From ea08c43b71455d24045b761312ad88ed187d9611 Mon Sep 17 00:00:00 2001
From: Claude <noreply@anthropic.com>
Date: Tue, 5 May 2026 21:27:35 +0000
Subject: [PATCH] =?UTF-8?q?fix(nests):=20don't=20teardown=20on=20wrapper?=
 =?UTF-8?q?=20Closed=20=E2=80=94=20let=20orchestrator=20reconnect?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

The cliff-detector recycle path emits Closed → Reconnecting → Connecting
→ Connected from the wrapper. The Closed branch was calling teardown(),
which cancels cliffDetectorJob, announcesJob, and the wrapper itself,
preventing the orchestrator from ever reopening the inner listener.

Result pre-fix (visible in receiver log): cliff-detector fires after 4s
of silence, teardown runs ~500ms later, wrapper never reopens, room
permanently dead.

User-initiated close goes through disconnect() / onCleared() which call
teardown directly; the wrapper's subsequent Closed emission is redundant
for those paths, so no-op'ing it is safe. UI still reflects Closed via
state.toUiState(ui.connection).

https://claude.ai/code/session_01UHN3fnXzWdj8UbSXnxSwwv
---
 .../commons/viewmodels/NestViewModel.kt       | 31 ++++++++++++++++---
 1 file changed, 27 insertions(+), 4 deletions(-)

diff --git a/commons/src/commonMain/kotlin/com/vitorpamplona/amethyst/commons/viewmodels/NestViewModel.kt b/commons/src/commonMain/kotlin/com/vitorpamplona/amethyst/commons/viewmodels/NestViewModel.kt
index 6776346a09..c221bdc20e 100644
--- a/commons/src/commonMain/kotlin/com/vitorpamplona/amethyst/commons/viewmodels/NestViewModel.kt
+++ b/commons/src/commonMain/kotlin/com/vitorpamplona/amethyst/commons/viewmodels/NestViewModel.kt
@@ -883,11 +883,34 @@ class NestViewModel(
                             // wait for a manual reconnect tap.
                         }
 
-                        // Server-initiated Closed: tear down stale
-                        // local state so any later user-driven reconnect
-                        // starts fresh.
                         NestsListenerState.Closed -> {
-                            teardown(targetState = ConnectionUiState.Closed)
+                            // Closed from the wrapper is *almost always*
+                            // transient — the orchestrator emits Closed
+                            // → Reconnecting → Connecting → Connected
+                            // around every cliff-detector recycle and
+                            // every JWT refresh. We must NOT teardown
+                            // here: teardown cancels `cliffDetectorJob`,
+                            // `announcesJob`, etc., AND calls
+                            // `wrapper.close()`, which cancels the
+                            // wrapper's orchestrator before it can
+                            // reopen the inner listener. End result
+                            // pre-fix: the very first cliff-detector
+                            // recycle permanently kills the room
+                            // instead of recovering it (visible in the
+                            // 15:56:25 receiver log: cliff fires →
+                            // teardown fires 521 ms later → wrapper
+                            // never reopens → `cliff-detector EXITED
+                            // closed=false`).
+                            //
+                            // User-initiated close goes through
+                            // `disconnect()` / `onCleared()` which call
+                            // `teardown` directly; the wrapper's
+                            // subsequent Closed emission here is a
+                            // redundant signal — no-op'ing it doesn't
+                            // change anything for those paths. The UI
+                            // still picks up the Closed via
+                            // `state.toUiState(ui.connection)` above
+                            // for visual feedback.
                         }
 
                         else -> { /* no extra side effect */ }