and 0.7.4 carried them and led with that. It fails safe either way — `src/server/index.ts`
+    refuses a save the rules reject, names the move it stopped at, and leaves the file untouched, so
+    an operator can put the old version back to finish a game that matters.
+
+### Settled, kept for the reasoning
+
+Closed items, newest first. Kept because several of them are the only record of a ruling or a lesson;
+the numbers stay so cross-references above and below still resolve.
+
+43. ~~**"Waiting on" said nobody while the game was stopped on the Superintendent.**~~ — done
+    2026-08-30 in v0.7.9. `Frame.actor` carried `clock.currentActor`, null for the whole Mainline
+    Phase, so all three interruptions (clearance, Yard Office, Red Flag) reported that the Division
+    was running itself. It carries `actingPlayer` now and a new `awaiting` field says what the
+    question is and which train it is about. **Worth knowing for the next Frame field:** the engine
+    had the right answer in `actingPlayer` since the Gitea#5 refactor, and the Frame simply did not
+    carry it. A view that reads one field of `clock` directly is a place this can happen again.
+
+34. ~~**`replay.ts` printed a raw outcome enum, exactly as the results screen used to.**~~ — done
+    2026-08-30 in v0.7.9. Its summary line rendered "loss — revenueFloor", the same defect Gitea#16
+    was filed about, alive in the dev-side viewer a release after the playable page was fixed.
+    `panels.ts`'s `reasonSentence` is exported and shared rather than reimplemented, so the replay and
+    the results screen cannot explain one ending two different ways; the drift test maps `win`/`loss`
+    to `won`/`lost` so it still checks agreement rather than spelling.
+
+29. ~~**Put the Fedora at the right-hand end of the phase row.**~~ — done 2026-08-30 in v0.7.9.
+    Details, and why the name does NOT go inside the Supervisor Shift pill, in Display below.
+
+42. ~~**Solitaire must ask before it deals, the same way multiplayer's lobby already does.**~~ — done
+    2026-08-29 in v0.7.5, and then fixed three more times. Jesse: "let the user choose their options
+    like the start of a multiplayer game"; "asking first is the only path." A `#solitairesetup` screen
+    asks the full shared block before a genuinely fresh visit deals anything; a saved game, an
+    explicit `?seed=`, or a URL a Deal already wrote all skip past it. The in-game dialog, the lobby
+    and this screen share one `wireGameTypeBlock()`/`commitNewGame()` pair.
+
+    **THE LESSON, AND IT IS THE MOST EXPENSIVE ONE IN THIS FILE.** The same report came back three
+    times, and each of the first two "fixes" corrected something real that was not the reported fault.
+
+    - **v0.7.6** — the splash's "Play solitaire" door landed in a leftover Co-op lobby: `start()`
+      checked a remembered multiplayer session before looking at solitaire's own state. Real bug.
+      Not the one reported.
+    - **v0.7.7** — every packaged build published the same cache-bust key (`?v=nogit`, because the
+      `.s9pk` build has no `.git` for `git rev-parse`) and the server sent no `Cache-Control`, so
+      neither release ever reached the browser that asked for it. Real bug. Still not the one
+      reported. **When a fix appears to have had no effect, check that it ARRIVED before
+      re-diagnosing it** — a hard reload would have answered it on the first report.
+    - **v0.7.8** — the actual fault: v0.7.5 skipped the setup screen whenever a save existed ("a
+      saved game is a game to resume"), so any browser that had ever played solitaire could never
+      reach it again. The private window that seemed to vindicate v0.7.7 simply had no save.
+
+    **Each of the three was reported as verified, and each verification read what the SERVER served
+    rather than exercising the path with the state a returning player actually has.** The thing that
+    worked was a failing test written before the fix. Next time a report repeats: reproduce the
+    user's state first, and treat "I verified it" as unearned until something failed the way they
+    described.
+
+38. ~~**Gitea#13, #5 and #19 — three rules corrections.**~~ — done 2026-08-29 in v0.7.4 (`5e34c73`,
+    `2280276`, `19a6a47`), wrapper `085b88b` as `0.7.4:0`, installed on `phoenix.local`. Each issue
+    carries a comment naming its commit and what was ruled. What they left behind is #39, #40 and #41
+    above.
+
+31. ~~**Bump the StartOS wrapper to 0.7.2.**~~ — done 2026-08-26 (`1bfea8d`). Folded into the
+    standing practice at #37 above, which is where the sequence now lives.
+
+25-27. ~~**Three Division-map drawing reports from the v0.7.0 build**~~ — no track geometry, the
+    buffer stop pointing the wrong way at two players, and east not always being to the right. All
+    answered by **Gitea#18**; the last was FIXED OUTRIGHT by it and is the reason a single row won.
+    Collapsed into one entry in Display below.
+
+19-22, 24. ~~**Five drawing items against the wrapped Division map**~~ — vertical track art, the
+    inter-row connector, overflowing captions, filling the dead centre, and seating the viewer at the
+    bottom. All **SUPERSEDED by Gitea#18**, which replaced the layout rather than fixing the drawing.
+    Two ideas survive them and are recorded in Display below: "SHARED in the middle, YOURS on the
+    right" (now Gitea#20's territory), and why rotating the map to seat a viewer was refused.
+
+9-11, 12a-12g. ~~**The v0.4.9d and v0.4.9e gameplay-testing reports**~~ — twelve bugs, all shipped in
+    v0.4.9e / v0.7.1. Two trains in one station answering to one button; a Freight House unloading the
+    boxcar it just loaded; the Grocer's Warehouse shipping and the Refinery receiving; Gitea#4 extra
+    train starts; #7 coach counts on four train cards; #6 then #9 on discarding train cards (#9
+    superseded #6 three days later — a Timetabled train may be tossed face-up to a Department slot,
+    an Extra may not); #8 the per-diem train and the caboose; and #10 the Day-rollover dialog.
+    Reasoning for each is in `docs/rules/implications.md` and `CHANGELOG.md`.
+
+    **Gitea#2 carries a RULING worth keeping.** Four porters, two passengers on the platform, and
+    only one may be worked — RULED AND FIXED, but not the way the report implied. The engine is
     faithful to the written rules at every step; what bites is that BOTH directions of porter work
     move coaches one-way into a Classification Yard that comes back only when the Division Yard is
     bare of all ~60 cars. Sixteen coaches in the game, and the reported save runs dry on Day 5 with
-    eight of them stranded in Classification. **Jesse's ruling: the shortage stays** — "it is
-    possible to run out, that's part of the strategy" — so the three balance options in Play Balance
-    below are declined, not deferred. What was actually wrong is that the game said NOTHING: a Porter
-    action that cannot be taken is simply absent from the menu, and the "why is nothing moving?"
-    panel covered freight facilities only. That half is fixed.
+    eight stranded. **Jesse's ruling: the shortage stays** — "it is possible to run out, that's part
+    of the strategy" — so the three balance options in Play Balance below are DECLINED, not deferred.
+    What was actually wrong is that the game said NOTHING: a Porter action that cannot be taken is
+    simply absent from the menu, and the "why is nothing moving?" panel covered freight facilities
+    only. That half is fixed.
 
-12e. ~~**Gitea#8 — the per-diem train could not couple a caboose**~~ — done in the release below.
-    `ROLLING_STOCK_SUPPLY` mints all six cabooses `loaded: true` because §2.2's "coloured is loaded,
-    white is empty" doubles as a PIECE COUNT there and there is no white caboose. X22 Pee-Dee, whose
-    whole card is "may only pick up MTs", read that literally and refused every caboose including the
-    one it was made up with — set it out and the train was stranded, which made it unplayable rather
-    than merely restricted. A caboose carries the crew, not freight, so it is never a load.
-
-12g. ~~**Gitea#9 — a Timetabled train may be discarded**~~ — done in the release below, and it
-    SUPERSEDES Gitea#6 (item 12c above), shipped three days earlier. A Timetabled train may be tossed
-    face-up to a Department slot, where a rival may pick it up — which needed no machinery, since
-    that is where every discard already goes. An Extra still may not: it never joins the timetable,
-    so it can never be what jams it. On this line the Timetabled half is a New Game setting
-    (`discardTimetabled`, on by default), because Jesse's reasoning is about games run longer than
-    five Days; the 0.4.9 line takes the plain rule. Reasoning in `docs/rules/implications.md` §6.2.
-
-12f. ~~**Gitea#10 — a dialog when the Day rolls over**~~ — done in the release below. "Hard to keep
-    track of time." Nothing on screen was wrong — the clock, the turn chart and the timetable all
-    said which Day it was — but a Day turns over inside the phases that run themselves, so it passes
-    between one click and the next, and the two transient signals the page had (the phase banner at
-    2.6s, the announcement flash at 4.2s) are both gone before a player reading the board notices.
-    A modal stops and waits, and carries the standings, the Days left and the combined target.
-    Suppressed on the first frame, on Undo stepping back across a rollover, and on the Day the game
-    ends — the outcome panel is the thing to read then.
-
-13. **Watch the other players and the bots actually make their moves** — raised by Jesse
-    2026-08-22, and **settled 2026-08-29 as the harder of the two readings**: "I want to be able to
-    watch other players and bots make their moves. It's not fun to do my turn and have magic happen
-    in the background and then have to figure out what others did."
-
-    So this is not the log-legibility fix. It is the ordered, per-action presentation of everyone
-    else's turns — **the same mechanism Gitea#20 step 4 specifies for the common board**, routed to a
-    player's own screen as well. Jesse: "this relates to issue #20 and will require a lot more
-    thinking." Do not start it as a standalone piece; it wants designing with #20. Reasoning in
-    Multiplayer below.
-
-14. **INVESTIGATE: stamp the history with wall-clock time** — even if nothing displays it yet, so
-    "how long did that turn take" can be answered afterwards. Reasoning in Replay / Save Games below.
-    Half of it already exists server-side and is read by nothing.
-
-15. **INVESTIGATE: a "most recent action" line under the status block** — above the Division map,
-    saying what just happened in the same words the history uses. Reasoning in Display below;
-    overlaps item 13 and should be decided with it.
-
-15a. **Build documentation FROM the implementation, starting with a card reference** — raised by
-    Jesse 2026-08-22. Reasoning in Other below. The prompt for it was finding train card data spread
-    across five documents of three different vintages, one of them superseded.
-
-Queued 2026-08-22, from a session looking at the screen rather than the rules. All Display below.
-
-16. **Three explicit display options for the Office map** — always hidden, always on, auto-hide. All
-    three modes already exist; only the BUTTON is a cycle, and it cannot reach every one of them.
-17. **The same three options for the Division map**, which today cannot be hidden at all. Auto there
-    means something different and useful: hide it now, bring it back at the end of the phase.
-18. **Give every phase a visible beat.** The automatic phases are not too fast — they are never
-    drawn at all, because `pump` runs them all before the page renders once.
-19. ~~**Turn the track art vertical on a Division card laid vertically**~~ — **SUPERSEDED by
-    Gitea#18** (Jesse, 2026-08-26). There are no vertical lanes any more: the Division draws as a
-    single row, west to east.
-20. ~~**Run the inter-row connector round the OUTSIDE**~~ — **SUPERSEDED by Gitea#18.** No second
-    row, so nothing to connect.
-21. ~~**The Division Point captions overflow the map**, and the buffer stops point the wrong way once
-    the route wraps~~ — **SUPERSEDED by Gitea#18.** Nothing wraps; both ends face outward.
-22. ~~**Fill the dead centre of the Division map with the common board**~~ — **SUPERSEDED by
-    Gitea#18.** A row has no centre to fill.
-23. **History: newest at the top?** Plus the timestamps question from item 14, which lands here.
-24. ~~**Put the viewer's own district at the BOTTOM of the Division map** and wrap the table around
-    them~~ — **DEFINITIVELY SUPERSEDED by Gitea#18** (Jesse's word, 2026-08-26). A single row and a
-    table wrapped around the viewer cannot both hold, and the row wins: being able to rely on east
-    meaning right is worth more than being seated at the table.
-
-Queued 2026-08-23, from Jesse playing the v0.7.0 build on StartOS. **All three were the same drawing
-pass as 19-21 and 24 above, and all three are answered by Gitea#18 rather than fixed:**
-
-25. ~~**The Division map does not draw track geometry at all**~~ — **SUPERSEDED by Gitea#18.** The
-    Division map stops drawing office-area detail altogether, so there is no Running Track on it to
-    draw geometry for. The geometry belongs to the Office map, which already draws it.
-26. ~~**CONFIRMED IN PLAY: the buffer stop points the wrong way** at two players~~ — **SUPERSEDED by
-    Gitea#18**, with item 21.
-27. ~~**CONFIRMED IN PLAY: east is not always to the right.**~~ — **FIXED OUTRIGHT by Gitea#18**, and
-    the reason it wins over item 24. One row means east is always to the right.
-
-28. **INVESTIGATE: move the game's settings off the top line and into a card of their own** — and
-    show ALL of them, not the four that fit. Reasoning in Display below.
-
-29. **Put the Fedora at the right-hand end of the phase row**, with (or in) the Supervisor Shift
-    pill. Reasoning in Display below.
-
-Queued 2026-08-25, from releasing 0.7.1 / 0.4.9g. **Both are blocked on the two commits being made
-and pushed** — they were still uncommitted when the session ended.
-
-30. ~~**Close the Gitea issues by hand, each with a comment naming the commit that fixed it.**~~ —
-    done, and done again for v0.7.2 / v0.4.9h (2026-08-26). #2, #6, #8, #9 and #10 carry their
-    comments from the v0.7.1 release, including the pointer on #6 saying #9 superseded it. #3, #14,
-    #15, #17 and #18 now carry theirs, each naming `2ab25e3` and `e9683cc` and the version each
-    shipped in — and, where the fix was not what the report implied, the ruling that decided it:
-    #15 was REVERSED on review (the placement is legal; what was confirmed is that no train crosses
-    the gap), and #14's comment lists the ten unbuilt cards that are held out, so they do not vanish
-    along with the issue.
-
-    **Keep doing this.** Auto-closing leaves an issue with no record of which commit or which release
-    answered it. The token and the API calls are in the workspace's `AGENTS.local.md`.
-
-31. ~~**Bump the StartOS wrapper to 0.7.2.**~~ — done 2026-08-26 (`1bfea8d`). Submodule pinned to
-    `v0.7.2`, `current.ts` at `0.7.2:0`, release notes rewritten in all five locales, `README.md` and
-    `instructions.md` updated. No new version file and no migration. Verified: `npm run check` clean
-    and `make x86` packs as `v0.7.2:0`.
-
-    **This is the first release that does NOT carry games in progress**, and the release notes lead
-    with it in every locale. A save is a seed plus the moves played; the deck going from 206 cards to
-    121 means a card id recorded under 0.7.1 refers to a different card or to none, so a save stops
-    replaying at its first `card.play`. There is nothing to migrate — those moves were made against a
-    deck that no longer exists — and it fails safe: `src/server/index.ts` refuses to resume a save the
-    rules reject, names the move it stopped at, and leaves the file untouched, so an operator can put
-    0.7.1 back on to finish a game that matters.
-
-32. **Tell the 0.4.9 playtesters their saves are dead, before they find out.** The same deck change
-    shipped as v0.4.9h, so every save filed before it — including the ones attached to Gitea#15 and
-    #17 — stops replaying at its first `card.play`, and any game a tester has in progress will be
-    declined on restart. They fail safe and the files are kept, but nobody has been told. Worth a
-    line wherever the tester build is announced, and worth knowing when the next bug report arrives
-    with a save that will not load.
-
-Queued 2026-08-29, from building Gitea#11 and #16 (both shipped in v0.7.3, main only):
-
-33. **The second pass on the results screen — badges, and the brainstorm Gitea#16 asks for.** The
-    first pass is in and reports everything the Frame and the event tally know. What it deliberately
-    does not have is the interesting half: "maybe create badges for anything interesting that
-    happened… there should be a whole conversation brainstorming session on what are the things that
-    might be interesting for people to be aware of at the end of the game." The raw material is
-    already being kept — `tally.trainsCompletedWithWork` is the switching-master join, `longestStand`
-    is the engine that sat on a siding — and because the statistics are DERIVED from the event stream
-    rather than recorded, a second pass can add any of them retroactively to games already played and
-    saved. Needs Jesse and a conversation, not code, to start.
-
-34. **`replay.ts` prints a raw outcome enum, exactly as the results screen used to.** Its summary
-    line is `` `${o.result} — ${o.reason}` ``, which renders "loss — revenueFloor" — the same defect
-    Gitea#16 was filed about, in the dev-side replay viewer rather than the playable page. The
-    sentences now exist (`panels.ts`'s `reasonSentence`), but they are written against a `Frame` and
-    the replay recorder has a `GameState`, so it is a small refactor rather than a one-line swap. Not
-    done in v0.7.3 because nothing about it is player-facing and the change earns its own look.
-
-35. **Extended play has never been played at a real table** — but it now works on a real server.
-    **Verified live on phoenix.local, 2026-08-29**, against the installed v0.7.3:0 rather than in
-    tests: a two-seat competitive game (one human client, one bot) was dealt over the HTTP API with
-    `days: 1`, played to the end of its timetable, and reached `awaitingExtension` on Day 2 with
-    `official = { win, winner 0, daysElapsed }` frozen at Day 1 (`config.days`) and votes
-    `[null, null]`. Voting yes as seat 0 was accepted, the bot followed as designed, and the game
-    returned to `active` with `extraDays: 1` and the official outcome **unchanged**. The tally rode
-    the Frame to the client. Both test games were deleted afterwards; the box is back to Jesse's own
-    `WHISTLE-4086` and the `FREIGHT-3230` lobby.
-
-    **The save carry-over claim was also checked rather than asserted**: phoenix held five saves
-    before the update, of which `WHISTLE-4086` resumed and three were already refused by the 0.7.2
-    deck change. After updating to 0.7.3 the log is identical — same game resumed with the same 7
-    intents, same three refusals at the same move with the same code.
-
-    **What is still untested is the part the item is named for: humans, at a table.** Nobody has sat
-    down and played a game off the end of its timetable, and the multiplayer vote has never been
-    driven through two browsers — what a second player sees while waiting on a first, and whether
-    "waiting on Carol" is legible once Carol has closed her laptop, are still unanswered. v0.7.3 is tested — engine, server,
-    replay, an all-bot regression — and compiled and exercised headlessly, but nobody has sat down,
-    run a game off the end of its timetable and voted. The multiplayer vote in particular has only
-    been driven through `session.intent`, never through two browsers: what a second player sees while
-    waiting on a first, and whether "waiting on Carol" is legible when Carol has closed her laptop,
-    are both unanswered. Worth being the first thing the next play session does.
-
-36. **There is no per-Stage "this train did not move" signal, so "longest an engine sat on a siding"
-    cannot be answered.** Gitea#16 asks for it and the comment on that issue said `trainStoodStill`
-    would supply it, "emitted per Stage, so a run of them is exactly the streak you describe". That
-    is wrong, and was found only by reading `advance.ts` while building the tally: the event fires
-    for a train whose profile sets `stopEarnsPoint` — the X18 Circus and nothing else — and
-    `tray.stopPointClaimed` guarantees it fires at most once per train per game. A streak folded from
-    it reads "1 Stage" for ever, which is what v0.7.3 built and then removed.
-    **What it would take:** either a new event emitted per Stage per stationary tray (cheap to emit,
-    but it is a lot of events for a statistic nothing scores), or sampling live state on the Stage
-    boundary the way `stats.ts`'s funnel probe does — which the tally cannot do today, because it
-    folds a batch of events AFTER `advance` has already mutated past the moment they describe. Worth
-    settling with the badge pass (#33) rather than on its own, since that is the only consumer.
-
-37. ~~**Bump the StartOS wrapper to 0.7.3.**~~ — done 2026-08-29 (`74aea24` in
-    `station-master-startos`). Submodule pinned to `v0.7.3` (`45580d8`), `current.ts` at `0.7.3:0`,
-    release notes in all five locales, `README.md` and `instructions.md` updated. No new version file
-    and no migration — the outgoing `0.7.2:0`'s `up` was empty, `versions.md`'s common case, so
-    `current.ts` bumped in place. Verified: `npm run check` clean, prettier clean, `make x86` packs
-    as `v0.7.3:0`. **NOT installed on a box and not played** — see #35.
-
-    **Unlike 0.7.2, games in progress survive this one**, and both docs lead with it. No card data
-    changed and the engine changes are additive, so every intent in a 0.7.2 save is still legal;
-    proven rather than assumed by `test/harness.test.ts`, which replays the three files in
-    `public/replays` (all recorded under an older ruleset) and asserts every intent still applies.
-
-    **The v0.7.3 tag is made locally and signed, but not pushed** — agents cannot push to
-    `origin`. Until Jesse pushes it, the tag exists only here and in the wrapper's submodule clone;
-    the recorded pin is the commit SHA, which is already on `origin`, so the wrapper is not blocked
-    on it.
+1-6. ~~**The 2026-08-20 multiplayer planning queue and the first StartOS play session**~~ — the New
+    Train phase car-placement round, unified victory conditions, Phase 2 server core, the lobby
+    offering every game parameter, deciding what the four `optionalRules` are (`sisterTrains`
+    deleted, `employeeRotation` implemented, the other two already live), and stopping every release
+    from destroying every game in progress — done by replaying the save rather than comparing version
+    strings. All shipped in v0.6.0; reasoning in Multiplayer below.
 
 ---
 
@@ -541,7 +563,19 @@ The replay viewer, the save format, and how a game gets shared.
 
 ## Play Balance
 
-- [ ] **FREIGHT GOT SCARCER WHEN THE MAINLINE WENT ONTO REGIONS, and nobody knows why yet.**
+- [ ] **A SOLITAIRE GAME CAN NOW END ON THE COLLISION FLOOR (v0.7.9, 2026-08-30) — small, but every
+      full-length figure in this file predates it.** §3.4's check was gated on
+      `mode === 'competitive' || mode === 'coop'`, so the two collision limits were live settings in
+      the solitaire dialog that did nothing; Jesse's ruling was that they should do what they say, and
+      the gate is gone. Measured over 200 standard developer-bot games: `loss/collisionFloor` fires in
+      **1 game in 200**, Days played 5.00 → 4.98 mean with a **minimum of 1**, and collisions per game
+      unchanged at 0.14 (max 3).
+
+      Nothing here needs acting on — the effect is smaller than the noise on every number in this
+      section — but it is a new way for a run to be short, so **a mean taken over games that all ran
+      five Days is no longer quite what is being sampled.** Worth remembering when the rebalance pass
+      re-measures, and worth watching if a future change makes collisions more common, because the
+      cost of one stops being "−1 Revenue" and starts being "the game".
       Gitea#3, measured 2026-08-26 across the same 100 games: freight share of gross fell **8% → 5%**,
       and completed freight loads went from something a 40-game sample caught reliably to needing
       200 — on `sim.test.ts`'s seeds, 40 games now yield 0 loads, 80 yield 3, 120 yield 10, 200
@@ -1341,11 +1375,19 @@ Deferred while planning the server; decisions and reasoning are in `docs/archite
 
 ## Display
 
-- [ ] **THE BOARD STILL DOES NOT SAY WHICH WAY A HEAVY GRADE CLIMBS.** RAR, twice: "grade should
+- [x] **~~THE BOARD STILL DOES NOT SAY WHICH WAY A HEAVY GRADE CLIMBS.~~** — done 2026-08-30 in
+      v0.7.9 (`cf018b4`). A brown wedge in the card's lower right rising toward the climb, with an
+      arrow lying along its slope. **It is Gitea#18 that made this drawable**: east is always to the
+      right on a single-row map, so a wedge can be read without a compass — under the wrapped layout
+      the same wedge would have pointed a different way at every seat. Orientation was measured
+      rather than assumed (400 seeds x 4 player counts, 535 grades, 0 without one, 276 east / 259
+      west). Original note below.
+
+      **The board did not say which way a Heavy Grade climbs.** RAR, twice: "grade should
       tell you which way is up." `gradeUp` is dealt at setup and drives which of Helpers or
       Brakeman/Airbrakes can ever pay, and Gitea#3 made it matter more — the modifiers now move a
       train's STARTING REGION, so playing the wrong one is three Stages of climb instead of two. The
-      tooltip says it (`mainlineDescription`); the map does not. Untouched by Gitea#3, which was
+      tooltip says it (`mainlineDescription`); the map did not. Untouched by Gitea#3, which was
       about the rules rather than the drawing.
 
 
@@ -1478,202 +1520,59 @@ What is on the screen and where. Split out of Other 2026-08-22; the rules are el
       **The open questions.** Which of the six stay on the top line — Revenue and the objective are
       glanced at constantly and clearly belong there, the seed and the code almost never are. Whether
       the card folds like `#district` does or is always open. Whether the collision counts belong in
-      it or beside the objective, since they are a live score rather than a setting. And it interacts
-      with the Display items about the right column and the middle of the Division map (item 22): if
-      the shared board moves to the centre, the right column is "yours", and a settings card is not
-      yours — it is the table's.
+      it or beside the objective, since they are a live score rather than a setting. And it inherits
+      the one principle that outlived the wrapped-map items (settled below): **SHARED in the middle,
+      YOURS on the right.** A settings card is not yours — it is the table's — so if the common board
+      of Gitea#20 ever comes back onto this page, the settings card belongs beside it rather than in
+      the right-hand column.
 
-- [ ] **INVESTIGATE: the Fedora belongs at the right-hand end of the phase row.** Raised by Jesse
-      2026-08-23, the day after it was added: "on the next row down, we recently added Superintendent
-      and saying who's got the Fedora. That information should be all the way on the right, where
-      currently it says 'Supervisor Shift'. Would it be possible to say 'Supervisor Shift — and then
-      the player name, who's the current supervisor'? If not, just moving the Superintendent and the
-      Fedora graphic to the right-hand side of Supervisor Shift is probably a better place for that."
+- [x] **~~INVESTIGATE: the Fedora belongs at the right-hand end of the phase row.~~** — done
+      2026-08-30 in v0.7.9 (`Next` #29). Jesse's fallback was the one taken: `.tc-super` moved to the
+      end of the row, after the pills, and wrapping under them rather than squeezing the chips on a
+      narrow screen.
 
-      **Where it is now.** `turnChartHtml` (`sim/turnchart.ts`, shared by the play screen and both
-      replay viewers) lays out four blocks in a row: the Day/Stage/clock, the phase, "waiting on
-      <name>", then the Fedora chip (`.tc-super`), and last the `<ol class="tc-phases">` of five phase
-      pills. So the Fedora sits in the middle of the row, immediately before the pills.
+      **The appealing version was tried and rejected.** Putting the name INSIDE the Supervisor Shift
+      pill reads as "this phase belongs to that player", which is not what the Fedora means — the
+      office holds the clearance ruling and starts every round, in every phase. Worth knowing if
+      anyone proposes it again.
 
-      **Putting the name IN the Supervisor Shift pill is the appealing version and needs thought.**
-      The pills are a WHERE-ARE-WE indicator — each lights violet while its phase is running and dims
-      once it is done — so a name inside one may read as "this phase belongs to that player", which is
-      not what the Fedora means (the office holds the clearance ruling and starts every round, in
-      every phase). The pill is also the one whose tooltip already explains the hat passing every
-      third Stage, which is why it is the natural home. Worth trying both and looking at them.
+      **The row was looked at as a whole**, per the note left here at the time, because item #15's
+      "most recent action" line still wants space in the same strip. Nothing was taken; the row has
+      room for one more block at the widths tested.
 
-      **The fallback Jesse names** — move `.tc-super` to the end of the row, after the pills — is a
-      two-line change and safe. Neither should be done without looking at the row as a whole: it is
-      the most-glanced-at strip on the page, and item 15's "most recent action" line wants space in
-      the same place.
 
-- [ ] **THE DIVISION MAP DRAWS NO TRACK GEOMETRY — a turnout is indistinguishable from a straight.**
-      Reported by Jesse 2026-08-23, playing the v0.7.0 build on StartOS: he upgraded a straight on his
-      Running Track to a turnout and "did not see the division map on my side updated. And in the
-      opponent's screen, they did not see any update in their division map either."
+- [x] **~~Five drawing items against the wrapped Division map~~ — ALL ANSWERED BY Gitea#18**
+      (closed 2026-08-26), which replaced the layout rather than fixing the drawing. They were:
+      the map drawing no track geometry; vertical track art on a card laid down a side lane; the
+      inter-row connector routed round the outside with angled corners; the Division Point captions
+      overflowing and the buffer stops pointing the wrong way once the route wrapped; and seating the
+      viewer at the bottom with the table wrapped around them. All five were the same pass, and
+      Jesse's instruction was to do them together — which is what made replacing the layout the
+      cheaper answer than drawing it five ways.
 
-      **It did update. The whole visible change is one word.** Measured rather than reasoned about —
-      a real two-player game played out ~260 turns, then a running-track straight upgraded to a
-      turnout, diffing the drawn SVG on both seats:
+      **What the single row settled.** The route is one line, west on the left and east on the right,
+      at every seat count (`board-svg.ts`). Both buffer stops face outward, so the two "wrong way"
+      reports cannot recur. The captions sit under their own Division Point rather than hanging off
+      the ends, so nothing clips. And **east is always to the right** — the report that decided it,
+      and the reason the horseshoe lost to a row that is 1,580px wide at four players against 842.
+      It is also what made the Heavy Grade wedge drawable at all (item above).
 
-      ```
-      own division view : CHANGED
-      other's view      : CHANGED
-      drawn SVG         : CHANGED
-        removed: <text class="bs-name" ...>straight</text>
-        added  : <text class="bs-name" ...>turnout</text>
-      ```
+      **The Division map no longer draws office-area detail**, so there is no Running Track on it to
+      draw turnout geometry for. That belongs to the Office map, which reads `cell.links` and already
+      draws it properly.
 
-      **Why.** `divisionSvg` draws the SAME generic rail on every Running Track cell —
-      `rail(c.x + 6, c.y + 32, c.x + c.w - 6)`, unconditionally — and prints the card's name under it.
-      So a straight, a curve and a turnout are pixel-identical on the map and only the caption
-      differs. `officeSvg` does it properly for the district grid: it reads `cell.links` and draws the
-      real geometry, 45° legs included.
+      **Two things survive their items and are NOT closed with them:**
 
-      **The data is already there and unused.** `RunningCardView` carries
-      `links: connectionsFor(card)` — the same field `officeSvg` draws from — and `divisionSvg` never
-      reads it. So this is a rendering change with no plumbing behind it.
+      - **"SHARED in the middle, YOURS on the right"** — Jesse's principle from the dead-centre item,
+        and worth keeping as the rule that decides where anything new on the screen belongs, even
+        though a row has no centre to fill. It is now **Gitea#20**'s territory: the common board is
+        the shared thing, and it got its own display rather than a hole in the map.
+      - **Why seating the viewer at the bottom was refused**, since it will be proposed again: the
+        Division is a LINE, and the open gap between the two buffer stops is the only thing that says
+        so. Rotate the map to seat a viewer and that gap moves with them — sometimes behind them,
+        out of the eye's path, which is exactly where the one feature saying "this is not a circle"
+        must not go. Jesse called it definitively superseded, 2026-08-26.
 
-      **What is NOT wrong, checked at the same time:** a card laid anywhere BELOW the Running Track
-      correctly changes nothing on the map — the map is the through route between the Limits, and
-      district interiors live in the Office Area grid. Of 4,000 bot steps in a two-player game, all
-      15 running-row plays changed both players' maps and none of the 19 below-the-row plays did. The
-      delta is not dropping anything: `deltaFrame` JSON-compares `division` and sends it whenever it
-      differs.
-
-      **Do it with items 19, 20, 21 and 24** — Jesse's instruction, 2026-08-23. Drawing real geometry
-      is the same pass as turning that geometry through 90° down a side lane, and both are wasted
-      work if the map is later rotated to seat the viewer at the bottom.
-
-- [ ] **INVESTIGATE: turn the track art vertical on a Division card laid vertically.** Raised by Jesse
-      2026-08-22: a side lane "looks like a bunch of disconnected left-right tracks stacked one on top
-      of another instead of looking like a continuous track."
-
-      **It does, and the cause is one line.** `divisionSvg` draws every cell's rail with
-      `rail(c.x + 6, c.y + 32, c.x + c.w - 6)` — horizontal, unconditionally, whatever side of the
-      table the cell was laid on. **`railV` already exists**, is already used for the connector
-      between stacked cells, and takes the same shape turned ninety degrees. So the cell needs to know
-      which lane it is in (`dir[i]` is `'top' | 'right' | 'bottom' | 'left'`, known at layout time and
-      currently not stored on the cell) and pick the one that matches.
-
-      Everything else inside a side cell — the name, the capacity line, the train chips — is laid out
-      horizontally too, so this is bigger than swapping one call: it is deciding whether a side cell is
-      a rotated card or a differently-arranged one. Worth a sketch before any code.
-
-- [ ] **INVESTIGATE: run the inter-row connector round the outside, as rail, with angled corners.**
-      Raised by Jesse 2026-08-22. **CONFIRMED IN PLAY 2026-08-23**, and it is worse than a drawing
-      complaint: "as a train traverses the board in a multiplayer game, east is not always to the
-      right. Sometimes, for train direction, east might be south, west, or north as it traverses the
-      different players." The route wrapping through the lanes is exactly this, and a player reading
-      direction off the screen is being told the wrong thing — not merely an ugly corner.
-
-      **Where it comes out is wrong.** The turn between two sides of the table is drawn from
-      `a.y + CH` — the BOTTOM edge of the last cell in the top row — across to `b.y`, the TOP edge of
-      the first cell in the next. So the route leaves the top row's rightmost Limits downward out of
-      its underside, when it should leave from its right-hand side; and on the bottom row it enters
-      through the top edge when it should come in from the right. Jesse: "it should be on the outside
-      circumference of the display."
-
-      **What it is drawn WITH is wrong too.** That connector is a single `<path class="bs-turn">`
-      — one plain polyline — while every other join on the map is `rail()` or `railV()`, two rails
-      with cross ties. The connector between two stacked cells down a side already uses `railV`, which
-      is why the vertical run looks like track and the corners do not.
-
-      **And the corners want an angled piece**, Jesse's wish-list item: a 45° rail between the
-      horizontal and vertical runs rather than a right-angled elbow, matching how the district's own
-      track geometry works (everything leaving a card's north or south edge does so at 45°). That
-      would want a `railD` beside `rail` and `railV`.
-
-      All three are the same drawing pass and should be done together. Note the current elbow routes
-      through `(ay + by) / 2` — the middle of the board — which is the space the item below wants to
-      fill, so these two interact.
-
-- [ ] **INVESTIGATE: the Division Point captions overflow, and the buffer stops point the wrong way
-      once the route wraps.** Raised by Jesse 2026-08-22. **CONFIRMED IN PLAY 2026-08-23** at two
-      players: "the eastern division point has the end marker to the right, so it overlays where the
-      track is into the person's area, instead of off the left at the actual end of the track." The
-      stop is drawn past the cell's right-hand edge whatever lane the cell ended up in, so at a
-      wrapped route it points back into the board — over a district, not away from the line.
-
-      **The captions.** "west end · in and out" and "east end · in and out" are centred under their own
-      cell at `x = c.x + c.w / 2`. That was a deliberate fix — they used to hang off the outside of
-      the board and print clipped mid-word — and it works for a single row. It does not survive a
-      wrap: a Running Track cell is 78px wide and the caption is not, and `boardW`/`boardH` are
-      computed from CELL extents plus `PAD = 22`, never from the text. **The viewBox does not know the
-      captions exist**, so any caption wider than its cell plus the padding is outside it. This is the
-      same class of bug as the one the comment above the code says was already fixed once; the fix
-      solved the one-row case.
-
-      **The buffer stops.** Both are drawn on the HORIZONTAL ends of their cell — `first.x - 9` and
-      `last.x + last.w + 9` — regardless of which side of the table the cell was laid on. With one row
-      that is right: the line ends to the left and to the right. Wrapped, the East Division Point can
-      end up on the bottom row or down a side, and its stop still points right, into the board rather
-      than away from the route. Jesse's read is that it belongs at the top there. The correct rule is
-      probably "away from the neighbour it joins", which is derivable rather than a special case — and
-      it depends on the vertical-card question above, so do that one first.
-
-- [ ] **INVESTIGATE: fill the dead centre of the Division map with the common board.** Raised by
-      Jesse 2026-08-22: with three or four seats the map is a ring with a large empty middle, while
-      the timetable sits in a column on the far right.
-
-      **The proposal, in his words:** move the timetable into the middle; maybe the yards, "because
-      that is too applicable to everybody"; maybe the Department and Salvage decks as well, since
-      those apply to all players. "Then the Division map becomes the common board area. The
-      right-hand side — your move, your cards, your facilities, what's blocked — is your side of
-      things."
-
-      **That is a genuinely good division of the screen** and it names a principle the layout does not
-      currently have: SHARED in the middle, YOURS on the right. Worth writing down as the rule even if
-      the move itself is deferred, because it decides where anything new belongs.
-
-      **What makes it awkward.** The middle is only empty at three and four seats — at one seat there
-      is no middle at all, and at two the rows face each other across a gap the width of `SIDE_GAP`.
-      So whatever goes there needs somewhere else to live at low seat counts, which is close to
-      building both layouts. And the map is an SVG built by `divisionSvg` while the timetable, yards
-      and decks are HTML panels (`panels.ts`), so "put them in the middle" means either
-      foreign-objecting HTML into the SVG or positioning HTML over it — neither free, and the map is
-      embedded into the replay by `toString()`, which constrains what it may reach for.
-
-- [ ] **INVESTIGATE: seat the viewer at the bottom of the Division map and wrap the table around
-      them.** Raised by Jesse 2026-08-22: "consider that the player being displayed is always at the
-      bottom of the display, and the rest of the table is wrapped around. This would give something
-      of a feel of sitting at an actual table with my cards in front of me and the other players in
-      front of me as well." His own note: not a decision, something to think about.
-
-      **It needs no new data.** `divisionSvg` already takes `roster.viewer` — "the player this map is
-      being drawn for" — and `main.ts` already passes it. The layout is a fixed lane order,
-      `top → right → bottom → left`, filled in ROUTE order west to east; rotating means offsetting
-      which lane the first seat lands in so the viewer's own district comes out on `bottom`. That is
-      an index shift in one array, not a new layout engine.
-
-      **The strongest argument for it is already in the code.** The map marks your district with a
-      colour AND spells out "(you)", and the comment says why: "a colour alone cannot say which of
-      four railroads is the reader's, and that is the first thing anybody wants to know at a table
-      they just sat down at." A fixed position answers that structurally — you would know before
-      reading anything. The marker stays as reinforcement rather than being the only signal.
-
-      **What it costs is the other thing the map says.** The section is headed "The Division — west
-      to east" and the route is a LINE, not a loop: it starts at the West Division Point and ends at
-      the East one, with buffer stops at both and a deliberately open gap between them. Today that
-      line starts top-left, where a reader starts reading. Rotate it and west starts wherever your
-      seat put it — and the open gap, which is the thing that stops the ring being read as a loop,
-      moves with it. Sometimes it would land behind you, out of the eye's path, which is exactly
-      where the one feature that says "this is not a circle" should not be.
-
-      **So the question is which of the two the map is FOR**, and it may not have the same answer at
-      every seat count. At one seat there is nothing to rotate. At two it is a swap, and free. At
-      three, rotating changes which way the horseshoe opens. At four it moves the break in the
-      square. A rule like "rotate at three and four, leave one and two alone" is defensible but has
-      to be decided rather than fallen into.
-
-      **One caller has no viewer at all**: the site's replay viewer calls `divisionSvg(f.division)`
-      with no roster (`web/replays.ts`). A replay watches every seat and belongs to none, so it has
-      nothing to rotate around — the feature has to degrade cleanly to today's layout there, which is
-      an argument for building it as an optional rotation rather than as the layout.
-
-      **Same drawing pass as items 19, 20 and 22** — vertical track art, the corner connectors and
-      the dead centre. All four move cells around the ring or change what is drawn inside them, and
-      doing them one at a time means laying the map out four times.
 
 - [ ] **INVESTIGATE: history newest-at-the-top, and timestamps on it.** Raised by Jesse 2026-08-22.
 

package.json

@@ -1,6 +1,6 @@
 {
   "name": "station-master",
-  "version": "0.7.4",
+  "version": "0.7.9",
   "private": true,
   "type": "module",
   "description": "Station Master — a railroad operations game",

scripts/build-web.ts

@@ -60,7 +60,21 @@ execFileSync(
  */
 function buildStamp(): string {
   const pkg = JSON.parse(readFileSync(join(root, 'package.json'), 'utf8')) as { version: string };
-  let git = 'nogit';
+  /**
+   * THE FALLBACK HAS TO BE UNIQUE PER BUILD, because this string is also the cache-bust key.
+   *
+   * It used to be the literal `nogit`, which is exactly what the `.s9pk` build produces — the
+   * Dockerfile copies the working tree in without `.git`, so `git rev-parse` fails there every time.
+   * Every packaged release therefore published `?v=nogit`, byte-identical to the release before it,
+   * and a returning player's browser had no reason to refetch a single module. v0.7.5's setup screen
+   * and v0.7.6's fix to it both shipped correctly to `phoenix.local` and neither reached the browser
+   * that asked for them (Jesse, twice, 2026-08-29 — "setup did not work").
+   *
+   * The version plus the build's own timestamp is always distinct, needs nothing from the
+   * environment, and stays honest: two builds of the same commit ARE two deploys, and a cache key
+   * that says so costs one refetch, while one that lies costs a release nobody receives.
+   */
+  let git = `${pkg.version}-${Date.now().toString(36)}`;
   try {
     const sha = execFileSync('git', ['rev-parse', '--short', 'HEAD'], { cwd: root })
       .toString()

src/engine/advance.ts

@@ -1636,11 +1636,25 @@ function shiftChange(s: GameState, events: GameEvent[]): AdvanceResult {
     events.push({ type: 'stageBegan', day: s.clock.day, stage: s.clock.stage });
   }
 
-  // §3.4 — every mode but competitive-and-coop-only: a Day's collisions against `maxCollisionsPerDay`
-  // and the game's running total against `maxCollisionsTotal`. `0` disables either check. Flat, not
-  // scaled by player count — Jesse's call, 2026-08-20: more players is more independent chances to
-  // collide, not a bigger shared budget.
-  if (s.config.mode === 'competitive' || s.config.mode === 'coop') {
+  /**
+   * §3.4 — EVERY MODE, SOLITAIRE INCLUDED: a Day's collisions against `maxCollisionsPerDay` and the
+   * game's running total against `maxCollisionsTotal`. `0` disables either check. Flat, not scaled
+   * by player count — Jesse's call, 2026-08-20: more players is more independent chances to collide,
+   * not a bigger shared budget.
+   *
+   * SOLITAIRE WAS EXCLUDED UNTIL 2026-08-30 and nothing said so. The gate here read `mode ===
+   * 'competitive' || mode === 'coop'`, while `SOLO_CONFIG` carried both limits and the New Game
+   * dialog offered them as live settings — so a solitaire player could set a collision limit, read
+   * "the game ends in a loss" beside it, and crash as often as they liked. Found reviewing that
+   * screen's wording (Jesse, 2026-08-30); his ruling is that the settings should do what they say,
+   * so the gate is gone rather than the controls.
+   *
+   * A SOLITAIRE GAME CAN THEREFORE NOW END EARLY, which no measurement in `TODO.md` was taken
+   * under. At the shipped defaults (3 a Day, 5 total) it is a rare ending rather than a common one —
+   * the bot averages 0.06 collisions a game — but any figure quoted from a full-length run predates
+   * it.
+   */
+  {
     const perDayBreach =
       s.config.maxCollisionsPerDay > 0 && s.collisionsToday >= s.config.maxCollisionsPerDay;
     const totalBreach =

src/server/http.ts

@@ -101,7 +101,13 @@ function sendJson(res: ServerResponse, status: number, body: unknown): void {
   res.end(text);
 }
 
-async function serveStatic(distDir: string, urlPath: string, res: ServerResponse): Promise<void> {
+async function serveStatic(
+  distDir: string,
+  urlPath: string,
+  res: ServerResponse,
+  /** The request's `?v=` build tag, when it has one — see the `Cache-Control` note below. */
+  buildTagged = false,
+): Promise<void> {
   const rel = urlPath === '/' ? '/index.html' : urlPath;
   // `normalize` collapses `..`, and the join is then checked to still be inside `distDir` — a request
   // for `/../../etc/passwd` must not escape the one directory this is allowed to read from.
@@ -113,7 +119,27 @@ async function serveStatic(distDir: string, urlPath: string, res: ServerResponse
   try {
     const info = await stat(full);
     if (!info.isFile()) throw new Error('not a file');
-    res.writeHead(200, { 'Content-Type': MIME[extname(full)] ?? 'application/octet-stream', 'Content-Length': info.size });
+    /**
+     * ONLY A URL CARRYING A BUILD TAG MAY BE CACHED, AND NOTHING ELSE MAY BE.
+     *
+     * Nothing here sent a `Cache-Control` at all before, so a browser applied its own heuristic to
+     * the pages as much as the modules. The pages are the one thing that CANNOT be versioned in
+     * their own URL — a player types the address or follows a bookmark — so a cached `play.html`
+     * pins that player to the entire build it names, including every `?v=` tag inside it. That is
+     * half of why v0.7.5 and v0.7.6 did not reach the browser that asked for them; `build-web.ts`
+     * publishing `?v=nogit` on every packaged release was the other half, and neither is enough on
+     * its own.
+     *
+     * `?v=` is the exact condition rather than "not HTML": `build-web.ts` tags the modules and the
+     * script tags that load them, and tags NOTHING else. An untagged URL — an image, the replay
+     * manifest — has no way to announce a change, so a year of `immutable` on one would outlive
+     * several releases of whatever it holds.
+     */
+    res.writeHead(200, {
+      'Content-Type': MIME[extname(full)] ?? 'application/octet-stream',
+      'Content-Length': info.size,
+      'Cache-Control': buildTagged ? 'public, max-age=31536000, immutable' : 'no-cache',
+    });
     createReadStream(full).pipe(res);
   } catch {
     res.writeHead(404, { 'Content-Type': 'text/plain' });
@@ -281,7 +307,7 @@ export function startServer(opts: ServerOptions): void {
         // Unset means the routes are not here — indistinguishable from any other unknown path, so
         // nothing advertises an administrative surface to someone probing for one.
         if (!opts.adminSecret) {
-          await serveStatic(opts.distDir, url.pathname, res);
+          await serveStatic(opts.distDir, url.pathname, res, url.searchParams.has('v'));
           return;
         }
         if (req.headers['x-admin-secret'] !== opts.adminSecret) {
@@ -700,7 +726,7 @@ export function startServer(opts: ServerOptions): void {
         return;
       }
 
-      await serveStatic(opts.distDir, url.pathname, res);
+      await serveStatic(opts.distDir, url.pathname, res, url.searchParams.has('v'));
     })().catch((err: unknown) => {
       sendJson(res, 500, { error: err instanceof Error ? err.message : 'internal error' });
     });

src/sim/board-svg.ts

@@ -55,11 +55,12 @@ export function divisionSvg(nodes: DivisionView[], roster?: DivisionRoster | nul
    *
    *   West DP · Mainline · [Limits … Office … Limits] · Mainline · [ … ] · Mainline · East DP
    *
-   * SEATING. Players sit around a table, so the route is laid out the way they do: one row alone,
-   * two rows facing, a horseshoe of three, a square of four. The Division is a LINE and not a loop
-   * — trains enter at one Division Point and leave at the other — so the shape is deliberately left
-   * open, with the two ends drawn as buffer stops facing each other across a marked gap. Closing it
-   * into a ring would promise a connection the rules do not have.
+   * ONE ROW AT EVERY SEAT COUNT (Gitea#18). It used to be laid out the way players sit — two rows
+   * facing, a horseshoe of three, a square of four — and the reasoning for dropping that is at the
+   * layout itself below. The Division is a LINE and not a loop — trains enter at one Division Point
+   * and leave at the other — so the row is deliberately left open, with the two ends drawn as
+   * buffer stops facing outward. Closing it into a ring would promise a connection the rules do
+   * not have.
    *
    * Self-contained on purpose: the replay embeds this by `toString()`, so it may not reach for
    * anything outside its own body.
@@ -108,7 +109,6 @@ export function divisionSvg(nodes: DivisionView[], roster?: DivisionRoster | nul
    * this is back to what a buffer stop actually needs.
    */
   const PAD = 22;
-  const SIDE_GAP = 34;
 
   const esc = (t: string): string =>
     String(t).replace(/[&<>"]/g, (c) => ({ '&': '&amp;', '<': '&lt;', '>': '&gt;', '"': '&quot;' })[c] ?? c);
@@ -146,6 +146,12 @@ export function divisionSvg(nodes: DivisionView[], roster?: DivisionRoster | nul
     below?: Cell['trains'];
     /** Mainline cards only: §2.1 divides one into two regions. 0 elsewhere — no bars are drawn. */
     regions: number;
+    /**
+     * Which way a Heavy Grade climbs, or null on every other card. Drawn as a wedge, because the
+     * tooltip said "climbs east" and the card itself showed nothing — so the one card whose
+     * orientation the PLAYER chooses was the one card you had to hover to read (Jesse, 2026-08-30).
+     */
+    gradeUp?: string | null;
     w: number;
     x: number;
     y: number;
@@ -259,6 +265,7 @@ export function divisionSvg(nodes: DivisionView[], roster?: DivisionRoster | nul
       seat: null,
       // A Division Point is one region — the queue trains enter and leave the Division through.
       regions: dp ? 1 : (n.regions ?? 0),
+      gradeUp: dp ? null : (n.gradeUp ?? null),
       w: dp ? CW.dp : CW.ml,
     });
   }
@@ -357,6 +364,73 @@ export function divisionSvg(nodes: DivisionView[], roster?: DivisionRoster | nul
       out += `<line class="bs-region" x1="${c.x + 6 + RW * r}" y1="${c.y + RAIL_Y - 14}" x2="${c.x + 6 + RW * r}" y2="${c.y + RAIL_Y + 10}"/>`;
     }
 
+    /**
+     * WHICH WAY A HEAVY GRADE CLIMBS, drawn rather than only said.
+     *
+     * The tooltip has said "climbs east" since the Frame carried `gradeUp`, and the card showed
+     * nothing — so the one Mainline card whose orientation the PLAYER sets, and the one where a
+     * Helpers or Brakeman modifier means opposite things at opposite ends, was the one you had to
+     * hover to read (Jesse, 2026-08-30).
+     *
+     * A wedge rising toward the climb, with an arrow up its slope. Two cues rather than one: the
+     * wedge alone asks the reader to judge which end is taller, which at eleven pixels of rise is a
+     * comparison rather than a glance. Bottom-right, clear of the left-aligned capacity line and
+     * below the region bars, so it never lands under a train chip.
+     *
+     * `east` is RIGHT on this map and always has been (Gitea#18) — that is what makes a wedge
+     * readable without a compass, and it is why the row layout is worth its width.
+     */
+    if (c.gradeUp === 'east' || c.gradeUp === 'west') {
+      const s = c.gradeUp === 'east' ? 1 : -1;
+      const GW = 58;
+      const RISE = 24;
+      const gx1 = c.x + c.w - 9 - GW;
+      const gx2 = c.x + c.w - 9;
+      const yb = c.y + CH - 6;
+      const peakX = s === 1 ? gx2 : gx1;
+      out += `<polygon class="bs-grade" points="${gx1},${yb} ${gx2},${yb} ${peakX},${yb - RISE}"/>`;
+
+      /**
+       * THE ARROW LIES ALONG THE WEDGE'S OWN SLOPE, CENTRED IN IT (Jesse, 2026-08-30).
+       *
+       * Parallel to the hypotenuse is the shape that fits: the perpendicular gap to the slope is
+       * then CONSTANT along the whole arrow, instead of closing at one end the way a steeper line
+       * does. The earlier 30° pass had to be tucked into the fat half to survive, because 30° is
+       * steeper than this wedge climbs — at 58×24 the slope is 22.5°, and the arrow simply lies on
+       * it.
+       *
+       * Centred on the TRIANGLE'S CENTROID (2/3 along the base, 1/3 up), which is the balance point
+       * of the form rather than of its bounding box — centring on the box would push the arrow into
+       * the thin corner where there is no height for it.
+       *
+       * The wedge grew 50×22 → 58×24 to pay for that: the centroid sits only ~7px from the
+       * hypotenuse, so a centred arrow has less room than an off-centre one and needs a bigger form
+       * to keep it. Sizes are the best fit found by search, clearing every edge by 2.88px;
+       * `web.test.ts` re-derives it and fails under 2px.
+       *
+       * Worked in the wedge's own frame — `u` along the base from the thin corner, `h` up from it —
+       * so a westward climb is one sign on `u` rather than a second set of coordinates.
+       */
+      const L = 20;
+      const HL = 6;
+      const HW = 3.5;
+      const T = 1.4;
+      const A = Math.atan2(RISE, GW);
+      const cos = Math.cos(A);
+      const sin = Math.sin(A);
+      const cu = (2 * GW) / 3;
+      const ch = RISE / 3;
+      const pt = (lx: number, ly: number): string => {
+        const u = cu + lx * cos - ly * sin;
+        const h = ch + lx * sin + ly * cos;
+        return `${s === 1 ? gx1 + u : gx2 - u},${yb - h}`;
+      };
+      const H = L / 2;
+      out +=
+        `<polygon class="bs-gradeup" points="${pt(-H, -T)} ${pt(H - HL, -T)} ${pt(H - HL, -HW)} ` +
+        `${pt(H, 0)} ${pt(H - HL, HW)} ${pt(H - HL, T)} ${pt(-H, T)}"/>`;
+    }
+
     /**
      * ON THE DIVISION MAP, A TRAIN IS A CHIP — name, which way it points, how many cars.
      *
@@ -1123,6 +1197,14 @@ export const BOARD_CSS = `
 /* The vertical bars a Mainline card is divided into (§2.1). Drawn faint: they are the ruler the
    train is measured against, not something to look at instead of the train. */
 .bs-region{stroke:#4a5361;stroke-width:1.2;stroke-dasharray:3 3}
+/* THE HEAVY GRADE WEDGE. Terrain, so it is coloured as terrain rather than as a warning.
+   SOLID BROWN, fill and border the same (Jesse, 2026-08-30) — the first pass paired a desaturated
+   fill with an amber arrow and the pair read reddish, which on a map that spends amber on "it is
+   happening here" made a fixed piece of landscape look like a live alert. One flat brown recedes
+   into scenery; the arrow is bone so the DIRECTION, which is the fact being reported, is the part
+   that carries. */
+.bs-grade{fill:#6b5334;stroke:#6b5334;stroke-width:1}
+.bs-gradeup{fill:#f2e8d5}
 .bs-slot{fill:none;stroke:#5f6b7a;stroke-width:1.1;stroke-dasharray:3 2}
 .bs-slot.bs-occ{stroke-dasharray:none;stroke-width:1.6}
 /* CAR TYPE BY COLOUR, LOAD STATE BY FILL — the SAME distinction the train tray draws, because they

src/sim/replay.ts

@@ -35,6 +35,7 @@ import { legalActions } from '../engine/legal.ts';
 import { createGame } from '../engine/setup.ts';
 import type { Facility, GameConfig, GameState } from '../engine/state.ts';
 import { actingPlayer } from '../engine/state.ts';
+import { reasonSentence } from '../web/panels.ts';
 import { developerBot, lastChoiceReason } from './bot.ts';
 import { carLabel, cuesFor, idleNote, isVisible, narrate } from './narrate.ts';
 // The view-model lives in its own module so the browser build can import it without dragging in
@@ -148,12 +149,26 @@ export function record(seed: number, length: GameLength, maxSteps = 100_000): Re
     push(applied.events);
   }
 
+  /**
+   * IN WORDS, NOT AS AN ENUM (`TODO.md` #34). This heading read `loss — revenueFloor`, which is
+   * exactly the defect Gitea#16 was filed about on the playable page — it just outlived the fix
+   * here, because nothing player-facing pointed at it. `reasonSentence` is shared rather than
+   * reimplemented, so the replay and the results screen cannot end up explaining the same ending
+   * two different ways.
+   *
+   * Fed the LAST frame, which is the state the outcome was decided in and is already recorded.
+   * Tags are stripped: this lands in an `<h1>` and in a console line, neither of which wants markup.
+   */
   const o = s.outcome;
+  const last = frames[frames.length - 1];
+  const why = o && last ? reasonSentence(last, o, last.day).replace(/<[^>]+>/g, '') : '';
   return {
     seed,
     length,
     frames,
-    outcome: o ? `${o.result} — ${o.reason} · final Revenue ${s.players[0]?.revenue ?? 0}` : 'unfinished',
+    outcome: o
+      ? `${o.result === 'win' ? 'won' : 'lost'} — ${why} Final Revenue ${s.players[0]?.revenue ?? 0}.`
+      : 'unfinished',
   };
 }
 

src/sim/turnchart.ts

@@ -19,6 +19,8 @@ export type TurnChartFrame = {
   phase: string;
   phaseKey: string;
   actor: number | null;
+  /** What the game has stopped to ask, when it has. Null while a phase is simply running. */
+  awaiting?: { asks: string; train: string } | null;
 };
 
 /**
@@ -93,8 +95,21 @@ export function turnChartHtml(f: TurnChartFrame, actorName: string | null, super
     );
   }).join('');
 
-  // An automatic phase is waiting on nobody, and saying so is more use than a blank.
+  /**
+   * WAITING ON WHOM, AND FOR WHAT.
+   *
+   * "nobody — the Division is running itself" is true of an automatic phase and was being printed
+   * over the top of three interruptions that are emphatically waiting on a person: §8.1's clearance
+   * ruling, the Yard Office offer and the Red Flag prompt. The Frame carried the phase's actor,
+   * which is null throughout the Mainline Phase, so a game stopped on a named player's decision
+   * reported that nobody was holding it up (Jesse, 2026-08-30). `Frame.actor` is `actingPlayer` now
+   * and answers who; `awaiting` says what, because "waiting on Bob" with no more than that is a
+   * game that looks stuck to everyone except Bob.
+   */
   const who = actorName ?? 'nobody — the Division is running itself';
+  const asked = f.awaiting
+    ? ` <span class="tc-asks">${esc(f.awaiting.asks)} · ${esc(f.awaiting.train)}</span>`
+    : '';
   const fedora =
     superName === null
       ? ''
@@ -105,9 +120,12 @@ export function turnChartHtml(f: TurnChartFrame, actorName: string | null, super
     `<div class="tc-when"><b>Day ${f.day}</b><span>Stage ${f.stage} of 12</span>` +
     `<span class="dim">${esc(f.clock)}</span></div>` +
     `<div class="tc-now">phase <b>${esc(f.phase)}</b></div>` +
-    `<div class="tc-who">waiting on <b>${esc(who)}</b></div>` +
-    fedora +
-    `<ol class="tc-phases">${chips}</ol>`
+    `<div class="tc-who">waiting on <b>${esc(who)}</b>${asked}</div>` +
+    // THE FEDORA RIDES AT THE END OF THE PHASE ROW (`TODO.md` #29, Jesse). It sat on its own line
+    // between the phases and everything above them, which put a thing that changes every third
+    // Stage in the middle of the things that change every Stage. The row it belongs beside is the
+    // one whose last chip is Supervisor Shift — the phase that passes it.
+    `<div class="tc-row"><ol class="tc-phases">${chips}</ol>${fedora}</div>`
   );
 }
 
@@ -130,12 +148,18 @@ export const TURNCHART_CSS = `
    says "Player Solitaire", so the chart should agree. In multiplayer this is the thing a table
    glances at most often, so it gets its own chip rather than hiding in the phase text. */
 .tc-who{display:flex;align-items:center;gap:6px;font-size:12px;color:#8b94a3}
+.tc-asks{color:#a99ac4;font-style:italic}
 .tc-who b{color:#b98cf0;background:rgba(150,110,230,.16);border:1px solid #8b6ad0;
   border-radius:11px;padding:1px 9px;font-size:12px}
 /* WHO HOLDS THE FEDORA. Violet like the rest of the chart — this is "where you are" news, not
    something to press — but unfilled, so the eye still lands on "waiting on" first: that is the one
    that changes every turn, while this changes four times a Day. */
-.tc-super{display:flex;align-items:center;gap:6px;font-size:12px;color:#8b94a3;cursor:help}
+/* The phase row and the Fedora on one line, the hat pushed to the far end (TODO.md #29): the row
+   is the Stage, and the Superintendent is who holds it. Wraps under the phases on a narrow screen
+   rather than squeezing the chips. */
+.tc-row{display:flex;align-items:center;gap:12px;flex-wrap:wrap}
+.tc-row ol.tc-phases{flex:1 1 auto}
+.tc-super{margin-left:auto;display:flex;align-items:center;gap:6px;font-size:12px;color:#8b94a3;cursor:help}
 .tc-super b{color:#cbb6f2;border:1px solid #6b5a94;border-radius:11px;padding:1px 9px;font-size:12px}
 ol.tc-phases{display:flex;gap:6px;list-style:none;margin:0;padding:0;flex-wrap:wrap}
 .tc-phase{display:flex;align-items:center;gap:6px;border:1px solid #2c333d;border-radius:14px;

src/sim/view.ts

@@ -47,7 +47,7 @@ import {
 } from '../engine/content.ts';
 import type { Intent } from '../engine/intents.ts';
 import type { Facility, GameConfig, GameState, PlayerIndex, SeatIndex, TrackCard, TurnoutOrientation } from '../engine/state.ts';
-import { carsOn, playerAtSeat, railFacingOf, seatOf, turnOf } from '../engine/state.ts';
+import { actingPlayer, carsOn, playerAtSeat, railFacingOf, seatOf, turnOf } from '../engine/state.ts';
 import type { Hand, HouseRules, TrackGeometry } from '../engine/content.ts';
 import type { Port } from '../engine/track.ts';
 import { connectionsFor, slopeOfPair, variantsFor } from '../engine/track.ts';
@@ -354,7 +354,21 @@ export type Frame = {
    * replay recorder, which sees the events; the live game keeps its own on the Game object.
    */
   cues?: string[];
+  /**
+   * WHO THE GAME IS WAITING ON — the phase's actor, or the owner of a pending interruption when
+   * there is one. It carried `clock.currentActor` alone until 2026-08-30, which is null during the
+   * Mainline Phase, so a game stopped dead on a Superintendent's clearance ruling reported "waiting
+   * on nobody — the Division is running itself" while it waited on a named person to click
+   * (reported by Jesse). The engine had the answer the whole time in `actingPlayer`.
+   */
   actor: number | null;
+  /**
+   * WHAT that player is being asked, when the game is stopped on a question rather than a turn.
+   * Null whenever the phase is simply running. Naming the person is not enough on its own: three
+   * different interruptions can be waiting, and "waiting on Bob" with no more than that is a game
+   * that looks stuck to everyone except Bob.
+   */
+  awaiting: { asks: string; train: string } | null;
   superintendent: number;
   revenue: number;
   /**
@@ -1342,7 +1356,19 @@ export function snapshot(
     clock: clockTime(s.clock.stage),
     phase: phaseLabel(s.clock.phase),
     phaseKey: s.clock.phase,
-    actor: s.clock.currentActor,
+    actor: actingPlayer(s),
+    /**
+     * The three interruptions §8.1 and Gitea#5/#19 can raise, said in the words the prompt itself
+     * uses. `decisionActor` above decides WHO; this is only what they are looking at.
+     */
+    awaiting: (() => {
+      const d = s.clock.pendingDecision;
+      if (!d) return null;
+      const train = trainName(s, d.train);
+      if (d.kind === 'clearance') return { asks: 'a clearance ruling', train };
+      if (d.kind === 'yardOffice') return { asks: 'the Yard Office offer', train };
+      return { asks: 'a Red Flag', train };
+    })(),
     superintendent: s.clock.superintendent,
     revenue: s.players[viewer]?.revenue ?? 0,
     lines,

src/web/game.ts

@@ -134,10 +134,25 @@ export type NewGameOptions = {
 
 /** The same config with the New Game dialog's answers in it. */
 export function configWith(opts: NewGameOptions): GameConfig {
+  const days = opts.days ?? SOLO_CONFIG.days;
   return {
     ...SOLO_CONFIG,
-    days: opts.days ?? SOLO_CONFIG.days,
-    minCombinedRevenue: opts.minCombinedRevenue ?? SOLO_CONFIG.minCombinedRevenue,
+    days,
+    /**
+     * DERIVED FROM THE DAYS ACTUALLY IN PLAY, not from `SOLO_CONFIG`'s five-Day constant.
+     *
+     * It fell back to the constant until 2026-08-30, so `configWith({ days: 1 })` asked a one-Day
+     * game to clear **15** — a floor a five-Day game averages barely half of — and
+     * `configWith({ days: 10 })` asked for the same 15 a five-Day game does. The two fields silently
+     * disagreed, which is the one thing a "build me a config" helper must not let happen.
+     *
+     * Not a live fault when it was found: `createLocalSession` is the only caller, and the page
+     * always writes `minCombinedRevenue` itself (`solitaireDefaults` re-derives it from the preset).
+     * Found by a throwaway probe that passed only `days` — which is exactly how the next caller
+     * would use this. At the default day count the answer is unchanged, since `SOLO_CONFIG`'s own
+     * floor is this same formula at `DEFAULT_DAYS`.
+     */
+    minCombinedRevenue: opts.minCombinedRevenue ?? collectiveRevenueFloor(1, days),
     maxCollisionsPerDay: opts.maxCollisionsPerDay ?? SOLO_CONFIG.maxCollisionsPerDay,
     maxCollisionsTotal: opts.maxCollisionsTotal ?? SOLO_CONFIG.maxCollisionsTotal,
     optionalRules: { ...SOLO_CONFIG.optionalRules, ...(opts.optionalRules ?? {}) },

src/web/index.html

@@ -79,13 +79,13 @@ footer{margin-top:26px;color:var(--dim);font-size:11px;display:flex;gap:18px;fle
       <span class="go" id="door-multiplayer-go">Set up a game &rarr;</span>
     </a>
 
-    <a class="door" href="./play.html">
+    <a class="door" href="./play.html?solitaire">
       <h2>Play solitaire</h2>
-      <p>Play by yourself and run the entire division for five full days. Your goal is 20 Revenue.
-         Your game data is saved in your browser &mdash; if you close the tab and reopen this site
+      <p>Play by yourself and run the entire division for five full days. Clear the Revenue floor of
+         15 by the end or the game is a loss. Your game data is saved in your browser &mdash; if you close the tab and reopen this site
          without clearing your cache, your game is preserved and you can continue automatically.
          During the game you can also explicitly save your progress for later replay.</p>
-      <span class="go">Start a game &rarr;</span>
+      <span class="go">Set up a game &rarr;</span>
     </a>
 
     <a class="door" href="./replays.html">
@@ -100,7 +100,7 @@ footer{margin-top:26px;color:var(--dim);font-size:11px;display:flex;gap:18px;fle
 
   <footer>
     <span>build <span id="build">__BUILD__</span></span>
-    <span>solitaire runs entirely in your browser &mdash; no server code required</span>
+    <span>Multiplayer runs on StartOS server. Solitaire runs entirely in your browser.</span>
   </footer>
 </main>
 

src/web/lobby.ts

@@ -256,25 +256,21 @@ export function runLobby(handlers: LobbyHandlers, resume?: { token: string; game
     else if (type === 'custom') type = base;
     for (const r of typeRadios()) r.checked = r.value === type;
 
-    const scoring = preset(base).scoring;
-    const note = $('lb-type-note');
-    if (type === 'custom') {
-      note.textContent =
-        `${gameTypeLabel('custom', scoring)} · ${differing.length} ` +
-        `${differing.length === 1 ? 'setting differs' : 'settings differ'} from ${preset(base).label}.`;
-      note.className = 'ng-note changed-note';
-      // A Custom game is nobody's default: open the block that says how it differs.
-      $<HTMLDetailsElement>('lb-settings').open = true;
-    } else {
-      note.textContent = preset(type as PresetName).blurb;
-      note.className = 'ng-note';
-    }
+    /**
+     * NO SENTENCE UNDER THE RADIOS since 2026-08-30 — it restated the type just chosen to the person
+     * who had just chosen it, and the row is already labelled and already carries its own
+     * description (Jesse: "There's no need to repeat it below"). `form.mark` still puts a hint on
+     * each row that actually differs, which is where a Custom game's differences can be acted on.
+     */
+    // A Custom game is nobody's default: open the block that says how it differs.
+    if (type === 'custom') $<HTMLDetailsElement>('lb-settings').open = true;
   }
 
   for (const r of typeRadios()) {
     // Solitaire is on this screen so the two screens read as one list, but there is nothing here to
-    // deal it with — the New Game dialog is where a solitaire game comes from.
-    if (r.value === 'solitaire') markUnavailable(r, 'dealt with the New game button, not here');
+    // deal it with. Dimmed and left to speak for itself: the heading says "Game type
+    // (multi-player)", which is the explanation (Jesse, 2026-08-30).
+    if (r.value === 'solitaire') markUnavailable(r);
     r.onchange = () => {
       if (!r.checked) return;
       if (r.value === 'custom') {
@@ -615,18 +611,15 @@ export function prefillCode(code: string): void {
  *
  * Reported by Jesse 2026-08-23: "solitaire is disabled, but really hard to tell." A bare `disabled`
  * on a radio leaves the whole row at full strength — the dot simply refuses the click, which reads
- * as a broken control rather than an unavailable one. Dims the row and says why, once.
+ * as a broken control rather than an unavailable one.
+ *
+ * The dimming is the whole signal now. It used to append a reason to the row as well, and dropped
+ * that in 2026-08-30 along with the same text on the solitaire screen: one heading naming which
+ * game the screen deals says it once, where three dimmed rows each said it again.
  */
-function markUnavailable(radio: HTMLInputElement, why: string): void {
+function markUnavailable(radio: HTMLInputElement): void {
   radio.disabled = true;
-  const row = radio.closest('label');
-  if (!row) return;
-  row.classList.add('disabled');
-  if (row.querySelector('.lb-why')) return;
-  const note = document.createElement('span');
-  note.className = 'lb-why';
-  note.textContent = ` — ${why}`;
-  row.querySelector('span')?.appendChild(note);
+  radio.closest('label')?.classList.add('disabled');
 }
 
 function escapeHtml(s: string): string {

src/web/main.ts

@@ -37,6 +37,7 @@ import {
 } from './presets.ts';
 import type { GameType, PresetName } from './presets.ts';
 import { settingsForm } from './settings-form.ts';
+import type { SettingsForm } from './settings-form.ts';
 
 const SAVE_KEY = 'station-master.save.v1';
 const SETTINGS_KEY = 'station-master.settings.v1';
@@ -505,7 +506,7 @@ const lobbyHandlers = {
     const record = readStore().games[gameId];
     if (!record) return;
     if (record.stage === 'game' && record.seat !== undefined) {
-      beginRemote({ ...record, seat: record.seat });
+      beginRemote({ ...record, seat: record.seat }, true);
       return;
     }
     // Still seated in a lobby that had not started: the stream puts us back on the seating screen,
@@ -532,6 +533,8 @@ const HANDOFF_STALL_MS = 8000;
 let handoffOpenedAt = 0;
 let handoffStall: number | null = null;
 let firstFrameSeen = false;
+/** Set when this page entered a game it was already seated in, so the first frame says so. */
+let rejoiningRemote = false;
 
 function setText(id: string, text: string): void {
   const el = document.getElementById(id);
@@ -611,7 +614,7 @@ function noteDayEnd(f: Frame): void {
  * nothing saying this was the game just set up. `#phasenote` cannot help: it announces a CHANGE of
  * phase, and there is no previous phase to have changed from.
  */
-function noteFirstFrame(f: Frame): void {
+function noteFirstFrame(f: Frame, rejoining = false): void {
   if (firstFrameSeen || isLocal(session)) return;
   firstFrameSeen = true;
   const held = Date.now() - handoffOpenedAt;
@@ -620,20 +623,43 @@ function noteFirstFrame(f: Frame): void {
   window.setTimeout(
     () => {
       closeHandoff();
+      /**
+       * "BEGUN" IS ONLY TRUE ONCE. `firstFrameSeen` is per page-load, so re-entering a game this
+       * browser already holds a seat in — a reload mid-game, or picking it out of the lobby's list
+       * — announced that the game had begun, to a player who had been playing it for an hour.
+       * Reported for the solitaire side by Jesse, 2026-08-30; the same line was wrong here.
+       */
       flashAnnounce(
-        `The game has begun — ${type} · ${f.players.length} players · Day ${f.day}, Stage ${f.stage}`,
+        `The game has ${rejoining ? 'resumed' : 'begun'} — ${type} · ${f.players.length} players · ` +
+          `Day ${f.day}, Stage ${f.stage}`,
       );
     },
     Math.max(0, HANDOFF_BEAT_MS - held),
   );
 }
 
-/** Toggles the two mutually-exclusive top-level screens `play.html` defines — `#lobby` (Phase 4)
- * and `#gameui` (the board, whether local or remote). Both start `hidden` in the markup so neither
- * ever flashes before `start()` decides which one this load actually needs. */
-function showScreen(which: 'lobby' | 'gameui'): void {
+/**
+ * COMING BACK TO A GAME, said out loud — the solitaire counterpart to `noteFirstFrame`.
+ *
+ * A restored game draws exactly like a dealt one: mid-Day, mid-phase, with a log already several
+ * turns deep. Nothing distinguished "this is the game you left" from "this is a game that has just
+ * started", and the multiplayer path had the opposite problem — it announced that the game had
+ * BEGUN to a player rejoining one (Jesse, 2026-08-30: "do not post a message that says 'The game
+ * has begun.' … it needs to say 'The game has resumed.'").
+ */
+function announceResumed(f: Frame): void {
+  flashAnnounce(`The game has resumed — Day ${f.day}, Stage ${f.stage}`);
+}
+
+/** Toggles the three mutually-exclusive top-level screens `play.html` defines — `#lobby` (Phase 4),
+ * `#gameui` (the board, whether local or remote), and `#solitairesetup` (asked before the first
+ * solitaire deal, the same way `#lobby` is asked before the first multiplayer one — Jesse,
+ * 2026-08-29). All three start `hidden` in the markup so none ever flashes before `start()` decides
+ * which one this load actually needs. */
+function showScreen(which: 'lobby' | 'gameui' | 'solitairesetup'): void {
   document.getElementById('lobby')!.hidden = which !== 'lobby';
   document.getElementById('gameui')!.hidden = which !== 'gameui';
+  document.getElementById('solitairesetup')!.hidden = which !== 'solitairesetup';
 }
 
 /**
@@ -642,7 +668,7 @@ function showScreen(which: 'lobby' | 'gameui'): void {
  * earlier visit. Either way the token is what makes reconnection work (`lobby-and-sessions.md` §1),
  * so it is always written back here before anything else happens.
  */
-function beginRemote(ready: LobbyReady): void {
+function beginRemote(ready: LobbyReady, rejoining = false): void {
   saveRemote({ token: ready.token, gameId: ready.gameId, gameCode: ready.gameCode, seat: ready.seat, stage: 'game' });
   gameCode = ready.gameCode;
   showScreen('gameui');
@@ -652,6 +678,7 @@ function beginRemote(ready: LobbyReady): void {
   // banner (`#presence`), and it holds a beat so the game visibly begins.
   openHandoff();
   session = createRemoteSession(ready.token, ready.seat, abandonRemote);
+  rejoiningRemote = rejoining;
   applyCapabilities();
   // A LocalSession has data the instant it is constructed; a RemoteSession does not — its first
   // real Frame only exists once the SSE connection's first push arrives, so the first render waits
@@ -714,12 +741,31 @@ function start(): void {
     return;
   }
 
+  /**
+   * ASKING FOR SOLITAIRE BEATS RESUMING A MULTIPLAYER SESSION TOO — same reasoning as `?lobby`
+   * above, for the door on the other side. A browser that has ever held a multiplayer seat carries
+   * `remembered` forever (`loadRemote` finds it below), and a bare `./play.html` load could not tell
+   * "I clicked Play solitaire" apart from "I reloaded mid-game" — so the splash's solitaire door
+   * always lost to whatever multiplayer game or lobby this browser last touched, and could never
+   * actually reach solitaire. Found 2026-08-29 verifying v0.7.5 on `phoenix.local`: the door landed
+   * back in a Co-op, four-seat LOBBY from unrelated earlier testing rather than solitaire's own new
+   * setup screen. The door now marks its intent explicitly, the same way `?lobby` already does —
+   * and so does everything else that already means "this is a solitaire navigation": an explicit
+   * `?seed=` (a shared or bookmarked deal) and `?hand=` (the setup screen's own Deal button writes
+   * it on every commit, so landing back here with it set is that navigation, not a bare reload).
+   * Checked here, ahead of `remembered`, rather than only below with `saved` — otherwise Deal would
+   * work once and then bounce the very next load into whatever multiplayer game this browser last
+   * touched, since its URL carries `hand=` but not `solitaire=`.
+   */
+  const wantsSolitaire =
+    params.get('solitaire') !== null || params.get('seed') !== null || params.has('hand');
+
   // Entered without checking it still exists — deliberately. Verifying up front would mean an
   // await before anything renders on the common path, where the game IS still there; instead the
   // session reports a dead game through `abandonRemote`, which lands in the lobby.
-  const remembered = loadRemote();
+  const remembered = wantsSolitaire ? null : loadRemote();
   if (remembered && remembered.stage === 'game' && remembered.seat !== undefined) {
-    beginRemote({ ...remembered, seat: remembered.seat });
+    beginRemote({ ...remembered, seat: remembered.seat }, true);
     return;
   }
   /**
@@ -739,23 +785,54 @@ function start(): void {
     return;
   }
 
-  showScreen('gameui');
+  // A saved game carries its OWN rules and re-deals itself under them, whatever the URL says — see
+  // `configFor`. Read once, here, so the same answer decides both whether to ask before dealing and
+  // (below) whether to restore.
+  const saved = load();
   const requested = params.get('seed');
+
+  /**
+   * ASK BEFORE THE FIRST DEAL, THE SAME WAY THE LOBBY ASKS BEFORE THE FIRST MULTIPLAYER GAME
+   * (Jesse, 2026-08-29 — "let the user choose their options like the start of a multiplayer game";
+   * "asking first is the only path").
+   *
+   * Three things answer the question and so skip the screen, in this order of precedence:
+   * `hand` (every `commitNewGame` write sets it, so this navigation IS the Deal button landing back
+   * here to deal), `seed` (a specific deal someone chose to share or bookmark), and — only when the
+   * player did not explicitly ask to set one up — an existing save, which is a game to resume.
+   *
+   * THE DOOR OUTRANKS A SAVED GAME, and getting that wrong is what made this feature unreachable
+   * for three releases. v0.7.5 skipped the screen whenever `load()` found ANYTHING, reasoned as "a
+   * saved game is a game to resume" — but a browser that has ever played solitaire always has one,
+   * so the door could never reach the screen again. Reported three times (Jesse, 2026-08-29 twice
+   * and 2026-08-30); a private window appeared to absolve it only because it had never played and
+   * so had no save. Clicking "Play solitaire" is a request to set a game up, not to resume one — a
+   * BARE reload is the resume case, and still is. `#ss-resume` is what keeps the save reachable, so
+   * this costs nobody the game they were playing.
+   */
+  const askedToSetUp = params.get('solitaire') !== null;
+  if (requested === null && !params.has('hand') && (askedToSetUp || !saved)) {
+    showScreen('solitairesetup');
+    runSolitaireSetup(params, saved !== null);
+    return;
+  }
+
+  showScreen('gameui');
   // A seed in the URL makes a game shareable and reproducible: same link, same deal.
   const seed = requested !== null ? Number(requested) || 1 : Math.floor(Math.random() * 1e9);
   const local = createLocalSession(seed, solitaireDefaults(gameOptionsFromUrl(params)));
   session = local;
-
-  // A saved game carries its OWN rules and re-deals itself under them, whatever the URL says — see
-  // `configFor`. That is why the restore happens after the session is built rather than feeding it.
-  const saved = load();
-  if (saved && requested === null) local.restore(saved);
+  const restored = Boolean(saved) && requested === null;
+  if (saved && restored) local.restore(saved);
 
   applyCapabilities();
   // Every render goes through the session, so the page redraws whenever the game says it changed —
   // which is what a remote session will use to push. Locally it fires on each accepted intent.
   session.subscribe(render);
   render();
+  // Coming back to a game is not the same event as being dealt one, and the board looks identical
+  // either way — mid-Day, mid-phase, with a log already deep (Jesse, 2026-08-30).
+  if (restored) announceResumed(session.view());
 }
 
 /**
@@ -884,7 +961,7 @@ function rulesSummary(f: Frame): string {
 function render(): void {
   const f = session.view();
   const menu = session.menu();
-  noteFirstFrame(f);
+  noteFirstFrame(f, rejoiningRemote);
 
   // Which squares the selected card or track piece may go on. Highlighting them is what turns the
   // coordinate list into a board: you pick the thing, then click where it goes.
@@ -1504,6 +1581,33 @@ function showResults(f: Frame): void {
   const body = document.getElementById('resultsbody');
   if (!dlg || !body) return;
   body.innerHTML = resultsHtml(f);
+
+  /**
+   * ASK THE EXTENSION QUESTION ON THE THING THAT IS ACTUALLY IN FRONT OF THE PLAYER.
+   *
+   * This dialog opens itself at every ending and it is MODAL, so `renderEnding`'s own "play one more
+   * Day" buttons — written into `#actions` — are behind it. The player read a results screen offering
+   * nothing but Close and concluded the game was over, which is exactly what it looked like (Jesse,
+   * 2026-08-30). Gitea#11 was verified over the HTTP API, where there is no dialog to be behind.
+   *
+   * The buttons in `#actions` stay, and are still correct: they are what remains after this is
+   * closed, and what a player who reopened the results with "see the full results" comes back to.
+   * Voting from either place submits the same intent.
+   */
+  const yes = document.getElementById('rs-extend-yes') as HTMLButtonElement | null;
+  const no = document.getElementById('rs-extend-no') as HTMLButtonElement | null;
+  const asking = f.status === 'awaitingExtension' && f.extensionVotes[f.viewer] === null;
+  if (yes && no) {
+    yes.hidden = !asking;
+    no.hidden = !asking;
+    if (asking) {
+      // `method="dialog"` closes it on click; the vote rides along. Assigned every time rather than
+      // once, because `f` is a fresh Frame on each ending.
+      yes.onclick = () => void session.submit({ type: 'game.extend', player: f.viewer, agree: true });
+      no.onclick = () => void session.submit({ type: 'game.extend', player: f.viewer, agree: false });
+    }
+  }
+
   // A redraw can arrive while it is open — `showModal` throws on an already-open dialog rather
   // than doing nothing (the same trap `noteDayEnd` documents).
   if (!dlg.open) dlg.showModal();
@@ -1895,189 +1999,265 @@ if (leaveBtn) {
   };
 }
 
-const newBtn = document.getElementById('newgame');
-const dlg = document.getElementById('newgamedlg') as HTMLDialogElement | null;
-if (newBtn && dlg) {
-  const field = <T extends HTMLElement>(id: string): T => document.getElementById(id) as T;
-  const ngForm = settingsForm('ng-');
+/**
+ * ONE GAME-TYPE BLOCK, WIRED — the type radios, the shared rules form beneath them, and the small
+ * glue between them (which type is currently selected, what its note says, how Days feeds the
+ * floor). The in-game "New game" dialog (`ng-`) and the pre-game setup screen (`ss-`, Gitea
+ * "let the user choose their options like the start of a multiplayer game", 2026-08-29) both need
+ * an identical copy of this — factored out once so the two cannot drift apart the way the rules
+ * block itself already had before `settings-form.ts` existed to stop it.
+ *
+ * PREFILLING IS DELIBERATELY LEFT TO THE CALLER. The dialog opens on the game CURRENTLY IN PLAY
+ * (so redealing to compare keeps comparing); the setup screen opens on the plain Solitaire
+ * defaults, because there is no game yet to read. `setBase` plus a direct `form.write(...)` is the
+ * seam that lets each caller do its own version of "what do these fields show at first paint"
+ * without this function having to guess which one it is wiring.
+ */
+type WiredGameType = {
+  form: SettingsForm;
+  days(): number;
+  refresh(): void;
+  /** The common case: prefill straight from a named type's own defaults, then repaint. */
+  selectPreset(name: PresetName): void;
+  /** The dialog's case: the caller writes the form itself (from a live game), then calls `refresh`
+   * — this only sets which type that write should be compared against. */
+  setBase(name: PresetName, type: GameType): void;
+};
 
-  /**
-   * THE SAME FIVE GAME TYPES THE LOBBY OFFERS, and the same shared rules block under them.
-   *
-   * The dialog used to carry its own copy of the questions and its own idea of the defaults, which
-   * is how it ended up with "where an Extra may start" that the lobby did not have and none of the
-   * three optional rules that it did. Both screens now read `presets.ts` and drive their block
-   * through `settings-form.ts`; only Solitaire can actually be DEALT here, so the three multiplayer
-   * types are shown disabled rather than hidden — what this screen offers and what the lobby offers
-   * should read as one list, not two.
-   */
-  let ngBase: PresetName = 'solitaire';
-  let ngType: GameType = 'solitaire';
+function wireGameTypeBlock(prefix: string, root: ParentNode): WiredGameType {
+  const field = <T extends HTMLElement>(id: string): T => document.getElementById(`${prefix}${id}`) as T;
+  const form = settingsForm(prefix);
+
+  let base: PresetName = 'solitaire';
+  let type: GameType = 'solitaire';
   /** As in the lobby: the floor is derived from the length until the player sets one themselves. */
-  let ngFloorTyped = false;
+  let floorTyped = false;
 
-  const ngDays = (): number => {
-    const raw = Number(field<HTMLInputElement>('ng-days').value);
+  const days = (): number => {
+    const raw = Number(field<HTMLInputElement>('days').value);
     return Number.isFinite(raw) && raw >= 1 ? Math.round(raw) : 5;
   };
 
-  const ngTypeRadios = (): HTMLInputElement[] =>
-    Array.from(dlg.querySelectorAll<HTMLInputElement>('input[name="ng-type"]'));
+  const typeRadios = (): HTMLInputElement[] =>
+    Array.from(root.querySelectorAll<HTMLInputElement>(`input[name="${prefix}type"]`));
 
-  function ngRefresh(): void {
-    const differing = ngForm.mark(ngBase, 1, ngDays());
-    if (differing.length > 0) ngType = 'custom';
-    else if (ngType === 'custom') ngType = ngBase;
-    for (const r of ngTypeRadios()) r.checked = r.value === ngType;
-    const note = field<HTMLElement>('ng-type-note');
-    note.textContent =
-      ngType === 'custom'
-        ? `${gameTypeLabel('custom', preset(ngBase).scoring)} · ${differing.length} ` +
-          `${differing.length === 1 ? 'setting differs' : 'settings differ'} from ${preset(ngBase).label}.`
-        : preset(ngType as PresetName).blurb;
+  function refresh(): void {
+    const differing = form.mark(base, 1, days());
+    if (differing.length > 0) type = 'custom';
+    else if (type === 'custom') type = base;
+    for (const r of typeRadios()) r.checked = r.value === type;
+    /**
+     * NO SENTENCE UNDER THE RADIOS. It restated the type just chosen — the row is already labelled
+     * and already carries its own one-line description — so it was the choice read back to the
+     * person who had just made it (Jesse, 2026-08-30: "It's obvious from what they selected above
+     * what they're playing. There's no need to repeat it below.").
+     *
+     * The Custom case said something the radios do NOT — how many settings differ, and from which
+     * type — and that is not lost: `form.mark` puts a hint on each row that actually differs, which
+     * is where a reader can act on it rather than a count they would then have to go and find.
+     */
   }
 
-  function ngSelectPreset(name: PresetName): void {
-    ngBase = name;
-    ngType = name;
-    ngFloorTyped = false;
-    const values = presetSettings(name, 1, ngDays());
-    ngForm.write(values, values);
-    ngRefresh();
+  function selectPreset(name: PresetName): void {
+    base = name;
+    type = name;
+    floorTyped = false;
+    const values = presetSettings(name, 1, days());
+    form.write(values, values);
+    refresh();
   }
 
-  for (const r of ngTypeRadios()) {
-    // Nothing here can deal a multiplayer game: a `LocalSession` runs the engine in this browser and
-    // a table needs a server. The lobby is the door, and the row says so rather than just refusing
-    // the click (Jesse, 2026-08-23 — a disabled radio that looks enabled reads as a broken one).
+  function setBase(name: PresetName, t: GameType): void {
+    base = name;
+    type = t;
+    floorTyped = false;
+  }
+
+  for (const r of typeRadios()) {
+    /**
+     * Nothing here can deal a multiplayer game: a `LocalSession` runs the engine in this browser and
+     * a table needs a server. Dimmed rather than hidden, so what this screen offers and what the
+     * lobby offers read as one list (Jesse, 2026-08-23 — a disabled radio that looks enabled reads
+     * as a broken one).
+     *
+     * NO REASON PRINTED BESIDE THEM since 2026-08-30. Each row used to gain "— use the Multiplayer
+     * button; a table needs a server", which is three unreachable types each explaining the same
+     * thing on a screen whose heading already says "Game type (solitaire)". Jesse: "grayed out with
+     * no additional explanation. The explanation above… is sufficient." The lobby dims Solitaire the
+     * same way and says nothing either, which is what lets one list serve both screens.
+     */
     if (r.value !== 'solitaire' && r.value !== 'custom') {
       r.disabled = true;
-      const row = r.closest('label');
-      if (row && !row.querySelector('.lb-why')) {
-        row.classList.add('disabled');
-        const note = document.createElement('span');
-        note.className = 'lb-why';
-        note.textContent = ' — use the Multiplayer button; a table needs a server';
-        row.querySelector('span')?.appendChild(note);
-      }
+      r.closest('label')?.classList.add('disabled');
     }
     r.onchange = () => {
       if (!r.checked) return;
       if (r.value === 'custom') {
-        ngType = 'custom';
-        ngRefresh();
+        type = 'custom';
+        refresh();
         return;
       }
-      ngSelectPreset(r.value as PresetName);
+      selectPreset(r.value as PresetName);
     };
   }
 
-  ngForm.onEdit((key) => {
-    if (key === 'minCombinedRevenue') ngFloorTyped = true;
-    ngType = 'custom';
-    ngRefresh();
+  form.onEdit((key) => {
+    if (key === 'minCombinedRevenue') floorTyped = true;
+    type = 'custom';
+    refresh();
   });
 
   // Days is a parameter, not a rule: it re-derives the floor and never makes a game Custom by itself.
-  field<HTMLInputElement>('ng-days').oninput = () => {
-    if (!ngFloorTyped) {
-      const values = ngForm.read();
-      const want = presetSettings(ngBase, 1, ngDays());
-      ngForm.write({ ...values, minCombinedRevenue: want.minCombinedRevenue }, want);
+  field<HTMLInputElement>('days').oninput = () => {
+    if (!floorTyped) {
+      const values = form.read();
+      const want = presetSettings(base, 1, days());
+      form.write({ ...values, minCombinedRevenue: want.minCombinedRevenue }, want);
     }
-    ngRefresh();
+    refresh();
   };
 
   // "Everyone moves one chair left" has no meaning at a table of one — disabled with the rest of the
-  // block still visible, so the two screens read the same.
-  ngForm.setEmployeeRotationAvailable(false);
+  // block still visible, so every screen that offers it reads the same.
+  form.setEmployeeRotationAvailable(false);
 
-  /**
-   * ASK FOR ALL OF IT, rather than documenting URL parameters in the title bar.
-   *
-   * It asked for the seed alone, through `prompt()`. The opening hand and the three revenue rates
-   * were constants in the source, so trying a variation meant an edit and a rebuild — and balance is
-   * the open question this game has (`TODO.md`). A dialog is what lets a playtest be a playtest.
-   *
-   * The dialog OPENS ON THE RULES IN PLAY rather than on the defaults: dealing a second game to
-   * compare against the first is the common case, and re-entering settings each time is how a
-   * comparison silently stops comparing. Which TYPE that is comes out of the comparison — a game
-   * dealt at the Solitaire defaults reopens on Solitaire, and one that was tuned reopens on Custom
-   * with every changed field marked.
-   */
-  newBtn.onclick = () => {
-    // The button itself is hidden for a session that cannot deal (`applyCapabilities`), but the
-    // dialog's whole answer-reading/URL-navigating flow below assumes a LocalSession throughout, so
-    // the guard is repeated — and `local` is captured as a `const` so the narrowing survives the
-    // closures below it (see `renderUndo`'s identical note on why `session` itself cannot be).
-    if (!isLocal(session)) return;
-    const local = session;
-    const f = local.view();
-    const day = f.day;
-    const started = f.status === 'active' && (day > 1 || f.stage > 1);
-    if (started && !confirm(`Forget this game (seed ${local.seed()}, Day ${day}) and deal a new one?`)) return;
+  return { form, days, refresh, selectPreset, setBase };
+}
 
-    field<HTMLInputElement>('ng-seed').value = '';
-    field<HTMLInputElement>('ng-days').value = String(f.days);
-    ngBase = 'solitaire';
-    ngType = 'solitaire';
-    ngFloorTyped = false;
-    // The rules actually in play, then the comparison decides what to call them.
-    ngForm.write(settingsOf(configFromFrame(f)), presetSettings('solitaire', 1, f.days));
-    ngForm.setEmployeeRotationAvailable(false);
-    ngRefresh();
-    dlg.showModal();
+/**
+ * THE COMMIT — reads a wired block's answers and turns them into a URL, the same path `?seed=`
+ * already took: `start()` reads it back out, so there is exactly one place that turns a URL into a
+ * game, whichever screen produced it.
+ */
+function commitNewGame(wired: WiredGameType, seedFieldValue: string): void {
+  const asked = seedFieldValue.trim();
+  // A seed the browser cannot parse is not a reason to refuse to deal — blank and unparseable both
+  // mean "surprise me", which is what leaving the box alone plainly asks for.
+  const seed = asked === '' || !Number.isFinite(Number(asked)) ? '' : String(Math.trunc(Number(asked)));
+  const settings = wired.form.read();
+  const rules = houseRules({
+    houseRules: {
+      startingHand: settings.startingHand,
+      extraStart: settings.extraStart,
+      discardTimetabled: settings.discardTimetabled,
+      revenue: {
+        passengerPerCoach: settings.passengerPerCoach,
+        freightPerLoad: settings.freightPerLoad,
+        trainPerTransit: settings.trainPerTransit,
+      },
+    },
+  });
+  const victory: NewGameOptions = {
+    days: Math.max(1, wired.days()),
+    minCombinedRevenue: settings.minCombinedRevenue,
+    maxCollisionsPerDay: settings.maxCollisionsPerDay,
+    maxCollisionsTotal: settings.maxCollisionsTotal,
+    optionalRules: {
+      reducedVisibility: settings.reducedVisibility,
+      // Never on at a table of one, whatever the box says — the control is disabled for the same
+      // reason, and this is the half that reaches the engine.
+      employeeRotation: false,
+      emergencyToolbox: settings.emergencyToolbox,
+    },
   };
 
+  clearSave();
+  const next = rulesToUrl(rules, victory, seed);
+  // Assigning the search string the page ALREADY has does nothing at all, which reads as a button
+  // that did not work — and it is the common case: deal a random seed, decide it was a bad deal,
+  // deal another at the same settings. Reload instead, and `start()` rolls a fresh seed.
+  if (next === location.search) location.reload();
+  else location.search = next;
+}
+
+/**
+ * THE IN-GAME "NEW GAME" BUTTON GOES TO THE SETUP SCREEN (Jesse, 2026-08-30 — "it should not go to
+ * a separate screen. We should reuse the Solitaire New Game Screen").
+ *
+ * `#newgamedlg` used to be a third copy of the same questions and the one that drifted: it carried
+ * multiplayer wording on a screen only a solitaire player ever sees. It is deleted; this navigates
+ * to the screen that already asks these questions properly.
+ *
+ * Nothing is lost on the way: `render()` calls `save()` every frame, so the game in progress is
+ * always on disk, and the setup screen offers "Continue saved game" to come back to it.
+ */
+const newBtn = document.getElementById('newgame');
+if (newBtn) {
+  newBtn.onclick = () => {
+    showScreen('solitairesetup');
+    // IN PLACE, not a navigation: the live session stays in memory, so the fields can open on the
+    // rules actually being played and "Continue saved game" is just showing the board
+    // again rather than a reload and a replay.
+    runSolitaireSetup(new URLSearchParams(), true, session.view());
+  };
+}
+
+/**
+ * THE PRE-GAME SETUP SCREEN — asked before the FIRST solitaire deal, the same way `#lobby` is
+ * already asked before the first multiplayer one (Jesse, 2026-08-29: "let the user choose their
+ * options like the start of a multiplayer game"; "asking first is the only path").
+ *
+ * Only reached for a genuinely fresh visit — `start()` is what decides that; by the time this runs,
+ * there is no saved game and no URL already carrying a deal's answers. It opens on the plain
+ * Solitaire defaults, since there is no live game to compare against yet, and reuses the identical
+ * `wireGameTypeBlock`/`commitNewGame` pair the in-game dialog uses — the two are one design, not two.
+ */
+function runSolitaireSetup(params: URLSearchParams, hasSave = false, live: Frame | null = null): void {
+  const screen = document.getElementById('solitairesetup');
+  const dealBtn = document.getElementById('ss-deal');
+  if (!screen || !dealBtn) return;
+
+  const ss = wireGameTypeBlock('ss-', screen);
+  // A `?seed=` with no other rules params still means SOMETHING — a shared or bookmarked link
+  // naming a specific deal — so it is honoured as a prefill rather than discarded because this
+  // visit happened to be routed through the screen that now asks first.
+  const seedField = document.getElementById('ss-seed') as HTMLInputElement | null;
+  if (seedField) seedField.value = params.get('seed') ?? '';
+
   /**
-   * One handler for every way the dialog can close — the Deal button, the Cancel button, and Esc,
-   * which `<dialog>` answers with an empty `returnValue` and no submit event at all.
+   * WHAT THE FIELDS OPEN ON, and it is not the same question in both directions.
    *
-   * The answers go into the URL and the page navigates, which is the same path `?seed=` already
-   * took: `start()` reads them back, so there is exactly one place that turns a URL into a game.
+   * Reached mid-game from "New game", this opens on the rules CURRENTLY IN PLAY — that is what the
+   * deleted dialog was good for, and losing it would make "change one dial and redeal to compare"
+   * impossible. Reached from the splash, there is no game to read, so it opens on the plain
+   * Solitaire defaults.
    */
-  dlg.addEventListener('close', () => {
-    if (dlg.returnValue !== 'deal') return;
+  if (live) {
+    const daysField = document.getElementById('ss-days') as HTMLInputElement | null;
+    if (daysField) daysField.value = String(live.days);
+    ss.setBase('solitaire', 'solitaire');
+    ss.form.write(settingsOf(configFromFrame(live)), presetSettings('solitaire', 1, live.days));
+    ss.refresh();
+  } else {
+    ss.selectPreset('solitaire');
+  }
 
-    const asked = field<HTMLInputElement>('ng-seed').value.trim();
-    // A seed the browser cannot parse is not a reason to refuse to deal — blank and unparseable
-    // both mean "surprise me", which is what leaving the box alone plainly asks for.
-    const seed = asked === '' || !Number.isFinite(Number(asked)) ? '' : String(Math.trunc(Number(asked)));
-    const settings = ngForm.read();
-    const rules = houseRules({
-      houseRules: {
-        startingHand: settings.startingHand,
-        extraStart: settings.extraStart,
-        discardTimetabled: settings.discardTimetabled,
-        revenue: {
-          passengerPerCoach: settings.passengerPerCoach,
-          freightPerLoad: settings.freightPerLoad,
-          trainPerTransit: settings.trainPerTransit,
-        },
-      },
-    });
-    const victory: NewGameOptions = {
-      days: Math.max(1, ngDays()),
-      minCombinedRevenue: settings.minCombinedRevenue,
-      maxCollisionsPerDay: settings.maxCollisionsPerDay,
-      maxCollisionsTotal: settings.maxCollisionsTotal,
-      optionalRules: {
-        reducedVisibility: settings.reducedVisibility,
-        // Never on at a table of one, whatever the box says — the control is disabled for the same
-        // reason, and this is the half that reaches the engine.
-        employeeRotation: false,
-        emergencyToolbox: settings.emergencyToolbox,
-      },
-    };
+  /**
+   * THE WAY BACK TO A GAME IN PROGRESS, and the reason the door is allowed to outrank a save at all.
+   * Dealing from here calls `clearSave()`, so a player who reached this screen from the splash — by
+   * clicking "Play solitaire", which nobody reads as "throw away what I was playing" — needs their
+   * game one button away and needs to be told what Deal costs.
+   *
+   * Mid-game the game is still in memory, so going back is just showing it again. From the splash
+   * there is nothing loaded yet, so it is a navigation to the bare URL and `start()` restores the
+   * save — one place that turns a URL into a game, either way.
+   */
+  const resumeBtn = document.getElementById('ss-resume');
+  const savedNote = document.getElementById('ss-saved-note');
+  const canResume = hasSave || live !== null;
+  if (resumeBtn) {
+    resumeBtn.hidden = !canResume;
+    resumeBtn.onclick = live
+      ? () => {
+          showScreen('gameui');
+          render();
+          announceResumed(session.view());
+        }
+      : () => void (location.search = '');
+  }
+  if (savedNote) savedNote.hidden = !canResume;
 
-    clearSave();
-    const next = rulesToUrl(rules, victory, seed);
-    // Assigning the search string the page ALREADY has does nothing at all, which reads as a button
-    // that did not work — and it is the common case: deal a random seed, decide it was a bad deal,
-    // deal another at the same settings. Reload instead, and `start()` rolls a fresh seed.
-    if (next === location.search) location.reload();
-    else location.search = next;
-  });
+  dealBtn.onclick = () => commitNewGame(ss, seedField?.value ?? '');
 }
 
 const zoomOutBtn = document.getElementById('zoomout') as HTMLButtonElement | null;

src/web/panels.ts

@@ -254,8 +254,13 @@ function collisionsHtml(f: Frame): string {
  * game read the words `GAME OVER — revenueFloor`: an internal enum value, printed at the one moment
  * the game has the player's whole attention. Each reason gets a sentence that says what actually
  * happened, with this game's own numbers in it.
+ *
+ * EXPORTED for the developer replay recorder (`sim/replay.ts`), which was still printing
+ * `loss — revenueFloor` into its own heading a release after this was written — the same defect the
+ * issue was filed about, surviving in the one place nobody had looked (`TODO.md` #34). One
+ * implementation, so the two cannot say the game ended for different reasons.
  */
-function reasonSentence(f: Frame, o: NonNullable<Frame['outcome']>, day: number): string {
+export function reasonSentence(f: Frame, o: NonNullable<Frame['outcome']>, day: number): string {
   const combined = f.players.reduce((n, p) => n + p.revenue, 0);
   switch (o.reason) {
     case 'daysElapsed':

src/web/play.html

@@ -76,11 +76,24 @@ dialog input:focus{outline:none;border-color:#4d6fa8}
   padding:5px 14px;cursor:pointer;font:inherit;font-size:13px}
 .ng-buttons button:hover{border-color:#4d6fa8}
 #ng-deal{background:#31527f;border-color:#4d6fa8}
+/* The save warning is the one thing on this screen that describes something IRREVERSIBLE, and it
+   sat in `.ng-note` — the same dim 11px grey as the twenty explanatory notes above it, which is
+   where the eye has already learned there is nothing to act on. Sized and coloured to be read
+   (Jesse, 2026-08-30). Amber rather than red: losing a saved game is a real cost, not a danger, and
+   red here would outrank the actual rules of the game sitting above it. */
+#ss-saved-note{font-size:15px;font-weight:500;line-height:1.55;color:#ffcf70;background:#332a15;
+  border:1px solid #b8912c;border-left:5px solid #e0a83c;border-radius:5px;padding:12px 14px;
+  margin:18px 0 0}
+#ss-saved-note b{color:#ffe3a6}
+/* Two live choices, so neither is the quiet one: `Create new game` keeps the primary blue it has when
+   it is the only button, and `Continue` is given the same weight rather than reading as a cancel. */
+#ss-deal{background:#31527f;border-color:#4d6fa8}
+#ss-resume{background:#2f5340;border-color:#4f8a68}
 main{display:grid;grid-template-columns:minmax(0,1fr) 400px;gap:14px;padding:14px;align-items:start}
 @media(max-width:1100px){main{grid-template-columns:1fr}}
 section{background:var(--panel);border:1px solid var(--line);border-radius:7px;
   padding:10px 12px;margin-bottom:12px}
-#lobby{max-width:1040px;margin:0 auto;padding:14px}
+#lobby,#solitairesetup{max-width:1040px;margin:0 auto;padding:14px}
 /* The create form is two short lists, not one long one: what game this is on the left, what its
    rules are on the right. Collapses to one column where there is no room for two. */
 .lb-two{display:grid;grid-template-columns:minmax(0,1fr) minmax(0,1.1fr);gap:22px;align-items:start}
@@ -95,9 +108,8 @@ section{background:var(--panel);border:1px solid var(--line);border-radius:7px;
    lobby and nothing on it said so — a radio that silently refuses reads as a broken radio. */
 .ng-radio.disabled{opacity:.45;cursor:not-allowed}
 .ng-radio.disabled:hover{background:none}
-.lb-why{color:#e0b060;font-size:11px}
-#lobby h2{margin-top:0}
-#lobby h3{margin-bottom:2px}
+#lobby h2,#solitairesetup h2{margin-top:0}
+#lobby h3,#solitairesetup h3{margin-bottom:2px}
 .lb-seat{display:flex;align-items:center;gap:8px;padding:5px 0;border-bottom:1px solid var(--line)}
 .lb-seat:last-child{border-bottom:none}
 .lb-seat .who{flex:1}
@@ -307,9 +319,10 @@ ul.blocked li{padding:2px 0}
 
 <!-- THE LOBBY (Phase 4) — shown instead of the game UI whenever there is no game yet to play: no
      stored session token, or a token whose game hasn't started. `lobby.ts` owns everything in here;
-     `main.ts` only decides whether THIS div or `#gameui` below is the one currently visible.
-     `#newgamedlg` at the very end of the body is solitaire-only, and asks the same questions through
-     the same shared module (`settings-form.ts`) — the two blocks are generated from one template. -->
+     `main.ts` only decides whether THIS div, `#solitairesetup` or `#gameui` is the one visible.
+     `#solitairesetup` asks the same questions of a solitaire player, through the same shared module
+     (`settings-form.ts`) — the two blocks are generated from one template, and since 2026-08-30
+     they are the ONLY two: the in-game dialog that was a third copy is gone. -->
 <div id="lobby" hidden>
   <header><b><a href="./index.html" class="home">Station Master</a></b> — <span class="dim">Multiplayer</span></header>
 
@@ -402,12 +415,13 @@ ul.blocked li{padding:2px 0}
          can be shared, compared or replayed. Leave it blank for a random one.</p>
       <!-- WITH THE TABLE SIZE IT IS ABOUT, not below the rules block — reported by Jesse, who found
            it separated from the control it explains by fifteen settings. -->
-      <p class="ng-note">Every chair has to be taken before the game can start — by a person or by a
-         bot. Pick the size of the table now; it cannot change once the game is created.</p>
+      <p class="ng-note">Every chair must be filled before the game can start. For solitaire,
+         there&rsquo;s only one player. For multiplayer, that must be filled by a person or a
+         bot. The number of players cannot be changed once the game is created.</p>
       </div>
 
       <div class="lb-col">
-      <h3>Game type</h3>
+      <h3>Game type (multi-player)</h3>
     <div class="set-row" id="lb-type-row">
       <label class="ng-radio"><input type="radio" name="lb-type" value="solitaire">
         <span><b>Solitaire</b><br><span class="dim">One railroad, one player. The whole Division is yours to run.</span></span></label>
@@ -421,7 +435,7 @@ ul.blocked li{padding:2px 0}
         <span><b>Custom</b><br><span class="dim">Whatever you set below. Selected for you the moment you change a rule; it is scored as the type you started from.</span></span></label>
     </div>
 
-      <p class="ng-note" id="lb-type-note"></p>
+      
       </div>
 
       <!-- FULL WIDTH WHEN IT OPENS. Reported by Jesse: opened inside the right-hand column it made a
@@ -431,7 +445,7 @@ ul.blocked li{padding:2px 0}
         <summary>Game settings</summary>
         <p class="ng-note">Every rule the game type sets, and every one of them yours to change.
            Changing any of them selects <b>Custom</b>, which keeps the scoring of the type you
-           started from; clicking a type again resets all of them back to it. They are fixed when the
+           started from; changing the game type resets all of them back to it. They are fixed when the
            game is created and cannot be changed once it starts.</p>
       <div class="set-groups">
 
@@ -505,13 +519,13 @@ ul.blocked li{padding:2px 0}
       </div>
       <div class="set-row" id="lb-colday-row">
         <label class="ng-gate"><input type="checkbox" id="lb-colday-on" checked>
-          <span>The game ends and everyone loses if collisions in one Day reach</span>
+          <span>The game ends immediately and results in a loss if collisions in one Day reach</span>
           <input id="lb-colday" type="number" min="0" step="1" class="gate-num"></label>
         <span class="set-hint" id="lb-colday-hint"></span>
       </div>
       <div class="set-row" id="lb-coltotal-row">
         <label class="ng-gate"><input type="checkbox" id="lb-coltotal-on" checked>
-          <span>The game ends and everyone loses after this many collisions in the whole game</span>
+          <span>The game ends immediately and results in a loss after this many collisions in the whole game</span>
           <input id="lb-coltotal" type="number" min="0" step="1" class="gate-num"></label>
         <span class="set-hint" id="lb-coltotal-hint"></span>
       </div>
@@ -522,7 +536,7 @@ ul.blocked li{padding:2px 0}
 
       <div class="set-group">
       <h3>Optional rules</h3>
-      <p class="ng-note">Off in every game type; each one changes how the game plays.</p>
+      <p class="ng-note">Each one changes how the game plays.</p>
       <div class="set-row" id="lb-visibility-row">
         <label class="ng-num"><span>Reduced Visibility — five switching Moves instead of six in the
           night Stages (1&ndash;3 and 11&ndash;12)</span>
@@ -555,7 +569,7 @@ ul.blocked li{padding:2px 0}
       </details>
 
       <div class="lb-span">
-        <button id="lb-create" type="button">Create game</button>
+        <button id="lb-create" type="button">Create new game</button>
         <p class="lb-error" id="lb-create-err" role="alert"></p>
       </div>
       </div>
@@ -592,6 +606,208 @@ ul.blocked li{padding:2px 0}
   </section>
 </div>
 
+<!-- ===================================================================
+     SOLITAIRE SETUP — the same question multiplayer already asks first,
+     now asked here too (Jesse, 2026-08-29): a genuinely fresh visit deals
+     nothing until this screen's own Deal button is pressed. A saved game,
+     an explicit `?seed=`, or a URL already carrying a Deal's answers (any
+     of the shared block's fields — `hand` names the one always written)
+     all skip straight past this screen, exactly as `?lobby` already skips
+     past it into the lobby: those are not "no plan yet", they are a
+     choice already made, elsewhere.
+
+     THE SAME BLOCK THE DIALOG AND THE LOBBY USE, same shared module
+     (`settings-form.ts`), same order — three screens are one design now
+     instead of two. Only Solitaire can be dealt from here, so the other
+     four types are shown exactly as the in-game dialog shows them: present,
+     disabled, with a note pointing at the Multiplayer door instead.
+     ==================================================================== -->
+<div id="solitairesetup" hidden>
+  <header><b><a href="./index.html" class="home">Station Master</a></b> — <span class="dim">Solitaire</span></header>
+
+  <section>
+    <h2>New solitaire game</h2>
+    <p class="ng-note">One railroad, one player, five full days by default — everything below is
+       yours to change before you deal. Clearing the Revenue floor wins; falling short loses.</p>
+
+    <!-- THE SAME THREE PARAMETERS THE LOBBY ASKS, in the same order, with the same note under them
+         (Jesse, 2026-08-30: "everything beneath that should be the same"). The table size is here
+         rather than hidden because it is one of the three things that describe a game, and leaving
+         it out made this screen a different form that happened to share a rules block. It is LOCKED
+         at one: a `LocalSession` runs the engine in this browser and a table needs a server, which
+         is the same reason the four multiplayer game types are shown disabled below. -->
+    <div class="lb-params">
+      <label class="ng-num"><span>Seed</span>
+        <input id="ss-seed" type="text" inputmode="numeric" autocomplete="off" placeholder="blank for a random seed"></label>
+      <label class="ng-num"><span>Players at the table</span>
+        <select id="ss-players" disabled>
+          <option value="1" selected>1</option>
+        </select></label>
+      <label class="ng-num"><span>Days</span>
+        <input id="ss-days" type="number" min="1" max="20" step="1" value="5"></label>
+    </div>
+    <p class="ng-note">The same seed and the same settings always deal the same railroad, so a game
+       can be shared, compared or replayed. Leave it blank for a random one.</p>
+    <p class="ng-note">Every chair must be filled before the game can start. For solitaire,
+       there&rsquo;s only one player. For multiplayer, that must be filled by a person or a
+       bot. The number of players cannot be changed once the game is created.</p>
+
+    <h3>Game type (solitaire)</h3>
+    <div class="set-row" id="ss-type-row">
+      <label class="ng-radio"><input type="radio" name="ss-type" value="solitaire" checked>
+        <span><b>Solitaire</b><br><span class="dim">One railroad, one player. The whole Division is yours to run.</span></span></label>
+      <label class="ng-radio"><input type="radio" name="ss-type" value="coop">
+        <span><b>Co-op</b><br><span class="dim">Everyone&#8217;s Revenue is one table score. You win together or lose together.</span></span></label>
+      <label class="ng-radio"><input type="radio" name="ss-type" value="competitive">
+        <span><b>Competitive</b><br><span class="dim">Highest Revenue wins — unless the table misses its combined minimum, and then everyone loses.</span></span></label>
+      <label class="ng-radio"><input type="radio" name="ss-type" value="cutthroat">
+        <span><b>Cutthroat</b><br><span class="dim">Highest Revenue wins, and nothing is shared — the only way everyone loses is three collisions in one Day.</span></span></label>
+      <label class="ng-radio"><input type="radio" name="ss-type" value="custom">
+        <span><b>Custom</b><br><span class="dim">Whatever you set below. Selected for you the moment you change a rule; it is scored as the type you started from.</span></span></label>
+    </div>
+
+    
+
+    <details id="ss-settings" open>
+      <summary>Game settings</summary>
+      <p class="ng-note">Every rule the game type sets, and every one of them yours to change.
+         Changing any of them selects <b>Custom</b>, which keeps the scoring of the type you
+         started from; changing the game type resets all of them back to it. They are fixed when the
+         game is created and cannot be changed once it starts.</p>
+      <div class="set-groups">
+
+      <div class="set-group">
+      <h3>Starting hand</h3>
+      <p class="ng-note">What each player is dealt before the first turn. The hand limit is three
+         either way — deal six and the first turn is spent choosing which of them to keep.</p>
+      <div class="set-row" id="ss-hand-row">
+        <label class="ng-radio"><input type="radio" name="ss-hand" value="threeRandom">
+          <span><b>Three random cards</b><br><span class="dim">The original rule. At the hand limit already, and no guarantee of track.</span></span></label>
+        <label class="ng-radio"><input type="radio" name="ss-hand" value="sixRandom" checked>
+          <span><b>Six random cards</b><br><span class="dim">Twice the choice, still no guaranteed track — the first turn is a discard.</span></span></label>
+        <label class="ng-radio"><input type="radio" name="ss-hand" value="threeTrackThreeOther">
+          <span><b>Three random track and three random non-track cards</b><br><span class="dim">Dealt from two piles, so the district you can build is dealt rather than waited for.</span></span></label>
+        <span class="set-hint" id="ss-hand-hint"></span>
+      </div>
+      </div>
+
+      <div class="set-group">
+      <h3>Where an Extra may start</h3>
+      <p class="ng-note">The player who plays an Extra Train card chooses where its Crew Tray goes,
+         and the place decides which way it runs — a Division Point sends it away from itself; in the
+         middle of the railroad the player picks east or west. The Division Points and the Interchange
+         belong to nobody and are always available. Starting one inside a district is the part that
+         favours a seat, so it is set here. An Office must be a Control Point whatever this says: a
+         Whistle Post never qualifies.</p>
+      <div class="set-row" id="ss-extra-row">
+        <label class="ng-radio"><input type="radio" name="ss-extra" value="divisionPointsOnly">
+          <span><b>Division Points and the Interchange only</b><br><span class="dim">The strictest reading. Every Extra begins on shared ground.</span></span></label>
+        <label class="ng-radio"><input type="radio" name="ss-extra" value="ownOffice">
+          <span><b>Also the playing player&#8217;s own Control Point</b><br><span class="dim">You may start one at home, but not in somebody else&#8217;s district.</span></span></label>
+        <label class="ng-radio"><input type="radio" name="ss-extra" value="anyOffice">
+          <span><b>Also any player&#8217;s Control Point</b><br><span class="dim">The most permissive — an Extra may be planted in another player&#8217;s district.</span></span></label>
+        <span class="set-hint" id="ss-extra-hint"></span>
+      </div>
+      </div>
+
+      <div class="set-group">
+      <h3>Revenue</h3>
+      <p class="ng-note">What each piece of work pays, 0 to 5. A coach pays when it is boarded and
+         again when it is detrained; a load pays when it is made up and again when it is broken. Zero
+         switches an economy off so the others can be read.</p>
+      <div class="set-row" id="ss-passenger-row">
+        <label class="ng-num"><span>Passenger revenue per coach</span>
+          <input id="ss-passenger" type="number" min="0" max="5" step="1" value="1"></label>
+        <span class="set-hint" id="ss-passenger-hint"></span>
+      </div>
+      <div class="set-row" id="ss-freight-row">
+        <label class="ng-num"><span>Freight revenue per load</span>
+          <input id="ss-freight" type="number" min="0" max="5" step="1" value="1"></label>
+        <span class="set-hint" id="ss-freight-hint"></span>
+      </div>
+      <div class="set-row" id="ss-transit-row">
+        <label class="ng-num"><span>Train revenue per transit</span>
+          <input id="ss-transit" type="number" min="0" max="5" step="1" value="0"></label>
+        <span class="set-hint" id="ss-transit-hint"></span>
+      </div>
+      <p class="ng-note">A transit pays every player, once, when a train runs off the end of the
+         Division — the one thing nobody has to work for.</p>
+      </div>
+
+      <div class="set-group">
+      <h3>Victory conditions</h3>
+      <p class="ng-note">The ways this game can end badly. Each one is switched on or off in its own
+         right; how long the game runs is set above, with the table size.</p>
+      <div class="set-row" id="ss-minrev-row">
+        <label class="ng-gate"><input type="checkbox" id="ss-minrev-on" checked>
+          <span>You lose if Revenue at the end is under</span>
+          <input id="ss-minrev" type="number" min="0" step="1" class="gate-num"></label>
+        <span class="set-hint" id="ss-minrev-hint"></span>
+      </div>
+      <div class="set-row" id="ss-colday-row">
+        <label class="ng-gate"><input type="checkbox" id="ss-colday-on" checked>
+          <span>The game ends immediately and results in a loss if collisions in one Day reach</span>
+          <input id="ss-colday" type="number" min="0" step="1" class="gate-num"></label>
+        <span class="set-hint" id="ss-colday-hint"></span>
+      </div>
+      <div class="set-row" id="ss-coltotal-row">
+        <label class="ng-gate"><input type="checkbox" id="ss-coltotal-on" checked>
+          <span>The game ends immediately and results in a loss after this many collisions in the whole game</span>
+          <input id="ss-coltotal" type="number" min="0" step="1" class="gate-num"></label>
+        <span class="set-hint" id="ss-coltotal-hint"></span>
+      </div>
+      <p class="ng-note">The opponent-directed cards — Derail, Watertower, Hobo Jungle and the
+         nineteen others, along with the seven that answer them — are not implemented yet, so no game
+         type deals them whatever else is set here.</p>
+      </div>
+
+      <div class="set-group">
+      <h3>Optional rules</h3>
+      <p class="ng-note">Each one changes how the game plays.</p>
+      <div class="set-row" id="ss-visibility-row">
+        <label class="ng-num"><span>Reduced Visibility — five switching Moves instead of six in the
+          night Stages (1&ndash;3 and 11&ndash;12)</span>
+          <input id="ss-visibility" type="checkbox"></label>
+        <span class="set-hint" id="ss-visibility-hint"></span>
+      </div>
+      <div class="set-row" id="ss-rotation-row">
+        <label class="ng-num"><span>Employee Rotation — not applicable for solitaire</span>
+          <input id="ss-rotation" type="checkbox" disabled></label>
+        <span class="set-hint" id="ss-rotation-hint"></span>
+      </div>
+      <div class="set-row" id="ss-toolbox-row">
+        <label class="ng-num"><span>Emergency Toolbox — start holding a Red Flag, so a hand of four;
+          play or discard down to three on the first turn</span>
+          <input id="ss-toolbox" type="checkbox"></label>
+        <span class="set-hint" id="ss-toolbox-hint"></span>
+      </div>
+      <div class="set-row" id="ss-tossloco-row">
+        <label class="ng-num"><span>A Timetabled train may be discarded — toss it face-up to a
+          Department slot. Turn this off and a train card can only ever be played onto the timetable.
+          An Extra is never discardable either way</span>
+          <input id="ss-tossloco" type="checkbox"></label>
+        <span class="set-hint" id="ss-tossloco-hint"></span>
+      </div>
+      </div>
+
+      </div>
+    </details>
+
+    <!-- Shown only when `station-master.save.v1` holds a game. Dealing from this screen CLEARS that
+         save (`commitNewGame` calls `clearSave`), so without a way back the door would be a way to
+         lose a game in progress — and the door is reached by clicking "Play solitaire", which nobody
+         reads as "discard what I was playing". -->
+    <p id="ss-saved-note" hidden><b>You have a solitaire game in progress.</b> Creating a new game
+       replaces it permanently &mdash; there is no undo. Choose <b>Continue saved game</b> to pick it
+       up where you left off.</p>
+
+    <menu class="ng-buttons">
+      <button id="ss-resume" type="button" hidden>Continue saved game</button>
+      <button id="ss-deal" type="button">Create new game</button>
+    </menu>
+  </section>
+</div>
+
 <div id="gameui" hidden>
 <div class="topbar">
 <header>
@@ -623,7 +839,7 @@ ul.blocked li{padding:2px 0}
   </span>
   <button id="undo" title="Take the last action back. The save is the seed plus the moves made, so this replays the game without the last one — as far back as you like.">Undo</button>
   <button id="savefile" title="Download this game as a save file you can replay or share">Save replay</button>
-  <button id="newgame" title="Deal a fresh game. You choose the seed, the opening hand and what the three economies pay. Undo steps back one action at a time; this throws the whole game away, so download the replay first if you want to keep it.">New game</button>
+  <button id="newgame" title="Set up a fresh game — the seed, the table, the opening hand and what the three economies pay. Opens the same screen a new solitaire game starts from, with your current rules filled in; your game in progress is kept until you press Deal, and Continue puts it straight back.">New game</button>
   <button id="multiplayer" title="Create or join a Competitive or Co-op game on this server, with other players.">Multiplayer</button>
   <!-- LEAVING A RUNNING GAME. Reported by Jesse 2026-08-23: "if I'm a player in the middle of the
        game and I need to leave, how do I leave the game, clear the token from my browser so I can
@@ -704,187 +920,20 @@ ul.blocked li{padding:2px 0}
   </div>
 </main>
 
-<!-- ===================================================================
-     NEW GAME — the seed, the opening hand, and what the three economies pay.
+<!-- THE IN-GAME "NEW GAME" BUTTON GOES TO THE SOLITAIRE SETUP SCREEN — there is no second
+     dialog any more (Jesse, 2026-08-30: "it should not go to a separate screen. We should reuse the
+     Solitaire New Game Screen… in general we should reuse what we already have").
 
-     It was a `prompt()` asking for a seed. Two of the three things that decide what kind of game
-     you are about to play had no way in at all: the opening hand had been changed twice with no
-     way back to the earlier rule, and the revenue rates were constants in the source. Balance is
-     the open question in this game (`TODO.md`), and the way to settle it is to deal several games
-     at different settings — which needs a dialog, not a rebuild.
+     `#newgamedlg` was a third copy of the same questions, and the one that drifted: it kept the
+     multiplayer wording ("Everyone loses if COMBINED Revenue…") on a screen only ever shown to a
+     solitaire player, and explained Employee Rotation in full beside a control it had disabled.
+     Deleting it removes the drift rather than re-wording it.
 
-     Every control has a default that is the recommended answer, so DEAL with nothing touched is a
-     complete, sensible game. The settings ride in the URL alongside the seed, because a seed alone
-     no longer names a game: `?seed=430` with a different opening hand is a different railroad.
-     ==================================================================== -->
+     Nothing is lost by navigating away mid-game: `render()` calls `save()` on every frame, so the
+     game in progress is always on disk, and the setup screen offers "Continue saved game"
+     to come straight back to it. -->
 </div><!-- /gameui -->
 
-<dialog id="newgamedlg" aria-labelledby="ng-title">
-  <form method="dialog" id="newgameform">
-    <h2 class="big" id="ng-title">New game</h2>
-
-    <!-- THE SAME BLOCK THE LOBBY USES, same shared module, same order — the two screens are one
-         design. Only Solitaire can be dealt here; the multiplayer types are shown disabled rather
-         than hidden, so what this screen offers and what the lobby offers read as one list. -->
-    <div class="lb-params">
-      <label class="ng-num"><span>Seed</span>
-        <input id="ng-seed" type="text" inputmode="numeric" autocomplete="off" placeholder="blank for a random seed"></label>
-      <label class="ng-num"><span>Days</span>
-        <input id="ng-days" type="number" min="1" max="20" step="1" value="5"></label>
-    </div>
-    <p class="ng-note">The same seed and the same settings always deal the same railroad, so a game
-       can be shared, compared or replayed. Leave it blank for a random one.</p>
-
-    <h3>Game type</h3>
-    <div class="set-row" id="ng-type-row">
-      <label class="ng-radio"><input type="radio" name="ng-type" value="solitaire">
-        <span><b>Solitaire</b><br><span class="dim">One railroad, one player. The whole Division is yours to run.</span></span></label>
-      <label class="ng-radio"><input type="radio" name="ng-type" value="coop" checked>
-        <span><b>Co-op</b><br><span class="dim">Everyone&#8217;s Revenue is one table score. You win together or lose together.</span></span></label>
-      <label class="ng-radio"><input type="radio" name="ng-type" value="competitive">
-        <span><b>Competitive</b><br><span class="dim">Highest Revenue wins — unless the table misses its combined minimum, and then everyone loses.</span></span></label>
-      <label class="ng-radio"><input type="radio" name="ng-type" value="cutthroat">
-        <span><b>Cutthroat</b><br><span class="dim">Highest Revenue wins, and nothing is shared — the only way everyone loses is three collisions in one Day.</span></span></label>
-      <label class="ng-radio"><input type="radio" name="ng-type" value="custom">
-        <span><b>Custom</b><br><span class="dim">Whatever you set below. Selected for you the moment you change a rule; it is scored as the type you started from.</span></span></label>
-    </div>
-
-    <p class="ng-note" id="ng-type-note"></p>
-
-    <details id="ng-settings" open>
-      <summary>Game settings</summary>
-      <p class="ng-note">Every rule the game type sets, and every one of them yours to change.
-         Changing any of them selects <b>Custom</b>; clicking a type again resets all of them back
-         to it.</p>
-      <div class="set-groups">
-
-      <div class="set-group">
-      <h3>Starting hand</h3>
-      <p class="ng-note">What each player is dealt before the first turn. The hand limit is three
-         either way — deal six and the first turn is spent choosing which of them to keep.</p>
-      <div class="set-row" id="ng-hand-row">
-        <label class="ng-radio"><input type="radio" name="ng-hand" value="threeRandom">
-          <span><b>Three random cards</b><br><span class="dim">The original rule. At the hand limit already, and no guarantee of track.</span></span></label>
-        <label class="ng-radio"><input type="radio" name="ng-hand" value="sixRandom" checked>
-          <span><b>Six random cards</b><br><span class="dim">Twice the choice, still no guaranteed track — the first turn is a discard.</span></span></label>
-        <label class="ng-radio"><input type="radio" name="ng-hand" value="threeTrackThreeOther">
-          <span><b>Three random track and three random non-track cards</b><br><span class="dim">Dealt from two piles, so the district you can build is dealt rather than waited for.</span></span></label>
-        <span class="set-hint" id="ng-hand-hint"></span>
-      </div>
-      </div>
-
-      <div class="set-group">
-      <h3>Where an Extra may start</h3>
-      <p class="ng-note">The player who plays an Extra Train card chooses where its Crew Tray goes,
-         and the place decides which way it runs — a Division Point sends it away from itself; in the
-         middle of the railroad the player picks east or west. The Division Points and the Interchange
-         belong to nobody and are always available. Starting one inside a district is the part that
-         favours a seat, so it is set here. An Office must be a Control Point whatever this says: a
-         Whistle Post never qualifies.</p>
-      <div class="set-row" id="ng-extra-row">
-        <label class="ng-radio"><input type="radio" name="ng-extra" value="divisionPointsOnly">
-          <span><b>Division Points and the Interchange only</b><br><span class="dim">The strictest reading. Every Extra begins on shared ground.</span></span></label>
-        <label class="ng-radio"><input type="radio" name="ng-extra" value="ownOffice">
-          <span><b>Also the playing player&#8217;s own Control Point</b><br><span class="dim">You may start one at home, but not in somebody else&#8217;s district.</span></span></label>
-        <label class="ng-radio"><input type="radio" name="ng-extra" value="anyOffice">
-          <span><b>Also any player&#8217;s Control Point</b><br><span class="dim">The most permissive — an Extra may be planted in another player&#8217;s district.</span></span></label>
-        <span class="set-hint" id="ng-extra-hint"></span>
-      </div>
-      </div>
-
-      <div class="set-group">
-      <h3>Revenue</h3>
-      <p class="ng-note">What each piece of work pays, 0 to 5. A coach pays when it is boarded and
-         again when it is detrained; a load pays when it is made up and again when it is broken. Zero
-         switches an economy off so the others can be read.</p>
-      <div class="set-row" id="ng-passenger-row">
-        <label class="ng-num"><span>Passenger revenue per coach</span>
-          <input id="ng-passenger" type="number" min="0" max="5" step="1" value="1"></label>
-        <span class="set-hint" id="ng-passenger-hint"></span>
-      </div>
-      <div class="set-row" id="ng-freight-row">
-        <label class="ng-num"><span>Freight revenue per load</span>
-          <input id="ng-freight" type="number" min="0" max="5" step="1" value="1"></label>
-        <span class="set-hint" id="ng-freight-hint"></span>
-      </div>
-      <div class="set-row" id="ng-transit-row">
-        <label class="ng-num"><span>Train revenue per transit</span>
-          <input id="ng-transit" type="number" min="0" max="5" step="1" value="0"></label>
-        <span class="set-hint" id="ng-transit-hint"></span>
-      </div>
-      <p class="ng-note">A transit pays every player, once, when a train runs off the end of the
-         Division — the one thing nobody has to work for.</p>
-      </div>
-
-      <div class="set-group">
-      <h3>Victory conditions</h3>
-      <p class="ng-note">The ways this game can end badly. Each one is switched on or off in its own
-         right; how long the game runs is set above, with the table size.</p>
-      <div class="set-row" id="ng-minrev-row">
-        <label class="ng-gate"><input type="checkbox" id="ng-minrev-on" checked>
-          <span>Everyone loses if combined Revenue at the end is under</span>
-          <input id="ng-minrev" type="number" min="0" step="1" class="gate-num"></label>
-        <span class="set-hint" id="ng-minrev-hint"></span>
-      </div>
-      <div class="set-row" id="ng-colday-row">
-        <label class="ng-gate"><input type="checkbox" id="ng-colday-on" checked>
-          <span>The game ends and everyone loses if collisions in one Day reach</span>
-          <input id="ng-colday" type="number" min="0" step="1" class="gate-num"></label>
-        <span class="set-hint" id="ng-colday-hint"></span>
-      </div>
-      <div class="set-row" id="ng-coltotal-row">
-        <label class="ng-gate"><input type="checkbox" id="ng-coltotal-on" checked>
-          <span>The game ends and everyone loses after this many collisions in the whole game</span>
-          <input id="ng-coltotal" type="number" min="0" step="1" class="gate-num"></label>
-        <span class="set-hint" id="ng-coltotal-hint"></span>
-      </div>
-      <p class="ng-note">The opponent-directed cards — Derail, Watertower, Hobo Jungle and the
-         nineteen others, along with the seven that answer them — are not implemented yet, so no game
-         type deals them whatever else is set here.</p>
-      </div>
-
-      <div class="set-group">
-      <h3>Optional rules</h3>
-      <p class="ng-note">Off in every game type; each one changes how the game plays.</p>
-      <div class="set-row" id="ng-visibility-row">
-        <label class="ng-num"><span>Reduced Visibility — five switching Moves instead of six in the
-          night Stages (1&ndash;3 and 11&ndash;12)</span>
-          <input id="ng-visibility" type="checkbox"></label>
-        <span class="set-hint" id="ng-visibility-hint"></span>
-      </div>
-      <div class="set-row" id="ng-rotation-row">
-        <label class="ng-num"><span>Employee Rotation — at the end of each Day everyone moves one
-          chair left and takes over the next station up the line. Your Revenue and the Fedora go with
-          you; the district stays where it is</span>
-          <input id="ng-rotation" type="checkbox"></label>
-        <span class="set-hint" id="ng-rotation-hint"></span>
-      </div>
-      <div class="set-row" id="ng-toolbox-row">
-        <label class="ng-num"><span>Emergency Toolbox — everyone starts holding a Red Flag, so a hand
-          of four; play or discard down to three on the first turn</span>
-          <input id="ng-toolbox" type="checkbox"></label>
-        <span class="set-hint" id="ng-toolbox-hint"></span>
-      </div>
-      <div class="set-row" id="ng-tossloco-row">
-        <label class="ng-num"><span>A Timetabled train may be discarded — toss it face-up to a
-          Department slot, where a rival may pick it up. Turn this off and a train card can only ever
-          be played onto the timetable. An Extra is never discardable either way</span>
-          <input id="ng-tossloco" type="checkbox"></label>
-        <span class="set-hint" id="ng-tossloco-hint"></span>
-      </div>
-      </div>
-
-      </div>
-    </details>
-
-    <menu class="ng-buttons">
-      <span class="ng-note" id="ng-multiplayer-note" style="margin:0 auto 0 0">Use the <b>Multiplayer</b> button instead — it creates or joins a game on this server.</span>
-      <button value="cancel" id="ng-cancel" type="submit" formnovalidate>Cancel</button>
-      <button value="deal" id="ng-deal" type="submit">Deal</button>
-    </menu>
-  </form>
-</dialog>
-
 <!-- THE DAY ROLLING OVER (Gitea#10). A Day turns inside the automatic phases, so it happens
      between one click and the next; the phase banner and the announcement flash both fade before
      someone reading the board notices them. A modal stops and waits, which is the whole request:
@@ -901,10 +950,19 @@ ul.blocked li{padding:2px 0}
 <!-- THE END-OF-GAME RESULTS (Gitea#16). Filled by `resultsHtml` and opened from `renderEnding`,
      which puts it up once per ending unasked and leaves a button to reopen it. Reopenable matters:
      Gitea#11 lets a table play past the end, and continuing must not cost you the results screen. -->
+<!-- THE EXTENSION QUESTION IS ASKED HERE, not only behind this dialog (Gitea#11 + #16).
+     This opens ITSELF at every ending, and it is modal — so `renderEnding`'s "play one more Day"
+     buttons, which it writes into `#actions`, sit underneath it. A player saw a results screen whose
+     only control was Close and reasonably concluded the game was over: reported by Jesse
+     2026-08-30, "Solitaire game ended. I did not have an option to extend the game by a day."
+     Gitea#11 was verified over the HTTP API, which renders no dialog, so the browser never was.
+     The two extension buttons are hidden unless the game is actually awaiting a vote. -->
 <dialog id="resultsdlg" aria-labelledby="rs-title">
   <form method="dialog">
     <div id="resultsbody"></div>
     <menu class="ng-buttons">
+      <button value="extend-yes" id="rs-extend-yes" type="submit" hidden>Play One More Day</button>
+      <button value="extend-no" id="rs-extend-no" type="submit" hidden>End the Game Here</button>
       <button value="ok" id="rs-ok" type="submit">Close</button>
     </menu>
   </form>

src/web/presets.ts

@@ -119,8 +119,14 @@ export const PRESETS: readonly Preset[] = [
     revenueFloor: (players, days) => collectiveRevenueFloor(players, days),
     rules: {
       startingHand: SIX,
-      // Nobody else's district exists, so "any Control Point" and "your own" are the same rule.
-      extraStart: 'anyOffice',
+      /**
+       * Nobody else's district exists, so "any Control Point" and "your own" are the same rule —
+       * `apply.ts` only ever rejects `ownOffice` when `start.seat !== seatOf(s, player)`, which
+       * cannot happen at one seat. It said `anyOffice` until 2026-08-30, which was true and read
+       * wrong: a solitaire player has no "any player" to contrast themselves with, so the permissive
+       * label described a permission nobody was being granted. Jesse's call; no gameplay effect.
+       */
+      extraStart: 'ownOffice',
       passengerPerCoach: 1,
       freightPerLoad: 1,
       trainPerTransit: 0,

test/advance.test.ts

@@ -655,13 +655,36 @@ describe('victory conditions (§3, Gap 10e) — unified 2026-08-20', () => {
     assert.notEqual(s.status, 'finished');
   });
 
-  it('solitaire never checks the collision floor, whatever the counts', () => {
+  it('solitaire checks the collision floor too, like every other mode', () => {
+    /**
+     * REVERSED 2026-08-30, and this test previously asserted the opposite ("solitaire never checks
+     * the collision floor, whatever the counts").
+     *
+     * The exclusion was never a stated rule — §3.4 does not carve solitaire out — and nothing on
+     * screen reflected it: `SOLO_CONFIG` carried both limits, the New Game dialog offered them as
+     * live settings, and the text beside them said the game would end in a loss. A solitaire player
+     * could set a limit of 1 and crash all game. Found reviewing that screen's wording; Jesse's
+     * ruling is that the settings do what they say.
+     */
     const s = game(1, { mode: 'solitaire', maxCollisionsPerDay: 1, maxCollisionsTotal: 1 });
     s.collisionsToday = 99;
     s.collisionsTotal = 99;
     s.clock.stage = 1;
     s.clock.phase = 'shiftChange';
     advance(s);
+    assert.equal(s.status, 'finished');
+    assert.equal(s.outcome!.reason, 'collisionFloor');
+  });
+
+  it('still lets a solitaire game switch the collision floor off with 0', () => {
+    // The disable path is what a player who does not want the new ending reaches for, so it has to
+    // work at one seat exactly as it does at four.
+    const s = game(1, { mode: 'solitaire', maxCollisionsPerDay: 0, maxCollisionsTotal: 0 });
+    s.collisionsToday = 99;
+    s.collisionsTotal = 99;
+    s.clock.stage = 1;
+    s.clock.phase = 'shiftChange';
+    advance(s);
     assert.notEqual(s.status, 'finished');
   });
 });

test/presets.test.ts

@@ -59,8 +59,11 @@ describe('what each game type is', () => {
     assert.equal(presetSettings('cutthroat', 4, 5).extraStart, 'anyOffice');
     assert.equal(presetSettings('coop', 4, 5).extraStart, 'ownOffice');
     assert.equal(presetSettings('competitive', 4, 5).extraStart, 'ownOffice');
-    // At one player the two rules are the same rule.
-    assert.equal(presetSettings('solitaire', 1, 5).extraStart, 'anyOffice');
+    // At one player the two rules ARE the same rule — `apply.ts` only rejects `ownOffice` when the
+    // start is another seat's, which cannot happen. Solitaire said `anyOffice` until 2026-08-30:
+    // true, and it read wrong, since a lone player has no "any player" to be contrasted with. The
+    // label changed and the behaviour did not.
+    assert.equal(presetSettings('solitaire', 1, 5).extraStart, 'ownOffice');
   });
 
   it('leaves every optional rule off, in every type', () => {

test/replay.test.ts

@@ -214,6 +214,24 @@ describe('a blocked platform says why (Gitea#2)', () => {
     assert.ok(f.inboundBox.length === 0, 'the red slots were free — the shortage is the only cause');
   });
 
+  it('says why the game ended in words, not as a raw enum (TODO #34)', () => {
+    /**
+     * The heading read `loss — revenueFloor` — the exact defect Gitea#16 was filed about on the
+     * playable page, still alive here a release after that was fixed, because nothing
+     * player-facing pointed at the developer replay. It shares `reasonSentence` with the results
+     * screen now, so the two cannot explain one ending in two ways.
+     */
+    const rec = record(1234, 'standard');
+    for (const raw of ['revenueFloor', 'daysElapsed', 'collisionFloor']) {
+      assert.ok(!rec.outcome.includes(raw), `the summary still prints the raw reason "${raw}"`);
+    }
+    assert.doesNotMatch(rec.outcome, /<[^>]+>/, 'markup leaked into a heading and a console line');
+    assert.match(rec.outcome, /Revenue/, 'the summary says nothing about how the game went');
+    // And the sentence is the shared one, with this game's own numbers in it.
+    assert.match(rec.outcome, /closed short|last on the timetable|declared unsafe/,
+      'the ending is not explained in the words the results screen uses');
+  });
+
   it('says nothing about a platform that is working fine', () => {
     // Passengers waiting AND a train with an empty coach to take them: no impediment.
     const s = createGame({ id: 'g', seed: 5, config, playerNames: ['p'] });
@@ -257,7 +275,10 @@ describe('replay recording', () => {
     const last = rec.frames[rec.frames.length - 1]!;
     assert.equal(last.revenue, stats.revenue.net, 'final revenue disagrees with the engine');
     assert.equal(last.day, s.clock.day, 'final Day disagrees with the engine');
-    assert.match(rec.outcome, new RegExp(stats.result));
+    // The summary says "won"/"lost" rather than the engine's `win`/`loss` (TODO #34 — it is a
+    // sentence for a reader now, not an enum). Mapped here so this still checks the two AGREE,
+    // which is what the test is for, rather than checking they are spelled the same.
+    assert.match(rec.outcome, new RegExp(stats.result === 'win' ? 'won' : 'lost'));
   });
 
   it('narrates every frame', () => {

test/web.test.ts

@@ -21,6 +21,7 @@ import { URLSearchParams as NodeURLSearchParams } from 'node:url';
 import { cardDescription, cardName, describeIntent, variantLabel } from '../src/sim/view.ts';
 import { variantsFor } from '../src/engine/track.ts';
 import { divisionSvg, officeSvg } from '../src/sim/board-svg.ts';
+import type { DivisionView } from '../src/sim/view.ts';
 import { ENHANCEMENT_RULES, STAGES_PER_DAY } from '../src/engine/content.ts';
 import { dayEndHtml, facilitiesHtml, resultsHtml, timetableHtml } from '../src/web/panels.ts';
 import { turnChartHtml } from '../src/sim/turnchart.ts';
@@ -32,6 +33,8 @@ import { createGame as createEngineGame } from '../src/engine/setup.ts';
 import { advance as advanceEngine } from '../src/engine/advance.ts';
 import type { GameConfig } from '../src/engine/state.ts';
 import {
+  configWith,
+  SOLO_CONFIG,
   overHandLimit,
   actionGroups,
   actionMenu,
@@ -2546,6 +2549,47 @@ describe('the static build', () => {
     }
   });
 
+  it('names who the game is waiting on when it stops to ask them something', () => {
+    /**
+     * REPORTED BY JESSE 2026-08-30: "waiting on shows 'nobody — the Division is running itself' BUT
+     * the system is actually waiting on the Superintendent."
+     *
+     * `Frame.actor` carried `clock.currentActor`, which is null for the whole Mainline Phase — so a
+     * game stopped dead on a §8.1 clearance ruling said nobody was holding it up, while it waited on
+     * a named person to click. `actingPlayer` had the answer the whole time; the Frame threw it
+     * away. Naming them is only half of it: three different interruptions can be pending, and
+     * "waiting on Bob" alone is a game that looks stuck to everyone except Bob.
+     */
+    const base = { day: 2, stage: 5, clock: '2:20', phase: 'Mainline', phaseKey: 'mainline' };
+
+    const idle = turnChartHtml({ ...base, actor: null, awaiting: null }, null, 'Bob');
+    assert.match(idle, /nobody — the Division is running itself/, 'an automatic phase should say so');
+
+    for (const [asks, train] of [
+      ['a clearance ruling', 'Train 4'],
+      ['the Yard Office offer', 'Train 7'],
+      ['a Red Flag', 'Train X18'],
+    ] as [string, string][]) {
+      const html = turnChartHtml({ ...base, actor: 1, awaiting: { asks, train } }, 'Bob', 'Bob');
+      assert.doesNotMatch(html, /running itself/, `"${asks}" still reported nobody`);
+      assert.match(html, /waiting on <b>Bob<\/b>/, `"${asks}" did not name who it waits on`);
+      assert.match(html, new RegExp(asks.replace(/ /g, ' ')), `"${asks}" did not say what is being asked`);
+      assert.match(html, new RegExp(train), `"${asks}" did not name the train it is about`);
+    }
+  });
+
+  it('puts the Fedora at the right-hand end of the phase row (TODO #29)', () => {
+    // It sat on a line of its own between the phases and everything above them, which put a thing
+    // that moves every third Stage in among the things that move every Stage. The row it belongs
+    // beside is the one whose last chip is Supervisor Shift — the phase that passes it.
+    const frame = { day: 1, stage: 4, clock: '2:00', phase: 'New Train', phaseKey: 'newTrain', actor: 1 };
+    const html = turnChartHtml(frame, 'Bob', 'Bob');
+    const row = /<div class="tc-row">([\s\S]*?)<\/div>\s*$/.exec(html)?.[1] ?? '';
+    assert.match(row, /<ol class="tc-phases">/, 'the phase row is not in the row wrapper');
+    assert.ok(row.indexOf('tc-super') > row.indexOf('tc-phases'), 'the Fedora is not after the phases');
+    assert.match(turnChartHtml(frame, 'Bob', null), /tc-row/, 'solitaire lost the row wrapper with the Fedora');
+  });
+
   it('names the Superintendent at a table, and stays quiet about it in solitaire', () => {
     /**
      * REPORTED BY JESSE 2026-08-23, playing two-player on StartOS: seat 1 played a train card and
@@ -2622,10 +2666,42 @@ describe('the static build', () => {
       assert.ok(!/https?:\/\//.test(html.replace(/<!--[\s\S]*?-->/g, '')), `${name} fetches something external`);
     }
     const splash = readFileSync(join(dist, 'index.html'), 'utf8');
-    assert.match(splash, /href="\.\/play\.html"/, 'the splash does not link to the game');
+    // `?solitaire` marks the door's intent explicitly (2026-08-29) so a browser that remembers a
+    // multiplayer session cannot swallow it — see `start()`'s own comment on `wantsSolitaire`.
+    assert.match(splash, /href="\.\/play\.html\?solitaire"/, 'the splash does not link to the game');
     assert.match(splash, /href="\.\/replays\.html"/, 'the splash does not link to the replays');
   });
 
+  it('cache-busts with a tag that varies per build even where there is no git', () => {
+    /**
+     * THE BUG THIS PINS COST TWO RELEASES. `buildStamp`'s no-git fallback was the literal `nogit`,
+     * and the `.s9pk` Dockerfile copies the working tree in WITHOUT `.git` — so every packaged
+     * release published `?v=nogit`, byte-identical to the one before it, and a returning player's
+     * browser refetched nothing. v0.7.5's setup screen and v0.7.6's fix to it both installed
+     * correctly on `phoenix.local` and neither reached the browser that asked for them.
+     *
+     * Asserted against the SCRIPT rather than a built page, because the property is about what the
+     * fallback does when `git rev-parse` fails, which a normal build here never exercises.
+     */
+    const src = readFileSync(join(root, 'scripts/build-web.ts'), 'utf8');
+    const fallback = /let git = ([^;]+);/.exec(src)?.[1] ?? '';
+    assert.ok(fallback !== '', 'the no-git fallback moved and this test cannot see it any more');
+    assert.doesNotMatch(fallback, /^'nogit'$|^"nogit"$/, 'the no-git fallback is a constant again');
+    assert.match(fallback, /Date\.now\(\)/, 'the no-git fallback carries nothing that varies per build');
+  });
+
+  it('lets a build-tagged URL be cached and nothing else', () => {
+    // The other half of the same bug: the pages carry the `?v=` tags but cannot be versioned in
+    // their own URL, so a cached `play.html` pins a player to the whole build it names. Only a
+    // request that actually carries `?v=` may be stored — an untagged image or the replay manifest
+    // has no way to announce a change.
+    const src = readFileSync(join(root, 'src/server/http.ts'), 'utf8');
+    assert.match(src, /'Cache-Control':\s*buildTagged\s*\?/, 'static responses no longer vary their caching');
+    assert.match(src, /immutable/, 'a tagged asset is not allowed to be cached at all');
+    assert.match(src, /serveStatic\(opts\.distDir, url\.pathname, res, url\.searchParams\.has\('v'\)\)/,
+      'the ?v= tag is not reaching serveStatic, so every response falls back to no-cache');
+  });
+
   it('opens the multiplayer door from the splash, straight into the lobby', () => {
     // This door sat `disabled` and labelled "Coming soon" from before the server existed until
     // v0.5.2 — Phases 2-4 built a working lobby and nothing ever linked to it, so a player with a
@@ -2800,6 +2876,107 @@ describe('the Division map shows the whole route', () => {
     return divisionSvg(snapshot(s, [], null).division);
   };
 
+  it('draws which way a Heavy Grade climbs, instead of only saying it in the tooltip', () => {
+    /**
+     * REPORTED BY JESSE 2026-08-30: "heavy grade mainline card tooltip states climbs east, but card
+     * doesn't show it." The Frame has carried `gradeUp` since the Division map was rebuilt, and the
+     * tip has read "climbs east" all along — but the one Mainline card whose orientation the PLAYER
+     * sets, and the one where Helpers and Brakeman mean opposite things at opposite ends, drew
+     * nothing to distinguish the two.
+     *
+     * Asserted on the geometry rather than on a screenshot: what can actually go wrong here is the
+     * wedge pointing the wrong way or hanging off the card, and both are numbers.
+     */
+    const card = (gradeUp: string | null): DivisionView =>
+      ({
+        kind: 'ml', label: 'Heavy Grade', trains: [], capacity: 1,
+        modifiers: [], gradeUp, regions: 3, what: 'three regions',
+      } as unknown as DivisionView);
+
+    const peakOf = (svg: string): [number, number][] | null => {
+      const m = /<polygon class="bs-grade" points="([^"]+)"/.exec(svg);
+      return m ? (m[1]!.split(' ').map((p) => p.split(',').map(Number) as [number, number])) : null;
+    };
+
+    // EAST IS RIGHT on this map (Gitea#18), which is the whole reason a wedge can be read without a
+    // compass — so the apex belongs at the greater x when the grade climbs east, and the lesser when
+    // it climbs west.
+    const east = peakOf(divisionSvg([card('east')]));
+    assert.ok(east, 'a Heavy Grade climbing east draws no wedge at all');
+    const [ea, eb, epeak] = east;
+    assert.equal(epeak![0], Math.max(ea![0], eb![0]), 'the wedge climbs the wrong way for east');
+
+    const west = peakOf(divisionSvg([card('west')]));
+    assert.ok(west, 'a Heavy Grade climbing west draws no wedge at all');
+    const [wa, wb, wpeak] = west;
+    assert.equal(wpeak![0], Math.min(wa![0], wb![0]), 'the wedge climbs the wrong way for west');
+
+    // Every other Mainline card is flat and must stay unmarked, or the wedge stops meaning anything.
+    assert.equal(peakOf(divisionSvg([card(null)])), null, 'a card with no grade drew one anyway');
+
+    /**
+     * THE ARROW LIVES INSIDE THE WEDGE, with room to spare.
+     *
+     * The first pass ran it up the hypotenuse, which starts at the wedge's thin corner where there
+     * is no height to draw in, so the head sat over the edge and read as clipped — reported exactly
+     * that way. A bounding-box check would have passed it: every point was on the card. What matters
+     * is containment in the TRIANGLE, and a margin big enough to see, so both are asserted.
+     */
+    for (const dir of ['east', 'west'] as const) {
+      const svg = divisionSvg([card(dir)]);
+      const wedge = peakOf(svg)!;
+      const [w1, w2, apex] = wedge;
+      const yBase = w1![1];
+      const xMin = Math.min(w1![0], w2![0]);
+      const xMax = Math.max(w1![0], w2![0]);
+      const arrow = /<polygon class="bs-gradeup" points="([^"]+)"/.exec(svg)?.[1];
+      assert.ok(arrow, `the ${dir} grade draws no arrow`);
+
+      const av = arrow.trim().split(/\s+/).map((p) => p.split(',').map(Number) as [number, number]);
+      for (const [ax, ay] of av) {
+        // 0 at the thin corner, 1 at the apex — the wedge's height scales with it.
+        const t = apex![0] === xMax ? (ax - xMin) / (xMax - xMin) : (xMax - ax) / (xMax - xMin);
+        const ceiling = yBase - (yBase - apex![1]) * t;
+        assert.ok(ax >= xMin && ax <= xMax, `the ${dir} arrow leaves the wedge sideways at ${ax}`);
+        assert.ok(
+          yBase - ay >= 2 && ay - ceiling >= 2,
+          `the ${dir} arrow comes within 2px of the wedge edge at ${ax},${ay} — it will read as clipped`,
+        );
+      }
+
+      /**
+       * IT LIES ALONG THE WEDGE'S OWN SLOPE, AND IS CENTRED IN THE FORM.
+       *
+       * Derived from the wedge rather than hardcoded, so resizing the wedge cannot leave the arrow
+       * at a stale angle — the assertion is "parallel", which is the property that makes the gap to
+       * the hypotenuse constant along the arrow instead of closing at one end.
+       *
+       * Measured on the AXIS: the two tail vertices straddle it by the shaft's half-thickness, so
+       * measuring from a corner under-reads by a few degrees. That mistake cost a wrong reading once
+       * already.
+       */
+      const tail: [number, number] = [(av[0]![0] + av[6]![0]) / 2, (av[0]![1] + av[6]![1]) / 2];
+      const tip = av[3]!;
+      const dx = tip[0] - tail[0];
+      const climb = (Math.atan2(-(tip[1] - tail[1]), Math.abs(dx)) * 180) / Math.PI;
+      const slope = (Math.atan2(yBase - apex![1], xMax - xMin) * 180) / Math.PI;
+      assert.ok(
+        Math.abs(climb - slope) < 0.5,
+        `the ${dir} arrow climbs at ${climb.toFixed(1)}° but the wedge rises at ${slope.toFixed(1)}°`,
+      );
+      assert.equal(dx > 0, dir === 'east', `the ${dir} arrow points away from the summit`);
+
+      // Centred on the TRIANGLE's centroid — 2/3 along the base toward the apex, 1/3 up — not on
+      // the bounding box, which would put it in the thin corner where there is no height for it.
+      const midX = (tail[0] + tip[0]) / 2;
+      const midY = (tail[1] + tip[1]) / 2;
+      const cX = apex![0] === xMax ? xMin + ((xMax - xMin) * 2) / 3 : xMax - ((xMax - xMin) * 2) / 3;
+      const cY = yBase - (yBase - apex![1]) / 3;
+      assert.ok(Math.abs(midX - cX) < 1, `the ${dir} arrow is off the centroid horizontally`);
+      assert.ok(Math.abs(midY - cY) < 1, `the ${dir} arrow is off the centroid vertically`);
+    }
+  });
+
   it('expands each Office into its Running Track, Limits to Limits', () => {
     // The map used to collapse a whole district into one "Office" box, so the track a train
     // actually runs along was invisible on the only view that shows where trains are.
@@ -3307,6 +3484,34 @@ describe('the Day rolling over says so (Gitea#10)', () => {
 
 // ---------------------------------------------------------------------------
 
+describe('configWith', () => {
+  /**
+   * The Revenue floor is a function of how long the game is, so a helper that takes `days` and
+   * defaults the floor has to derive it. It fell back to `SOLO_CONFIG`'s five-Day constant until
+   * 2026-08-30, which let the two fields disagree: a one-Day game was asked to clear 15, a floor a
+   * five-Day game averages barely half of. Found by a probe that passed only `days` — which is how
+   * the next caller would reach for it.
+   */
+  it('derives the Revenue floor from the days it was given', () => {
+    assert.equal(configWith({ days: 1 }).minCombinedRevenue, 3, 'a one-Day game kept the five-Day floor');
+    assert.equal(configWith({ days: 10 }).minCombinedRevenue, 30, 'a ten-Day game kept the five-Day floor');
+  });
+
+  it('leaves the default day count exactly where it was', () => {
+    // The fix must not move the floor anybody is actually playing against: SOLO_CONFIG's own floor
+    // is this same formula at DEFAULT_DAYS, so the default path is unchanged.
+    assert.equal(configWith({}).minCombinedRevenue, SOLO_CONFIG.minCombinedRevenue);
+    assert.equal(configWith({ days: SOLO_CONFIG.days }).minCombinedRevenue, SOLO_CONFIG.minCombinedRevenue);
+  });
+
+  it('still lets a caller name a floor that has nothing to do with the length', () => {
+    // Deriving is the DEFAULT, not a rule — the New Game dialog and the lobby both let a player set
+    // a floor directly, and that has to survive being passed through here.
+    assert.equal(configWith({ days: 10, minCombinedRevenue: 4 }).minCombinedRevenue, 4);
+    assert.equal(configWith({ days: 3, minCombinedRevenue: 0 }).minCombinedRevenue, 0, 'an off floor was re-derived');
+  });
+});
+
 describe('the end-of-game results screen (Gitea#16)', () => {
   /**
    * A finished game, built by running the clock off the end of a real one rather than by hand — the
@@ -3330,6 +3535,36 @@ describe('the end-of-game results screen (Gitea#16)', () => {
     return snapshot(s, [], null);
   };
 
+  it('asks the extension question ON the results dialog, not only behind it (Gitea#11)', () => {
+    /**
+     * REPORTED BY JESSE 2026-08-30: "Solitaire game ended. I did not have an option to extend the
+     * game by a day."
+     *
+     * The engine and the Frame were right all along — a solitaire game at the end of its timetable
+     * reaches `awaitingExtension` with an uncast vote, asserted below. What went wrong is that
+     * `renderEnding` writes its two buttons into `#actions` and then opens `#resultsdlg`, which is
+     * MODAL: the question was underneath a dialog whose only control was Close. Gitea#11 was
+     * verified over the HTTP API, which renders no dialog, so the browser path never was.
+     */
+    const f = finished();
+    assert.equal(f.status, 'awaitingExtension', 'a solitaire game no longer pauses to ask');
+    assert.equal(f.extensionVotes[f.viewer], null, 'the viewer has somehow already voted');
+
+    const html = readFileSync(join(dist, 'play.html'), 'utf8');
+    for (const id of ['rs-extend-yes', 'rs-extend-no']) {
+      assert.ok(html.includes(`id="${id}"`), `the results dialog cannot ask: no #${id}`);
+    }
+
+    // And they are wired: shown only while a vote is pending, and casting the same intent the
+    // `#actions` buttons do. Read from the source, since the dialog needs a DOM to drive.
+    const src = readFileSync(join(root, 'src/web/main.ts'), 'utf8');
+    const fn = /function showResults\(f: Frame\): void \{[\s\S]*?\n\}/.exec(src)?.[0] ?? '';
+    assert.ok(fn !== '', 'showResults moved and this test cannot see it');
+    assert.match(fn, /awaitingExtension/, 'the dialog does not know the game is asking');
+    assert.match(fn, /'game\.extend'/, 'the dialog offers no way to answer');
+    assert.match(fn, /yes\.hidden = !asking/, 'the buttons are not hidden on a game that is over');
+  });
+
   it('never prints a raw enum at the player', () => {
     // The bug the issue opens on: `GAME OVER — revenueFloor` is an internal identifier, shown at
     // the one moment the game has the player's whole attention.
@@ -3545,9 +3780,9 @@ describe('the lobby screen', () => {
       'lb-hand': group(['threeRandom', 'sixRandom', 'threeTrackThreeOther'], 'sixRandom'),
       'lb-extra': group(['divisionPointsOnly', 'ownOffice', 'anyOffice'], 'anyOffice'),
       'lb-type': group(['solitaire', 'coop', 'competitive', 'cutthroat', 'custom'], 'coop'),
-      'ng-hand': group(['threeRandom', 'sixRandom', 'threeTrackThreeOther'], 'sixRandom'),
-      'ng-extra': group(['divisionPointsOnly', 'ownOffice', 'anyOffice'], 'anyOffice'),
-      'ng-type': group(['solitaire', 'coop', 'competitive', 'cutthroat', 'custom'], 'solitaire'),
+      'ss-hand': group(['threeRandom', 'sixRandom', 'threeTrackThreeOther'], 'sixRandom'),
+      'ss-extra': group(['divisionPointsOnly', 'ownOffice', 'anyOffice'], 'anyOffice'),
+      'ss-type': group(['solitaire', 'coop', 'competitive', 'cutthroat', 'custom'], 'solitaire'),
     };
     const matching = (sel: string): Radio[] => {
       const name = /name="([^"]+)"/.exec(sel)?.[1] ?? '';
@@ -3671,8 +3906,13 @@ describe('the lobby screen', () => {
     els.get('lb-freight')!['value'] = '3';
     (els.get('lb-freight')!['oninput'] as () => void)();
     assert.equal(chosen(groups, 'lb-type'), 'custom', 'changing a RULE should have selected Custom');
-    assert.match(String(els.get('lb-type-note')!['textContent']), /scored as Co-op/);
-    assert.match(String(els.get('lb-type-note')!['textContent']), /1 setting differs from Co-op/);
+    // The summary sentence under the radios is gone (2026-08-30). What says HOW a Custom game
+    // differs is the hint on the row that differs, which is where it can be acted on.
+    assert.match(
+      String(els.get('lb-freight-hint')!['textContent']),
+      /Co-op default/,
+      'the changed row does not say what it was changed from',
+    );
   });
 
   it('clicking a type again resets every rule, and leaves the parameters alone', async () => {
@@ -3800,7 +4040,7 @@ describe('the lobby screen', () => {
   });
 });
 
-describe('the lobby and the dialog ask the same questions', () => {
+describe('the lobby and the setup screen ask the same questions', () => {
   /**
    * THE DRIFT GUARD.
    *
@@ -3816,8 +4056,10 @@ describe('the lobby and the dialog ask the same questions', () => {
   };
 
   it('carries every field of the shared block on both screens', () => {
+    // `ss-` joined `lb-`/`ng-` 2026-08-29: the pre-game solitaire setup screen drives the identical
+    // block ("asking first is the only path"). Same drift guard, one more prefix.
     const html = page();
-    for (const prefix of ['lb-', 'ng-']) {
+    for (const prefix of ['lb-', 'ss-']) {
       for (const selector of fieldSelectors(prefix)) {
         assert.ok(html.includes(selector), `the ${prefix} block is missing ${selector}`);
       }
@@ -3826,7 +4068,7 @@ describe('the lobby and the dialog ask the same questions', () => {
 
   it('offers all five game types on both screens', () => {
     const html = page();
-    for (const prefix of ['lb-', 'ng-']) {
+    for (const prefix of ['lb-', 'ss-']) {
       for (const type of ['solitaire', 'coop', 'competitive', 'cutthroat', 'custom']) {
         assert.ok(
           html.includes(`name="${prefix}type" value="${type}"`),
@@ -3853,13 +4095,17 @@ describe('the lobby and the dialog ask the same questions', () => {
     // The opponent-directed cards are unbuilt, and `buildDeck` holds them out however the config is
     // set — so the checkbox could not do anything, on either screen. The fact is stated in words.
     const html = page();
-    assert.ok(!html.includes('id="ng-pvp"'), 'the dialog still has the dead PvP checkbox');
     assert.ok(!html.includes('id="lb-pvp"'), 'the lobby still has the dead PvP checkbox');
+    assert.ok(!html.includes('id="ss-pvp"'), 'the setup screen still has the dead PvP checkbox');
+    // The in-game dialog was a THIRD copy of this block and the one that drifted — it kept the
+    // multiplayer wording on a solitaire-only screen. Deleted 2026-08-30; nothing may reintroduce
+    // a prefix that no screen owns.
+    assert.ok(!/id="ng-/.test(html), 'the deleted in-game dialog has come back');
     assert.match(html, /opponent-directed cards[\s\S]{0,120}not implemented yet/i);
   });
 });
 
-describe('the New Game dialog', () => {
+describe('the solitaire setup screen', () => {
   /**
    * DRIVEN THROUGH THE EMITTED BUNDLE, like the highlight test above, because the thing that can go
    * wrong here is wiring rather than logic: an id that does not match the HTML, a handler on the
@@ -3873,7 +4119,7 @@ describe('the New Game dialog', () => {
    * questions through `settings-form.ts`, which addresses its radio groups by NAME through the
    * DOCUMENT — so the stub keeps one set of groups and answers for both the document and the dialog.
    */
-  const load = async (search: string) => {
+  const load = async (search: string, stored: Record<string, string> = {}) => {
     execFileSync('node', ['scripts/build-web.ts'], { cwd: root, stdio: 'pipe' });
 
     const served = new Set(
@@ -3901,10 +4147,13 @@ describe('the New Game dialog', () => {
       }));
     // ONE set per page, not one per element: the block is addressed through the document, and a stub
     // that handed each element its own copy would let a broken selector still pass.
+    // ONE set, matching the markup's own `checked` defaults. There were two until 2026-08-30, one
+    // per screen, because the in-game dialog was a second copy of this block — it is deleted, and
+    // the setup screen it folded into is the only thing on the page driving these radios now.
     const groups: Record<string, Radio[]> = {
-      'ng-hand': group(['threeRandom', 'sixRandom', 'threeTrackThreeOther'], 'sixRandom'),
-      'ng-extra': group(['divisionPointsOnly', 'ownOffice', 'anyOffice'], 'anyOffice'),
-      'ng-type': group(['solitaire', 'coop', 'competitive', 'cutthroat', 'custom'], 'coop'),
+      'ss-hand': group(['threeRandom', 'sixRandom', 'threeTrackThreeOther'], 'sixRandom'),
+      'ss-extra': group(['divisionPointsOnly', 'ownOffice', 'anyOffice'], 'anyOffice'),
+      'ss-type': group(['solitaire', 'coop', 'competitive', 'cutthroat', 'custom'], 'solitaire'),
     };
     const matching = (sel: string): Radio[] => {
       const name = /name="([^"]+)"/.exec(sel)?.[1] ?? '';
@@ -3957,7 +4206,7 @@ describe('the New Game dialog', () => {
       origin: 'http://box.local',
       pathname: '/play.html',
     };
-    const store = new Map<string, string>();
+    const store = new Map<string, string>(Object.entries(stored));
     g['localStorage'] = {
       getItem: (k: string) => store.get(k) ?? null,
       setItem: (k: string, v: string) => void store.set(k, v),
@@ -3973,15 +4222,15 @@ describe('the New Game dialog', () => {
 
   /** What every field of the block reads, so a test can assert the whole form at once. */
   const readForm = (els: Map<string, Record<string, unknown>>, groups: Record<string, { value: string; checked: boolean }[]>) => ({
-    hand: groups['ng-hand']!.find((r) => r.checked)?.value,
-    extra: groups['ng-extra']!.find((r) => r.checked)?.value,
-    type: groups['ng-type']!.find((r) => r.checked)?.value,
-    passenger: els.get('ng-passenger')!['value'],
-    freight: els.get('ng-freight')!['value'],
-    transit: els.get('ng-transit')!['value'],
-    days: els.get('ng-days')!['value'],
-    minrev: els.get('ng-minrev')!['value'],
-    tossloco: els.get('ng-tossloco')!['checked'],
+    hand: groups['ss-hand']!.find((r) => r.checked)?.value,
+    extra: groups['ss-extra']!.find((r) => r.checked)?.value,
+    type: groups['ss-type']!.find((r) => r.checked)?.value,
+    passenger: els.get('ss-passenger')!['value'],
+    freight: els.get('ss-freight')!['value'],
+    transit: els.get('ss-transit')!['value'],
+    days: els.get('ss-days')!['value'],
+    minrev: els.get('ss-minrev')!['value'],
+    tossloco: els.get('ss-tossloco')!['checked'],
   });
 
   it('opens on the rules in play, so a second game can be dealt to compare with the first', async () => {
@@ -3990,9 +4239,9 @@ describe('the New Game dialog', () => {
     const { els, groups } = await load('?seed=430&hand=sixRandom&passenger=2&freight=3&transit=4');
     (els.get('newgame')!['onclick'] as () => void)();
 
-    const dlg = els.get('newgamedlg')!;
-    assert.equal(dlg['open'], true, 'the New game button did not open the dialog');
-    assert.equal(els.get('ng-seed')!['value'], '', 'the seed box kept the last game’s seed');
+    assert.equal(els.get('solitairesetup')!['hidden'], false, 'the New game button did not open the setup screen');
+    assert.equal(els.get('gameui')!['hidden'], true, 'the board is still showing over the setup screen');
+    assert.equal(els.get('ss-seed')!['value'], '', 'the seed box kept the last game’s seed');
     const form = readForm(els, groups);
     assert.equal(form.passenger, '2');
     assert.equal(form.freight, '3');
@@ -4026,39 +4275,43 @@ describe('the New Game dialog', () => {
     const tuned = await load('?seed=430&transit=4');
     (tuned.els.get('newgame')!['onclick'] as () => void)();
     assert.equal(readForm(tuned.els, tuned.groups).type, 'custom', 'a game paying for transits still read as Solitaire');
+    // As in the lobby: the row that differs carries the difference, not a sentence under the radios.
     assert.match(
-      String(tuned.els.get('ng-type-note')!['textContent']),
-      /1 setting differs from Solitaire/,
-      'the note did not say what differs',
+      String(tuned.els.get('ss-transit-hint')!['textContent']),
+      /Solitaire default/,
+      'the changed row does not say what it was changed from',
     );
   });
 
   it('offers the multiplayer types, disabled — one list across both screens, dealt from one of them', async () => {
     const { els, groups } = await load('?seed=430');
     (els.get('newgame')!['onclick'] as () => void)();
-    const disabled = groups['ng-type']!
+    const disabled = groups['ss-type']!
       .filter((r) => (r as { disabled: boolean }).disabled)
       .map((r) => r.value);
     assert.deepEqual(disabled, ['coop', 'competitive', 'cutthroat'], 'the wrong game types are dealable here');
     // Deal is never disabled here any more: the only types this screen can SELECT are the two it can
     // deal, so a disabled button would be answering a question the radios no longer ask.
-    assert.ok(
-      readFileSync(join(dist, 'play.html'), 'utf8').includes('id="ng-multiplayer-note"'),
-      'nothing on the dialog says where a multiplayer game comes from',
-    );
+    // No reason is printed beside the dimmed rows any more (Jesse, 2026-08-30 — "grayed out with no
+    // additional explanation"). The heading is what says which game the screen deals, once, and it
+    // is the same shape on the lobby, which is what lets one list serve both.
+    const page = readFileSync(join(dist, 'play.html'), 'utf8');
+    assert.match(page, /Game type \(solitaire\)/, 'the setup screen does not name the game it deals');
+    assert.match(page, /Game type \(multi-player\)/, 'the lobby does not name the game it deals');
+    assert.ok(!page.includes('lb-why'), 'the per-row reason is back, on the page or in its stylesheet');
   });
 
   it('clicking a game type resets every rule below to it, and leaves the parameters alone', async () => {
     const { els, groups } = await load('?seed=430&transit=4');
     (els.get('newgame')!['onclick'] as () => void)();
-    els.get('ng-days')!['value'] = '8';
-    (els.get('ng-days')!['oninput'] as () => void)();
+    els.get('ss-days')!['value'] = '8';
+    (els.get('ss-days')!['oninput'] as () => void)();
     // 3 × 1 player × 8 days: the floor follows the length, and changing the length is not a rule
     // change, so this is still Solitaire rather than Custom.
     assert.equal(readForm(els, groups).minrev, '24', 'the Revenue floor did not follow the Day count');
 
-    const solitaire = groups['ng-type']!.find((r) => r.value === 'solitaire')!;
-    for (const r of groups['ng-type']!) r.checked = r === solitaire;
+    const solitaire = groups['ss-type']!.find((r) => r.value === 'solitaire')!;
+    for (const r of groups['ss-type']!) r.checked = r === solitaire;
     (solitaire as { onchange: (() => void) | null }).onchange!();
 
     const form = readForm(els, groups);
@@ -4071,19 +4324,17 @@ describe('the New Game dialog', () => {
     const { els, groups, nav: n } = await load('?seed=430');
     (els.get('newgame')!['onclick'] as () => void)();
 
-    const dlg = els.get('newgamedlg')!;
-    els.get('ng-seed')!['value'] = '99';
-    for (const r of groups['ng-hand']!) r.checked = r.value === 'threeTrackThreeOther';
-    els.get('ng-passenger')!['value'] = '5';
-    els.get('ng-freight')!['value'] = '0';
-    els.get('ng-transit')!['value'] = '2';
-    els.get('ng-toolbox')!['checked'] = true;
-    dlg['returnValue'] = 'deal';
-    (dlg['close'] as () => void)();
+    els.get('ss-seed')!['value'] = '99';
+    for (const r of groups['ss-hand']!) r.checked = r.value === 'threeTrackThreeOther';
+    els.get('ss-passenger')!['value'] = '5';
+    els.get('ss-freight')!['value'] = '0';
+    els.get('ss-transit')!['value'] = '2';
+    els.get('ss-toolbox')!['checked'] = true;
+    (els.get('ss-deal')!['onclick'] as () => void)();
 
     assert.equal(
       n.search,
-      '?seed=99&hand=threeTrackThreeOther&extra=anyOffice&passenger=5&freight=0&transit=2' +
+      '?seed=99&hand=threeTrackThreeOther&extra=ownOffice&passenger=5&freight=0&transit=2' +
         '&days=5&minrev=15&colday=3&coltotal=5&tool=1',
     );
   });
@@ -4091,38 +4342,62 @@ describe('the New Game dialog', () => {
   it('a switched-off victory condition deals as 0, which is what the engine calls off', async () => {
     const { els, nav: n } = await load('?seed=430');
     (els.get('newgame')!['onclick'] as () => void)();
-    const dlg = els.get('newgamedlg')!;
-    els.get('ng-coltotal-on')!['checked'] = false;
-    (els.get('ng-coltotal-on')!['onchange'] as () => void)();
-    dlg['returnValue'] = 'deal';
-    (dlg['close'] as () => void)();
+    els.get('ss-coltotal-on')!['checked'] = false;
+    (els.get('ss-coltotal-on')!['onchange'] as () => void)();
+    (els.get('ss-deal')!['onclick'] as () => void)();
     assert.match(n.search, /coltotal=0/, 'unticking the total-collision condition did not switch it off');
   });
 
-  it('deals nothing on cancel, and nothing on Esc', async () => {
-    // Esc closes a <dialog> with an empty returnValue and fires no submit at all, so "not deal" has
-    // to be the test rather than "cancel" — the two arrive identically.
-    for (const returnValue of ['cancel', '']) {
-      const { els, nav: n } = await load('?seed=430');
-      (els.get('newgame')!['onclick'] as () => void)();
-      const dlg = els.get('newgamedlg')!;
-      dlg['returnValue'] = returnValue;
-      (dlg['close'] as () => void)();
-      assert.equal(n.search, '?seed=430', `closing with "${returnValue}" navigated`);
-      assert.equal(n.reloads, 0, `closing with "${returnValue}" reloaded`);
-    }
+  it('says the game has RESUMED, not begun, when you come back to one', async () => {
+    /**
+     * REPORTED BY JESSE 2026-08-30: "when you are continuing the saved game out of that screen, do
+     * not post a message that says 'The game has begun.' … it needs to say 'The game has resumed.'"
+     *
+     * A restored game draws exactly like a dealt one — mid-Day, mid-phase, log already deep — so
+     * nothing distinguished the two. Checked in both directions here: the wording itself, and that
+     * the multiplayer line which DOES say "begun" cannot be said to somebody rejoining.
+     */
+    const save = JSON.stringify({ seed: 12345, history: [] });
+    const { els } = await load('', { 'station-master.save.v1': save });
+    assert.match(String(els.get('announce')!['textContent']), /has resumed/, 'a restored game said nothing');
+    assert.doesNotMatch(String(els.get('announce')!['textContent']), /has begun/, 'a restored game claimed to be new');
+
+    // A freshly dealt game must NOT claim to be resumed — the announcement has to mean something.
+    const fresh = await load('?hand=sixRandom');
+    assert.doesNotMatch(String(fresh.els.get('announce')?.['textContent'] ?? ''), /resumed/,
+      'a brand new game announced itself as a resume');
+
+    // And the multiplayer first-frame line is conditional now rather than always "begun".
+    const src = readFileSync(join(root, 'src/web/main.ts'), 'utf8');
+    assert.match(src, /rejoining \? 'resumed' : 'begun'/, 'rejoining a table still says the game has begun');
+  });
+
+  it('backing out deals nothing and puts the same game back on screen', async () => {
+    /**
+     * WHAT CANCEL USED TO BE. The dialog had a Cancel button and an Esc key, and this pinned that
+     * neither dealt. The screen that replaced it has neither — it has "Continue Existing Saved
+     * Game", which has to do the same job and one more besides: the game was never navigated away
+     * from, so going back is showing the board again rather than reloading and replaying it.
+     */
+    const { els, nav: n } = await load('?seed=430');
+    (els.get('newgame')!['onclick'] as () => void)();
+    assert.equal(els.get('ss-resume')!['hidden'], false, 'mid-game there is no way back to the game');
+
+    (els.get('ss-resume')!['onclick'] as () => void)();
+    assert.equal(els.get('gameui')!['hidden'], false, 'backing out did not return to the board');
+    assert.equal(els.get('solitairesetup')!['hidden'], true, 'the setup screen stayed up');
+    assert.equal(n.search, '?seed=430', 'backing out navigated');
+    assert.equal(n.reloads, 0, 'backing out reloaded, losing the game in memory');
   });
 
   it('reloads when the answers are the URL the page already has, so a re-deal is not a no-op', async () => {
     // Dealing a random seed, disliking it and dealing again at the same settings produces the same
     // search string — and assigning `location.search` the value it already holds does nothing.
     const url =
-      '?hand=sixRandom&extra=anyOffice&passenger=1&freight=1&transit=0&days=5&minrev=15&colday=3&coltotal=5';
+      '?hand=sixRandom&extra=ownOffice&passenger=1&freight=1&transit=0&days=5&minrev=15&colday=3&coltotal=5';
     const { els, nav: n } = await load(url);
     (els.get('newgame')!['onclick'] as () => void)();
-    const dlg = els.get('newgamedlg')!;
-    dlg['returnValue'] = 'deal';
-    (dlg['close'] as () => void)();
+    (els.get('ss-deal')!['onclick'] as () => void)();
 
     assert.equal(n.search, url, 'the URL should be unchanged — that is the whole case');
     assert.equal(n.reloads, 1, 'a re-deal at the same settings did nothing at all');
@@ -4142,11 +4417,105 @@ describe('the New Game dialog', () => {
     assert.match(String(els.get('gametype')!['title']), /Days: 5/, 'the tooltip does not carry the victory conditions');
   });
 
+  it('asks before the first deal — a bare visit shows the setup screen, not a dealt game', async () => {
+    // Jesse, 2026-08-29: "let the user choose their options like the start of a multiplayer game";
+    // "asking first is the only path". A saved game, an explicit seed, or a URL a Deal already wrote
+    // (checked via `hand`, below) all skip this screen — nothing else does.
+    const { els } = await load('');
+    assert.equal(els.get('solitairesetup')!['hidden'], false, 'the setup screen stayed hidden');
+    assert.equal(els.get('gameui')!['hidden'], true, 'a game was dealt before anyone chose anything');
+  });
+
+  it('the solitaire door reaches the setup screen even when a solitaire game is saved', async () => {
+    /**
+     * REPORTED THREE TIMES BY JESSE (2026-08-29, twice, and 2026-08-30). v0.7.5 skipped the setup
+     * screen whenever `load()` found a save, reasoned as "a saved game is a game to resume" — but
+     * that means ANY browser that has ever played solitaire can never reach the setup screen from
+     * the door again, which is the whole feature. The private window that appeared to prove the
+     * caching fix had simply never played, so its `localStorage` was empty.
+     *
+     * The door is an explicit request to set a game up. A BARE reload still resumes (below).
+     */
+    const save = JSON.stringify({ seed: 12345, history: [] });
+    const { els } = await load('?solitaire', { 'station-master.save.v1': save });
+    assert.equal(els.get('solitairesetup')!['hidden'], false, 'a saved game swallowed the door');
+    assert.equal(els.get('gameui')!['hidden'], true, 'the saved game was resumed instead of asking');
+  });
+
+  it('offers a way back to the saved game, since dealing from the door destroys it', async () => {
+    // Deal calls `clearSave()`. The door is reached by clicking "Play solitaire", which nobody reads
+    // as "discard what I was playing" — so the save has to be one button away, and the cost of Deal
+    // has to be stated. Resuming navigates to the bare URL and lets `start()` do it.
+    const save = JSON.stringify({ seed: 12345, history: [] });
+    const { els, nav } = await load('?solitaire', { 'station-master.save.v1': save });
+    assert.equal(els.get('ss-resume')!['hidden'], false, 'no way back to the game in progress');
+    assert.equal(els.get('ss-saved-note')!['hidden'], false, "Deal's cost to the save is not stated");
+    (els.get('ss-resume')!['onclick'] as () => void)();
+    assert.equal(nav.search, '', 'resuming did not go back to the plain resume path');
+  });
+
+  it('hides the resume button when there is no saved game to go back to', async () => {
+    const { els } = await load('?solitaire');
+    assert.equal(els.get('ss-resume')!['hidden'], true, 'a resume button with nothing to resume');
+    assert.equal(els.get('ss-saved-note')!['hidden'], true, 'warns about replacing a save that does not exist');
+  });
+
+  it('a bare reload still resumes a saved solitaire game rather than asking again', async () => {
+    // The other half: `?solitaire` is what changed, not resuming itself. Reopening the tab must not
+    // put a question in front of somebody who just wants their game back (D11's zero-friction case).
+    const save = JSON.stringify({ seed: 12345, history: [] });
+    const { els } = await load('', { 'station-master.save.v1': save });
+    assert.equal(els.get('gameui')!['hidden'], false, 'a bare reload did not resume the saved game');
+    assert.equal(els.get('solitairesetup')!['hidden'], true, 'the setup screen interrupted a resume');
+  });
+
+  it('the solitaire door reaches solitaire even when this browser remembers a multiplayer game', async () => {
+    // Found 2026-08-29 verifying v0.7.5 on phoenix.local: a browser with ANY remembered multiplayer
+    // seat (`station-master.remote.v1`) could never reach solitaire's setup screen at all — a bare
+    // `./play.html` load and the splash's "Play solitaire" door were indistinguishable from a reload
+    // mid-multiplayer-game, and `start()` checked the remembered session first. The door now marks
+    // its intent with `?solitaire`, the same way `?lobby` already does for the door on the other side.
+    const remembered = JSON.stringify({
+      games: { g1: { token: 't1', gameId: 'g1', gameCode: 'FREIGHT-3230', seat: 0, stage: 'game' } },
+      last: 'g1',
+    });
+    const { els } = await load('?solitaire', { 'station-master.remote.v1': remembered });
+    assert.equal(els.get('solitairesetup')!['hidden'], false, 'the door lost to the remembered game');
+    assert.equal(els.get('gameui')!['hidden'], true, 'the remembered multiplayer game was resumed instead');
+  });
+
+  it('a bare reload still resumes a remembered multiplayer game, unlike the solitaire door', async () => {
+    // The other half of the fix above: `?solitaire` must be what changed, not remembered-session
+    // resume itself, which is the correct behaviour for an actual reload mid-game (D11/D14).
+    const remembered = JSON.stringify({
+      games: { g1: { token: 't1', gameId: 'g1', gameCode: 'FREIGHT-3230', seat: 0, stage: 'game' } },
+      last: 'g1',
+    });
+    const { els } = await load('', { 'station-master.remote.v1': remembered });
+    assert.equal(els.get('gameui')!['hidden'], false, 'a bare reload did not resume the remembered game');
+    assert.equal(els.get('solitairesetup')!['hidden'], true, 'the setup screen wrongly took priority');
+  });
+
+  it("the setup screen's own Deal does not bounce into a remembered multiplayer game", async () => {
+    // The same bug one level deeper: `commitNewGame` writes `?hand=...`, not `?solitaire=...`, so the
+    // very next load after pressing Deal has to be recognised as a solitaire navigation too — checked
+    // via `hand`, the same signal `start()` already uses to skip the setup screen a second time.
+    const remembered = JSON.stringify({
+      games: { g1: { token: 't1', gameId: 'g1', gameCode: 'FREIGHT-3230', seat: 0, stage: 'game' } },
+      last: 'g1',
+    });
+    const { els } = await load('?hand=sixRandom', { 'station-master.remote.v1': remembered });
+    assert.equal(els.get('gameui')!['hidden'], false, "the Deal button's own URL was not honoured");
+    assert.equal(els.get('solitairesetup')!['hidden'], true, 'the setup screen re-asked its own answer');
+  });
+
   it('deals six cards by default now, matching what the lobby calls Solitaire', async () => {
     // Jesse, 2026-08-23: every game type opens with six. `SOLO_CONFIG` — the ENGINE's fallback, which
-    // every sim measurement is taken against — deliberately did not move; this is the page's deal.
-    const { els } = await load('');
-    assert.match(String(els.get('houserules')!['textContent']), /6 cards/);
+    // every sim measurement is taken against — deliberately did not move; this is what the setup
+    // screen deals when nothing on it is touched, the same way the dialog always has.
+    const { els, nav } = await load('');
+    (els.get('ss-deal')!['onclick'] as () => void)();
+    assert.match(nav.search, /hand=sixRandom/, "the setup screen's own default was not six cards");
   });
 
   it('ignores a seed the browser cannot parse rather than refusing to deal', async () => {
@@ -4154,14 +4523,12 @@ describe('the New Game dialog', () => {
     // optional box is not worth writing.
     const { els, nav: n } = await load('?seed=430');
     (els.get('newgame')!['onclick'] as () => void)();
-    const dlg = els.get('newgamedlg')!;
-    els.get('ng-seed')!['value'] = 'not a number';
-    dlg['returnValue'] = 'deal';
-    (dlg['close'] as () => void)();
+    els.get('ss-seed')!['value'] = 'not a number';
+    (els.get('ss-deal')!['onclick'] as () => void)();
 
     assert.equal(
       n.search,
-      '?hand=sixRandom&extra=anyOffice&passenger=1&freight=1&transit=0&days=5&minrev=15&colday=3&coltotal=5',
+      '?hand=sixRandom&extra=ownOffice&passenger=1&freight=1&transit=0&days=5&minrev=15&colday=3&coltotal=5',
       'a bad seed was carried into the URL',
     );
   });