fix(rig): Ardour hears the surface again — two stale identities, one webcam

PLN's report: "no impact of lcxl in ardour... why would i need to remap midi
controls there across lcxl gens?" He is right: he never needs to remap. The
session file has the learned bindings (v2 CC 77-84 + 13-16, channel 1) and the
lcxl3-driver emits exactly that. The path was broken in TWO places, both by
name-vs-identity rot:

1. midi-autoconnect resolved its source by NAME and connected "name:0" — but
   aconnect resolves CLIENT names only, and the driver's virtual port carries
   its name on the PORT line (client = python-rtmidi's 'RtMidiOut Client').
   The connect failed into 2>/dev/null for exactly the source the candidate
   list most prefers, so Midi Through carried NOTHING in DAW mode. Now the
   resolver returns a numeric client:port address (awk over aconnect -i,
   matching client AND port lines). Fourth name-identity bug of the day.

2. Ardour's Generic-MIDI input connection was saved by device identity, and
   that identity was "ALSA;;C922 Pro Stream Webcam" — the webcam had taken
   over the slot the surface once owned. The purest stale-binding specimen the
   rig has produced. tidal-ardour-autoroute now asserts
   Midi-Bridge Midi Through -> ardour:MIDI Control In on its 2 s cadence,
   no-op until Ardour registers the port, permanent once it does.

Midi Through stays the rendezvous ON PURPOSE: v2 raw or v3-translated, both
feed it, which is what keeps the classic LCXL plug-and-play (the BC
constraint). Ardour 9.7 note for the archives: the Control Surfaces page is
now a vendor tree built from the .map files' manufacturer field — the old
"Generic MIDI" lives under vendor "Any" (GenericMIDI.map, manufacturer="Any"),
NOT under "Other", and Novation's "Launch Control XL (generic)" entry is the
factory map that would OVERRIDE the learned bindings — do not enable it.

tools/midi-autoconnect.sh

@@ -74,12 +74,23 @@ disconnect() {  # $1 = source, $2 = dest ; silent when there was nothing to cut
 
 # Echo the resolved source ONCE per change, never every INTERVAL: a line every
 # 2s is not a log, it is a way to make a real change invisible.
+#
+# RESOLVE TO A NUMERIC ADDRESS, not a name. The driver's virtual port is named
+# on the PORT line ('ParVagues LCXL3') while its CLIENT is python-rtmidi's
+# 'RtMidiOut Client' — and `aconnect "name:0"` matches CLIENT names only. So
+# the old name-based connect failed silently (2>/dev/null) for exactly the
+# source we most prefer, and Midi Through carried NOTHING in DAW mode: Ardour
+# heard no fader because this loop was connecting a name that cannot resolve.
+# Fourth name-vs-port-identity bug of 2026-08-29. Identity = client:port id.
 resolve_lcxl() {
-    local names; names="$(aconnect -i 2>/dev/null; aconnect -o 2>/dev/null)"
-    local cand
+    local cand addr
     for cand in "${LCXL_CANDIDATES[@]}"; do
-        if printf '%s' "$names" | grep -qF "$cand"; then
-            printf '%s' "$cand"
+        addr="$(aconnect -i 2>/dev/null | awk -v want="$cand" '
+            /^client /  { cid = $2; sub(":", "", cid); cline = $0 }
+            /^[ \t]/    { if (index(cline, want) || index($0, want)) {
+                              print cid ":" $1; exit } }')"
+        if [ -n "$addr" ]; then
+            printf '%s %s' "$addr" "$cand"
             return 0
         fi
     done
@@ -87,20 +98,21 @@ resolve_lcxl() {
 }
 
 reconcile() {
-    local src
-    if ! src="$(resolve_lcxl)"; then
+    local resolved addr name
+    if ! resolved="$(resolve_lcxl)"; then
         if [ "$LAST_SRC" != "__none__" ]; then
             echo "midi-autoconnect: no surface present (looked for: ${LCXL_CANDIDATES[*]})"
             LAST_SRC="__none__"
         fi
         return 0
     fi
-    if [ "$src" != "$LAST_SRC" ]; then
-        echo "midi-autoconnect: surface = '$src'"
-        LAST_SRC="$src"
+    addr="${resolved%% *}"; name="${resolved#* }"
+    if [ "$name ($addr)" != "$LAST_SRC" ]; then
+        echo "midi-autoconnect: surface = '$name' at $addr"
+        LAST_SRC="$name ($addr)"
     fi
-    connect    "$src:$LCXL_PORT" "$THRU:$THRU_PORT"   # rule 1: controls out
-    disconnect "$THRU:$THRU_PORT" "$src:$LCXL_PORT"   # rule 2: kill the loop
+    connect    "$addr" "$THRU:$THRU_PORT"             # rule 1: controls out
+    disconnect "$THRU:$THRU_PORT" "$addr"             # rule 2: kill the loop
 }
 
 echo "midi-autoconnect: reconciling every ${INTERVAL}s"

tools/tidal-ardour-autoroute.sh

@@ -42,7 +42,22 @@ sc_hw_playback_links() {
         }'
 }
 
+# Ardour's Generic-MIDI control input. Ardour SAVES this connection by device
+# identity, and on 2026-08-29 the saved identity was "ALSA;;C922 Pro Stream
+# Webcam" — the webcam had taken over the slot the surface once owned, so the
+# faders PLN learned (v2 CC 77-84 + 13-16, ch 1, stored in the session file)
+# received nothing. A binding resolved once is a stale binding waiting to
+# happen; this asserts the one true feed on the same 2 s cadence as the orbits.
+# Midi Through is deliberately the source, NOT the surface: Midi Through is the
+# stable rendezvous across controller generations (v2 raw, or v3 translated by
+# lcxl3-driver), which is what keeps the classic LCXL plug-and-play.
+CTRL_IN="ardour:MIDI Control In"
+CTRL_SRC="Midi-Bridge:Midi Through: Port-0 (capture)"
+
 reconcile() {
+    # ADD: the control surface feed (no-op until Ardour registers the port)
+    link "$CTRL_SRC" "$CTRL_IN"
+
     # ADD: orbit pair -> matching Tidal track
     while read -r nn; do
         [ -n "${nn:-}" ] || continue