Ask about the thumb, not the window width

There is a careful touch-target pass in this stylesheet -- 44px fields, 44px delete
buttons, a padded hit area for the rulings link. Every rule in it sits behind
`max-width: 700px`.

A phone held sideways is 844px wide.

So none of that work applied to the device it was written for, half the time. Measured
on an emulated touch device with the sidebar open -- twelve controls under 44px at
844x390 and again at 915x412:

  the rulings link            91x14
  the dedicated-dealer box    13x13
  chip-colour delete          28px wide
  blinds-direction select     32px tall
  five panel headers          38px

The rules are restated against `pointer: coarse`. The width-scoped block keeps its
layout changes -- one column, stacked dialogs -- because those really are about how
much room there is. This is only about what a thumb can hit, so it asks about the
thumb.

  844x390 touch   12 controls under 44px -> 1
  915x412 touch   12 -> 1
  390x844 touch    6 -> 1
  360x640 touch    6 -> 1

The one left is the bare 24x24 checkbox, whose row is 44px tall -- the row is the
target, which is the existing design and the only thing that works for a checkbox.

Two things were too small on every device, not just touch, and are fixed outright.
The panel headers -- the sidebar's entire navigation -- were 38px everywhere, phones
included; the pass below never reached them. And the rulings link was 91x14, which
misses the 24px minimum with a mouse as well as with a thumb. Under 24px anywhere is
now zero, at every size measured.

Two traps worth recording. A <summary> only draws its disclosure triangle while it is
display: list-item, so the headers get there on padding rather than a flex box. And a
<summary> is content-box, so `min-height: 44px` was *added* to the padding and gave
71px -- the same box-model trap as the header buttons one PR ago, in a different tag.
Both are stated in the CSS where the next person will hit them.

214 tests still passing; layout audit and contrast clean at six sizes across touch and
pointer contexts; no console errors; panels still toggle and the triangle is still
there.

Co-Authored-By: Claude Opus 5 <[email protected]>
Claude-Session: https://claude.ai/code/session_015bnp7xKUPFFSpwkYNFaPgb

JoshHeaps.Net/wwwroot/css/casino/holdem.css

@@ -212,6 +212,20 @@ button,
 }
 
 .holdemBtn {
+    /* Stated, because this class is worn by both <button> and <a> and the browser
+       does not give them the same box. Chromium's own stylesheet makes a button
+       border-box and leaves a link content-box, so `min-height: 46px` on the header
+       row came out as 46px on the buttons and 64px on the two links — and the row
+       stretches to its tallest item, so every button became 64px too. A 46px touch
+       target was quietly costing 64px of a phone screen, twice over, because the
+       header wraps to two rows. */
+    box-sizing: border-box;
+    /* And the touch floor is stated here rather than only under a width breakpoint.
+       It used to arrive by accident — the content-box links were 64px tall and the
+       row stretched everything to match — so once the boxes agree, a landscape phone
+       (wider than the phone breakpoint, still a thumb) would have been left with 42px
+       buttons. 44px is the floor everywhere; the narrow layout still asks for 46. */
+    min-height: 44px;
     background: linear-gradient(180deg, #2e2e2e, #232323);
     color: #eee;
     border: 1px solid #3a3a3a;
@@ -695,6 +709,23 @@ button,
     display: inline-block;
 }
 
+/* Seats on the bottom of the ring wear their badge underneath, because that's the
+   edge the pot isn't on. Bets sit between a seat and the pot, so on the bottom arc
+   the chip comes down onto the seat's top edge — which is exactly where this badge
+   was, and the chip is a layer above it. See render.js for the measurement. */
+.seat.seatBelowCentre .blindBadge {
+    top: auto;
+    bottom: -9px;
+}
+
+/* And the badge outranks the chips regardless, so a layout this doesn't anticipate
+   can't hide it again. Which blind a seat is posting is the one thing on that puck
+   nothing else on the felt repeats — a bet chip's amount is on the seat's own total
+   line as well. Measured across 12 table sizes: no chip loses its amount to this. */
+.blindBadge {
+    z-index: 4;
+}
+
 /* Both badges failed WCAG AA against their own gradients — measured at the worst
    stop, the small blind was 3.52:1 and the big blind 2.15:1, in 9px bold text a
    dealer is meant to read at a glance across a table.
@@ -982,9 +1013,24 @@ button,
     text-align: center;
 }
 
+/* A bare checkbox ships at 13x13, which misses the 24px minimum with a mouse as well
+   as with a thumb. The row around it is the real target on touch, but the box itself
+   should be findable on any device. */
+#dedicatedDealerToggle {
+    width: 24px;
+    height: 24px;
+}
+
 .linkBtn {
     background: none;
     border: none;
+    /* An underlined word was a 91x14 target, which misses the 24px minimum with a
+       mouse as well as with a thumb. The coarse-pointer block takes it to 44; this is
+       the floor everywhere else, and it changes nothing about how it looks. */
+    display: inline-flex;
+    align-items: center;
+    box-sizing: border-box;
+    min-height: 24px;
     padding: 0;
     color: #d8b46a;
     font-family: inherit;
@@ -1439,7 +1485,16 @@ button,
 
 .sidebarPanel summary {
     cursor: pointer;
-    padding: 0.7rem 0;
+    /* These headers are the sidebar's whole navigation and were 38px tall on every
+       device, phones included — the touch pass further down never reached them.
+       Padding rather than a flex box with min-height, because a <summary> only draws
+       its disclosure triangle while it is display: list-item. */
+    padding: 0.85rem 0;
+    /* A <summary> is content-box, so min-height would have been *added* to the padding
+       — measured at 71px before this line went in. Same trap the header buttons were
+       in, one file over. */
+    box-sizing: border-box;
+    min-height: 44px;
     font-size: 0.9rem;
     color: #eee;
     letter-spacing: 0.03em;
@@ -3154,3 +3209,66 @@ textarea:focus-visible {
     font-size: 0.72rem;
     color: #9a9a9a;
 }
+
+/* ---------- Touch targets on a wide touch screen ----------
+
+   Every touch-target rule in this file sits behind `max-width: 700px`, and width was
+   never the question. A phone held sideways is 844px wide, so none of that work
+   applied to it. Measured on an emulated touch device at 844x390, with the sidebar
+   panels open: twelve controls under 44px, including the rulings link at 91x14 and
+   the dedicated-dealer checkbox at 13x13.
+
+   Restated here against the pointer instead of the viewport. The width-scoped block
+   keeps its layout changes — one column, stacked dialogs — because those really are
+   about how much room there is. This is only about what a thumb can hit, so it asks
+   about the thumb. */
+@media (pointer: coarse) {
+    .panelBody input:not([type="checkbox"]):not([type="color"]),
+    .subPanelBody input:not([type="checkbox"]):not([type="color"]),
+    .panelBody select {
+        font-size: 16px;
+        min-height: 44px;
+        padding: 0.4rem 0.6rem;
+    }
+
+    .panelBody .holdemBtnSmall,
+    .handHistoryActions .holdemBtn,
+    .toastAction {
+        min-height: 44px;
+    }
+
+    /* Destructive and tiny is the worst combination: these delete a chip colour or a
+       custom field, and are 28px wide. */
+    .removeChipColorBtn,
+    .removeFieldBtn {
+        min-width: 44px;
+        min-height: 44px;
+    }
+
+    /* flex-shrink would take a stated width straight back off again. */
+    .chipColorInput {
+        flex: none;
+        width: 44px;
+        height: 44px;
+    }
+
+    /* A bare checkbox can't grow into a target, but the row it sits in can — the
+       label is what a tap actually lands on. */
+    .panelBody .fieldRow {
+        min-height: 44px;
+    }
+
+    #dedicatedDealerToggle {
+        width: 24px;
+        height: 24px;
+    }
+
+    /* An underlined word is a 14px-tall target; give it room without making it look
+       like a button. */
+    .linkBtn {
+        display: inline-flex;
+        align-items: center;
+        min-height: 44px;
+        padding: 0 0.4rem;
+    }
+}

JoshHeaps.Net/wwwroot/js/casino/holdem/betting.js

@@ -90,6 +90,28 @@ const Betting = {
         };
     },
 
+    // Does handing this seat chips also put them back in the game?
+    //
+    // At a real table those are one event, and here they were two: a busted player got
+    // their chips and stayed sat out, silently skipped hand after hand until somebody
+    // noticed and found "Deal in" in the seat menu. Nothing said so — the toast just
+    // read "Ben added $100".
+    //
+    // Not for a seat that was sat out while *holding* chips. That's a dealer who sat
+    // somebody out on purpose and is topping them up, and dealing them back in would be
+    // the tool guessing at something it wasn't told. Only a seat with no stack of its
+    // own: busted at zero, or a chair being bought into for the first time.
+    //
+    // Deliberately reads player.stack rather than stackOf(), which substitutes the
+    // table's starting stack for a seat that has none — the question here is what this
+    // seat was actually recorded as holding.
+    buyInDealsThemIn(player) {
+        if (!player || player.inHand) return false;
+        const own = player.stack;
+        const recorded = own !== null && own !== undefined && String(own).trim() !== "";
+        return !recorded || Number(own) <= this.EPSILON_CHIPS;
+    },
+
     // "bet" or "raise" — what putting more in is actually called right now.
     //
     // A wager into a pot nobody has wagered into is a bet; everything after that is a

JoshHeaps.Net/wwwroot/js/casino/holdem/main.js

@@ -140,11 +140,27 @@ document.addEventListener("DOMContentLoaded", () => {
         editSeat: index => Render.toggleSeatEditor(index),
 
         addChips: index => AddChips.open(state, index, (seatIndex, amount) => {
-            seatAction(seatIndex, `${nameOf(seatIndex)} added ${Money.format(amount)}.`, player => {
+            // Buying chips puts you back in the game — see Betting.buyInDealsThemIn.
+            const dealBackIn = Betting.buyInDealsThemIn(state.seats[seatIndex]);
+            const handRunning = !Betting.onlyBlindsIn(state) || Betting.potSoFar(state) > 0;
+
+            const message = `${nameOf(seatIndex)} added ${Money.format(amount)}` +
+                (dealBackIn
+                    ? handRunning ? " — in from the next hand." : " — dealt in."
+                    : ".");
+
+            seatAction(seatIndex, message, p => {
                 // Chips in front and cash in are different questions — the stack moves
                 // every hand, what they've handed over only moves when they buy.
-                player.stack = Money.round(Betting.stackOf(state, player) + amount);
-                player.buyIns = Money.round((Number(player.buyIns) || 0) + amount);
+                p.stack = Money.round(Betting.stackOf(state, p) + amount);
+                p.buyIns = Money.round((Number(p.buyIns) || 0) + amount);
+
+                if (dealBackIn) {
+                    p.inHand = true;
+                    // Same rule "Deal in" follows: you can't buy into a pot you haven't
+                    // been paying for, and `folded` is exactly that state for one hand.
+                    p.folded = handRunning;
+                }
             });
         }),
     };

JoshHeaps.Net/wwwroot/js/casino/holdem/render.js

@@ -171,6 +171,16 @@ const Render = {
             node.style.top = `${pos.top}%`;
             node.dataset.seatIndex = String(index);
 
+            // Which side of the seat the pot is on, which is the side its bet chip
+            // arrives from — the chip sits on the same spoke, closer in. For a seat on
+            // the top of the ring that's from below; on the bottom, from above.
+            //
+            // The blind badge has to be on the other side or the chip lands on top of
+            // it. Measured on a 390px phone: the big blind's own $2 chip covered 196 of
+            // the badge's 288 square pixels, and it sits a layer higher, so the marker
+            // telling the dealer who the big blind is was simply gone.
+            node.classList.toggle("seatBelowCentre", pos.top > 50);
+
             node.querySelector(".seatChairNumber").textContent = index + 1;
 
             const dealerChip = node.querySelector(".dealerChip");

tests/holdem/buyIn.test.js

@@ -0,0 +1,76 @@
+// Whether handing a seat chips also puts them back in the game.
+const test = require("node:test");
+const assert = require("node:assert");
+const { Betting, makeTable } = require("./load");
+
+test("a busted player who has been sat out comes back in", () => {
+    const state = makeTable({ seatCount: 6, seated: [0, 1, 2], stack: 100 });
+    state.seats[1].stack = 0;
+    Betting.beginHand(state);           // the deal sits them out
+
+    assert.equal(state.seats[1].inHand, false);
+    assert.equal(Betting.buyInDealsThemIn(state.seats[1]), true);
+});
+
+test("a chair being bought into for the first time comes in", () => {
+    const state = makeTable({ seatCount: 6, seated: [0, 1], stack: 100 });
+    state.seats[4].name = "Eve";
+    state.seats[4].stack = null;
+
+    assert.equal(Betting.buyInDealsThemIn(state.seats[4]), true);
+});
+
+test("a seat sat out while holding chips stays out", () => {
+    // A dealer sat them out on purpose and is topping them up. Dealing them back in
+    // would be the tool guessing at something it wasn't told.
+    const state = makeTable({ seatCount: 6, seated: [0, 1, 2], stack: 100 });
+    state.seats[2].inHand = false;
+
+    assert.equal(Betting.buyInDealsThemIn(state.seats[2]), false);
+});
+
+test("somebody already in the hand is left exactly where they are", () => {
+    const state = makeTable({ seatCount: 6, seated: [0, 1, 2], stack: 100 });
+    assert.equal(Betting.buyInDealsThemIn(state.seats[0]), false);
+
+    // Including a player who is in but out of chips — they're all-in, not sat out.
+    state.seats[2].stack = 0;
+    assert.equal(state.seats[2].inHand, true);
+    assert.equal(Betting.buyInDealsThemIn(state.seats[2]), false);
+});
+
+test("a blank stack field reads as nothing recorded, not as a number", () => {
+    const state = makeTable({ seatCount: 6, seated: [0, 1], stack: 100 });
+    state.seats[3].name = "Dee";
+    state.seats[3].stack = "";
+
+    assert.equal(Betting.buyInDealsThemIn(state.seats[3]), true);
+});
+
+test("it doesn't ask the table what a missing stack would have been", () => {
+    // stackOf substitutes the table's starting stack for a seat that has none, which
+    // would make a seat with nothing recorded look like it was holding chips.
+    const state = makeTable({ seatCount: 6, seated: [0, 1], stack: 100 });
+    state.defaultStack = 500;
+    state.seats[4].name = "Eve";
+    state.seats[4].stack = null;
+
+    assert.equal(Betting.stackOf(state, state.seats[4]), 500);
+    assert.equal(Betting.buyInDealsThemIn(state.seats[4]), true);
+});
+
+test("asking doesn't change anything", () => {
+    const state = makeTable({ seatCount: 6, seated: [0, 1, 2], stack: 100 });
+    state.seats[2].stack = 0;
+    state.seats[2].inHand = false;
+    const before = JSON.stringify(state);
+
+    Betting.buyInDealsThemIn(state.seats[2]);
+
+    assert.equal(JSON.stringify(state), before);
+});
+
+test("nothing to decide about a seat that isn't there", () => {
+    assert.equal(Betting.buyInDealsThemIn(null), false);
+    assert.equal(Betting.buyInDealsThemIn(undefined), false);
+});