Merge the indicator polish pass: sign-off note lifted, overclaim corrected, duplicated rationale trimmed

src/core/ui/footer_bar.h

@@ -88,11 +88,10 @@ bool modeSegmentEnabled(bool isActiveSegment, bool transportRunning, bool routab
 // Idle is "nothing moving", never "no work queued" — the shell's enum owns that reasoning.
 enum class FxDrainState { Idle, Online, Offline };
 
-// DRAFT wording, pending sign-off. Short on purpose: these share the count box, whose
-// default 64px holds roughly a dozen glyphs at the panel's smallest font before the
-// shell's kit end-ellipsizes it. The two directions read differently because they ARE
-// different acts — Offline takes the inactive modes' chains out, Online brings the
-// entering mode's back.
+// Short on purpose: these share the count box, whose default 64px holds roughly a
+// dozen glyphs at the panel's smallest font before the shell's kit end-ellipsizes it.
+// The two directions read differently because they ARE different acts — Offline
+// takes the inactive modes' chains out, Online brings the entering mode's back.
 inline constexpr const char* kFxParkingReadout   = "parking FX";
 inline constexpr const char* kFxRestoringReadout = "restoring FX";
 

src/shell/panel/CLAUDE.md

@@ -64,8 +64,11 @@ live in `shell/bank_ops`, a sibling directory, not here.
   reused WDL pieces are the toolkit; reject any path that pulls in a new framework.
 - **The FX-drain indicator asks `fxReclaimActivity()` and decides nothing itself.**
   `panel_input` polls it once per tick into `g_panel.fxDrain`; the footer readout and the
-  `WM_SETCURSOR` busy cursor both read that cached state, so paint and cursor cannot
-  disagree within a tick. The transport gate, the park debounce and the open-undo-block
+  `WM_SETCURSOR` busy cursor both read that cached state, so the two never disagree about
+  the drain's own direction. The cursor additionally suppresses on `dragArmed`/`dragging` (a
+  live card drag outranks it), so mid-drag the footer can still read "parking FX"/"restoring
+  FX" while the cursor shows the drag cue — that is the one case where paint and cursor
+  diverge, and it is deliberate. The transport gate, the park debounce and the open-undo-block
   stand-off are all already folded into that answer, against the reclaim queue's own pinned
   project — do not re-evaluate any of them here, and never drive this off `fxReclaimPending`
   (true throughout the debounce, when only the opposite direction is eligible).

src/shell/panel/panel_render.cpp

@@ -147,9 +147,8 @@ void drawFooter(LICE_IBitmap* bmp, int w, int h) {
         }
     }
 
-    // The one passive readout beside the toggle: a moving FX drain, else the per-mode member
-    // count (footerReadout owns which). Passive text, not a control. The drain draws
-    // text/primary against the count's text/dim so a transient state reads as one.
+    // footerReadout (core/ui) owns which string shows; the drain draws text/primary against
+    // the count's text/dim so the transient state visually reads as one.
     if (!fb.count.empty()) {
         const bool draining = g_panel.fxDrain != FxDrainState::Idle;
         const std::string label = footerReadout(g_panel.fxDrain, activeModeMemberCount());

src/shell/panel/panel_window.cpp

@@ -94,15 +94,10 @@ void handleDropFiles(HDROP hDrop) {
     if (!paths.empty()) ingestDroppedFiles(paths);
 }
 
-// IDC_APPSTARTING — arrow plus hourglass, "busy in the background, still usable" — is the
-// exact reading for the FX reclaimer's drain, and IDC_WAIT is not: nothing here blocks.
-// SWELL defines neither anywhere in vendor/WDL/WDL/swell, so on mac/Linux there is no busy
-// cursor at all and the footer readout carries the state alone. The cursor reaches OUR
-// windows only; REAPER's arrange and mixer answer their own WM_SETCURSOR and the SDK
-// exposes nothing host-wide to ask for. A global SetSystemCursor or a message hook would
-// reach further and is deliberately not done. [verify — DAW] whether REAPER's docker even
-// routes WM_SETCURSOR to the docked child dialog at all — the premise the whole Windows
-// half of this feature rests on.
+// IDC_APPSTARTING ("busy in the background, still usable"), not IDC_WAIT ("blocking") —
+// nothing here blocks. Platform reach and the SetSystemCursor rejection: see this
+// directory's CLAUDE.md. [verify — DAW] whether REAPER's docker even routes WM_SETCURSOR
+// to the docked child dialog at all — the premise the whole Windows half of this rests on.
 #ifdef _WIN32
 HCURSOR busyCursor() { return LoadCursor(nullptr, IDC_APPSTARTING); }
 #else

tests/test_footer_bar.cpp

@@ -265,9 +265,8 @@ static void testActiveSegmentStaysLiveEvenWhenUnroutable() {
 // --- The readout: count, and the drain's pre-emption of it -------------------
 
 static void testIdleReadsTheCountAndNeitherDrainString() {
-    // Idle is the whole no-work-moving family — an empty queue, an item inside its park
-    // debounce, one held by the transport, one held by an open undo block. The readout
-    // cannot tell them apart and must show the count in every one of them.
+    // Idle covers the whole no-work-moving family: empty queue, park debounce, transport
+    // hold, open undo block — indistinguishable to this predicate (see footer_bar.h).
     CHECK(footerReadout(FxDrainState::Idle, 0) == "0 tracks");
     CHECK(footerReadout(FxDrainState::Idle, 1) == "1 track");
     CHECK(footerReadout(FxDrainState::Idle, 2) == "2 tracks");
@@ -294,11 +293,13 @@ static void testTheTwoDirectionsReadDifferently() {
     CHECK(footerReadout(FxDrainState::Online, 3) != footerReadout(FxDrainState::Offline, 3));
 }
 
-static void testDrainReadoutsFitTheCountBox() {
-    // They borrow the count box, so neither may be long enough to make the kit's
-    // end-ellipsis the normal case. Twelve glyphs is the default 64px box's rough capacity
-    // at the panel's smallest font; a longer string wants a wider box, not a silent
-    // truncation.
+static void testDrainReadoutsStayShort() {
+    // NOT a pixel-width bound. Font::Micro is proportional (~10px, DT_END_ELLIPSIS in
+    // shell/panel/draw_kit.cpp) and this pure footer_bar test links no LICE, so it cannot
+    // measure rendered glyph width without pulling shell drawing into core/ui, which the
+    // pure/shell split (core/ui/CLAUDE.md) forbids. This only catches a readout that grew
+    // unreasonably long in character count; a real overflow end-ellipsizes rather than
+    // breaking layout, so fit is confirmed by the DAW pass, not by this assertion.
     CHECK(std::string(kFxParkingReadout).size() <= 12);
     CHECK(std::string(kFxRestoringReadout).size() <= 12);
     CHECK(!std::string(kFxParkingReadout).empty());
@@ -326,7 +327,7 @@ int main() {
     testIdleReadsTheCountAndNeitherDrainString();
     testEachDirectionPreEmptsTheCount();
     testTheTwoDirectionsReadDifferently();
-    testDrainReadoutsFitTheCountBox();
+    testDrainReadoutsStayShort();
 
     if (g_fail == 0) std::printf("footer_bar: all tests passed\n");
     else std::printf("footer_bar: %d CHECK(s) FAILED\n", g_fail);