From d8c1bdba245f69edade45c8f0b65d1a4d3cd9aa3 Mon Sep 17 00:00:00 2001
From: gramps <gramps@llamachile.shop>
Date: Thu, 5 Nov 2026 14:53:06 -0800
Subject: [PATCH] =?UTF-8?q?feat(stream):=20RECONNECTING=20driven=20by=20th?=
 =?UTF-8?q?e=20status=20poll=20=E2=80=94=20no=20invented=20countdown=20(TA?=
 =?UTF-8?q?SK=2032=C2=B72)?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Creator challenge (2026-11-05): "I really don't think that's going to be a constant — it's going to
depend on so many variables, some that are dark to us." Correct, so the measured-grace constant and
the countdown are dropped from the spec. YouTube's cutoff is server-side and depends on ingest health,
prior stream quality, region and load; a pinned number would be a guess wearing a timer, and it would
be wrong exactly when the creator is staring at it.

What ships instead:

- **Amber RECONNECTING** in the top bar with measured facts only: elapsed seconds and the attempt
  number ("0:12 · attempt 2 of 4"). The seconds come from a stopwatch started at the drop.
- **The status poll speeds up while recovering** (30 s → 4 s) and drives the sign; it restores 30 s
  when recovery ends.
- **`chatEndedEvent`** — already in the chat feed we poll and previously skipped — now raises a signal
  that prompts an immediate status check. It never decides anything: `liveBroadcasts.list`
  `lifeCycleStatus` stays authoritative, with `liveStreams.list` health as the push-received signal.
- At `complete`: red OFF AIR (the existing stop path) rather than a pretend-live sign.

Tests: `ChatEndedSignalTests` (the marker is not a chat row, the service raises the signal, and a
complete broadcast ends recovery with the honest message). FAST gate: 543 passed, 0 warnings.
---
 Controls/TopBar.xaml                  |  6 ++
 HANDOFF.md                            |  7 ++-
 Services/YouTubeChatService.cs        |  9 +++
 TASKS.md                              |  2 +-
 TASKS/task-32-stream-resilience.md    | 22 +++++--
 ViewModels/MainViewModel.Recovery.cs  | 87 ++++++++++++++++++++++++++-
 ViewModels/MainViewModel.Streaming.cs | 22 +++++--
 ViewModels/MainViewModel.cs           |  1 +
 ytLive.Tests/ChatEndedSignalTests.cs  | 63 +++++++++++++++++++
 9 files changed, 201 insertions(+), 18 deletions(-)
 create mode 100644 ytLive.Tests/ChatEndedSignalTests.cs

diff --git a/Controls/TopBar.xaml b/Controls/TopBar.xaml
index c393a21..0f267a8 100644
--- a/Controls/TopBar.xaml
+++ b/Controls/TopBar.xaml
@@ -120,6 +120,12 @@
                            Foreground="White" Margin="8,0,0,0" VerticalAlignment="Center"/>
                 <TextBlock Text="{Binding LiveElapsedText}" Foreground="White" FontSize="14"
                            FontFamily="Consolas" VerticalAlignment="Center" Margin="8,0,0,0"/>
+                <!-- TASK 32·2: measured facts only — elapsed seconds and which attempt. No countdown:
+                     YouTube's cutoff is not a constant we can promise. -->
+                <TextBlock x:Name="ReconnectDetail" Text="{Binding ReconnectDetailText}"
+                           Foreground="{StaticResource Brush.Attention}" FontSize="13"
+                           FontFamily="Consolas" VerticalAlignment="Center" Margin="10,0,0,0"
+                           Visibility="{Binding IsReconnecting, Converter={StaticResource BoolToVis}}"/>
                 <Button Content="{Binding EndButtonText}" Style="{StaticResource YtButton}"
                         Background="{StaticResource Brush.Primary}" Margin="14,0,0,0"
                         Command="{Binding EndStreamCommand}"/>
diff --git a/HANDOFF.md b/HANDOFF.md
index a7ce994..f4ffa58 100644
--- a/HANDOFF.md
+++ b/HANDOFF.md
@@ -13,8 +13,11 @@
 
 ## Next action
 
-- **TASK 32·1 shipped** (blip retry, immediate/+1/+2/+3 s, alert-stack episodes, complete-broadcast check).
-  Next: 32·5 (ingest pre-flight),
+- **32·1 + 32·2 shipped**: blip retry (immediate/+1/+2/+3 s), and the RECONNECTING sign driven by the
+  status poll (4 s cadence while recovering) with **measured** elapsed/attempt only — the countdown was
+  dropped after the creator pointed out YouTube's cutoff depends on variables we cannot see. The
+  `chatEndedEvent` in the chat feed now prompts an immediate status check.
+- Next: 32·5 (ingest pre-flight),
   then 32·2/32·3 (countdown + Back on air) which need YOUR measurement take: pull the network on a
   private stream and time how long YouTube holds the broadcast.
 - Open loop: `32·4` is proven by construction only — kill ffmpeg mid-recording and play the file.
diff --git a/Services/YouTubeChatService.cs b/Services/YouTubeChatService.cs
index a960370..18502f3 100644
--- a/Services/YouTubeChatService.cs
+++ b/Services/YouTubeChatService.cs
@@ -25,6 +25,14 @@ public class YouTubeChatService : IDisposable
 
     public event Action<ChatMessage>? MessageReceived;
 
+    internal void NotifyChatEnded() => ChatEnded?.Invoke();
+
+    /// <summary>Raised when YouTube puts a <c>chatEndedEvent</c> in the feed — the chat is over, which
+    /// means the broadcast is. TASK 32·2 uses it as a cheap prompt to check the broadcast status
+    /// immediately instead of waiting for the next health poll. It never decides anything by itself:
+    /// <c>liveBroadcasts.list</c> stays authoritative.</summary>
+    public event Action? ChatEnded;
+
     private static readonly JsonSerializerOptions JsonOpts = new()
     {
         PropertyNameCaseInsensitive = true
@@ -80,6 +88,7 @@ public class YouTubeChatService : IDisposable
 
             var json = await response.Content.ReadAsStringAsync();
             var page = ParsePage(json);
+            if (json.Contains("chatEndedEvent", StringComparison.Ordinal)) NotifyChatEnded();
 
             if (page.NextPageToken != null)
                 _nextPageToken = page.NextPageToken;
diff --git a/TASKS.md b/TASKS.md
index fad605e..e322732 100644
--- a/TASKS.md
+++ b/TASKS.md
@@ -48,7 +48,7 @@ blocker; the smallest slice is 32·4 (fragmented MP4 on the record block).
 | 55·2–3 | ☐ | Disk hygiene: delete `%TEMP%` alert clips on close + sweep at startup; low-disk guard before REC (slice 1 shipped) | [`task-55`](TASKS/task-55-disk-hygiene.md) |
 | 40·A/C/D | ☐ | App Settings: active-camera picker; persist output resolution + surface transition defaults + confirm-before-End; accent consolidation | [`task-40`](TASKS/task-40-app-settings-round.md) |
 | 3·20 | ☐ | RewardEvent persistence: canonical `RewardEvents` table + `superChatEvents.list` 30-day backfill (the session report and journey read it) | [`task-03`](TASKS/task-03-capture-pipeline.md) |
-| 32 | 🔶 | Stream resilience: **fragmented MP4 + blip retry done 2026-11-05**; remaining — measured grace + RECONNECTING countdown → one-click Back on air → ingest pre-flight | [`task-32`](TASKS/task-32-stream-resilience.md) |
+| 32 | 🔶 | Stream resilience: **fragmented MP4 + blip retry + poll-driven RECONNECTING done 2026-11-05** (no countdown — YouTube's cutoff is not a constant); remaining — one-click Back on air → ingest pre-flight | [`task-32`](TASKS/task-32-stream-resilience.md) |
 | 33 | ☐ | Bandwidth auto step-down: pure policy over dropped-frame rate, one-way within a broadcast | [`task-33`](TASKS/task-33-bandwidth-step-down.md) |
 | 34 | ☐ | Scheduled streams (Text-drawer version) | [`task-34`](TASKS/task-34-scheduled-streams.md) |
 | 35 | ☐ | Scene-linked audio | [`task-35`](TASKS/task-35-scene-linked-audio.md) |
diff --git a/TASKS/task-32-stream-resilience.md b/TASKS/task-32-stream-resilience.md
index d5821b9..93b13dd 100644
--- a/TASKS/task-32-stream-resilience.md
+++ b/TASKS/task-32-stream-resilience.md
@@ -23,12 +23,22 @@ the API — creator ruling 2026-09-01): no lying in the copy.
    `FramePump.EncoderFailed` separating a dead push (recoverable) from a fatal pump fault. Each episode
    reports into the left panel's alert stack; a reconnect that lands on a **complete** broadcast is
    reported as such instead of showing a live sign over a dead stream. Tests: `StreamRecoveryTests`.
-2. ☐ **Measured grace + the sign** — while the push is dead, escalate the existing health poll
-   (30s → ~5s). ON-AIR dot goes **amber "RECONNECTING"** with a mm:ss countdown to the observed
-   YouTube cutoff; the grace length is a **measured, pinned constant** — probe it on a real private
-   test stream, cite the observation (spin-guard rule; do NOT invent a number). Poll reports the
-   broadcast `complete` → dot **red "OFF AIR"**, primary button relabels **"● Back on air"** and
-   pulses. Test: fake status provider drives the dot through live → reconnecting → cut-off.
+2. 🔶 **Poll-driven sign — NO invented countdown** (spec revised 2026-11-05, creator challenge:
+   *"I really don't think that's going to be a constant — it's going to depend on so many variables, some
+   that are dark to us"*). The original plan was a measured grace constant and a countdown to YouTube's
+   cutoff. **Dropped**: YouTube's cutoff is server-side and depends on things we cannot observe
+   (ingest node health, prior stream quality, region, load), so any pinned number is a guess wearing a
+   timer. What ships instead:
+   - **Amber RECONNECTING** in the top bar, with **measured** facts only: elapsed time and attempt
+     count ("RECONNECTING · 0:12 · attempt 2 of 4"). No countdown.
+   - **The status poll speeds up** while recovering (30 s → ~4 s) and drives the truth. Signals used,
+     cheapest first: `liveBroadcasts.list` `status.lifeCycleStatus` (authoritative `complete`),
+     `liveStreams.list` `status.streamStatus`/`healthStatus` (is the push being received), and the
+     **`chatEndedEvent`** YouTube already puts in the chat feed we poll — free, and previously skipped.
+   - At `complete`: **red OFF AIR**, stop retrying, and (slice 3) offer **Back on air**.
+   Test: a fake status provider drives the sign through live → reconnecting → off air; a `chatEndedEvent`
+   triggers an immediate status check rather than a decision of its own.
+
 3. ☐ **One-click Back on air** — the relabeled button fires the existing go-live path with saved
    `Broadcast.*` + cached stream, dialog skipped: new broadcast, new VOD (accepted cost, stated
    in-product). Test: seeded saved form + fake stream service → click creates + pushes, no dialog.
diff --git a/ViewModels/MainViewModel.Recovery.cs b/ViewModels/MainViewModel.Recovery.cs
index 22b5b09..9e01884 100644
--- a/ViewModels/MainViewModel.Recovery.cs
+++ b/ViewModels/MainViewModel.Recovery.cs
@@ -19,11 +19,79 @@ public partial class MainViewModel
 {
     private CancellationTokenSource? _recoveryCts;
 
-    /// <summary>True while a push is being relaunched — slice 2 binds the amber sign to this.</summary>
-    public bool IsReconnecting { get; private set; }
+    /// <summary>True while a push is being relaunched — the top bar shows amber RECONNECTING.</summary>
+    public bool IsReconnecting
+    {
+        get => _isReconnecting;
+        private set
+        {
+            if (!SetProperty(ref _isReconnecting, value)) return;
+            RaiseTopBarModes();
+            OnPropertyChanged(nameof(ReconnectDetailText));
+            ApplyReconnectPollCadence();
+        }
+    }
+
+    private bool _isReconnecting;
 
     /// <summary>The attempt currently in flight (0 when idle).</summary>
-    public int ReconnectAttempt { get; private set; }
+    public int ReconnectAttempt
+    {
+        get => _reconnectAttempt;
+        private set
+        {
+            if (!SetProperty(ref _reconnectAttempt, value)) return;
+            OnPropertyChanged(nameof(ReconnectDetailText));
+        }
+    }
+
+    private int _reconnectAttempt;
+
+    /// <summary>Whole seconds since the drop — measured, never predicted.</summary>
+    public int ReconnectElapsedSeconds
+    {
+        get => _reconnectElapsedSeconds;
+        private set
+        {
+            if (!SetProperty(ref _reconnectElapsedSeconds, value)) return;
+            OnPropertyChanged(nameof(ReconnectDetailText));
+        }
+    }
+
+    private int _reconnectElapsedSeconds;
+    private System.Windows.Threading.DispatcherTimer? _reconnectTick;
+
+    /// <summary>While recovering, ask YouTube much more often — the status IS the sign, so it has to
+    /// keep up with a reconnect that can succeed or die inside a few seconds.</summary>
+    private void ApplyReconnectPollCadence()
+    {
+        if (IsReconnecting)
+        {
+            _healthPollTimer.Interval = TimeSpan.FromSeconds(4);
+            _reconnectTick ??= new System.Windows.Threading.DispatcherTimer
+            {
+                Interval = TimeSpan.FromSeconds(1),
+            };
+            _reconnectTick.Tick -= OnReconnectTick;
+            _reconnectTick.Tick += OnReconnectTick;
+            _reconnectTick.Start();
+        }
+        else
+        {
+            _healthPollTimer.Interval = TimeSpan.FromSeconds(30);
+            if (_reconnectTick != null)
+            {
+                _reconnectTick.Stop();
+                _reconnectTick.Tick -= OnReconnectTick;
+            }
+        }
+    }
+
+    private void OnReconnectTick(object? sender, EventArgs e)
+        => ReconnectElapsedSeconds = (int)Math.Round(_reconnectStarted?.Elapsed.TotalSeconds ?? 0);
+
+    // Measured from the drop, so the displayed seconds are facts rather than a projection.
+    private System.Diagnostics.Stopwatch? _reconnectStarted;
 
     private void OnEncoderFailed(object? sender, string message)
     {
@@ -38,8 +106,11 @@ public partial class MainViewModel
 
     private async Task RecoverStreamAsync(string reason, CancellationToken cancellationToken)
     {
+        ReconnectElapsedSeconds = 0;
+        _reconnectStarted = System.Diagnostics.Stopwatch.StartNew();
         IsReconnecting = true;
         ReconnectAttempt = 0;
+        var dropped = _reconnectStarted;
 
         var logFrom = AppLog.SnapshotCurrentLaunch().Count > 0
             ? AppLog.SnapshotCurrentLaunch()[^1].Sequence + 1
@@ -126,6 +197,7 @@ public partial class MainViewModel
         }
         finally
         {
+            _reconnectStarted = null;
             IsReconnecting = false;
             ReconnectAttempt = 0;
             _recoveryCts?.Dispose();
@@ -145,4 +217,13 @@ public partial class MainViewModel
 
     /// <summary>The creator asked to stop: never race a stop against a relaunch.</summary>
     private void CancelStreamRecovery() => _recoveryCts?.Cancel();
+
+    /// <summary>YouTube ended the chat: ask the broadcast status now rather than at the next tick.
+    /// The API decides — this signal only saves up to 30 s of lying to the creator.</summary>
+    private void OnChatEnded()
+    {
+        if (!IsLive) return;
+        AppLog.Write("Chat ended event received — checking the broadcast status now");
+        _ = PollHealthAsync();
+    }
 }
diff --git a/ViewModels/MainViewModel.Streaming.cs b/ViewModels/MainViewModel.Streaming.cs
index 153c757..cf9d14c 100644
--- a/ViewModels/MainViewModel.Streaming.cs
+++ b/ViewModels/MainViewModel.Streaming.cs
@@ -271,16 +271,26 @@ public partial class MainViewModel : ViewModelBase
 
     /// <summary>The reality word: TEST while a test broadcast runs, LIVE while on the
     /// air, REC during a local-record session.</summary>
-    public string RunningText => IsTestStream ? "TEST" : IsLive ? "LIVE" : "REC";
+    public string RunningText => IsReconnecting
+        ? "RECONNECTING"
+        : IsTestStream ? "TEST" : IsLive ? "LIVE" : "REC";
+
+    /// <summary>Measured facts only while reconnecting: how long, and which attempt. No countdown —
+    /// YouTube's cutoff is not a constant we can promise (spec revised 2026-11-05).</summary>
+    public string ReconnectDetailText => IsReconnecting
+        ? $"0:{ReconnectElapsedSeconds:00} · attempt {Math.Max(1, ReconnectAttempt)} of {StreamRetryPolicy.Default.Attempts}"
+        : string.Empty;
 
     /// <summary>Reality dot color, one meaning per mode: gold = test, red = live,
     /// green = recording. The pre-spec pulsing lamp never shipped; the word + bar
     /// color carry the state and the dot just pins it.</summary>
-    public string RunningDotBrush => IsTestStream
-        ? ThemeBrush.Hex("Brush.TestContainer")
-        : IsLive
-            ? ThemeBrush.Hex("Brush.Primary")
-            : ThemeBrush.Hex("Brush.Positive");
+    public string RunningDotBrush => IsReconnecting
+        ? ThemeBrush.Hex("Brush.Attention")
+        : IsTestStream
+            ? ThemeBrush.Hex("Brush.TestContainer")
+            : IsLive
+                ? ThemeBrush.Hex("Brush.Primary")
+                : ThemeBrush.Hex("Brush.Positive");
 
     /// <summary>Raises every flag the top bar renders, so any state change re-renders
     /// exactly the world that is now true.</summary>
diff --git a/ViewModels/MainViewModel.cs b/ViewModels/MainViewModel.cs
index f104d61..e65dd28 100644
--- a/ViewModels/MainViewModel.cs
+++ b/ViewModels/MainViewModel.cs
@@ -226,6 +226,7 @@ public partial class MainViewModel : ViewModelBase
         ApplyStreamQuality(_selectedQuality);
 
         _youtubeChat.MessageReceived += OnChatMessageReceived;
+        _youtubeChat.ChatEnded += OnChatEnded;
 
         _liveTimer = new DispatcherTimer { Interval = TimeSpan.FromSeconds(1) };
         _liveTimer.Tick += OnLiveTimerTick;
diff --git a/ytLive.Tests/ChatEndedSignalTests.cs b/ytLive.Tests/ChatEndedSignalTests.cs
new file mode 100644
index 0000000..920f3f7
--- /dev/null
+++ b/ytLive.Tests/ChatEndedSignalTests.cs
@@ -0,0 +1,63 @@
+using System;
+using System.Collections.Generic;
+using System.Threading;
+using System.Threading.Tasks;
+using Xunit;
+using ytLive.Services;
+
+namespace ytLive.Tests;
+
+/// <summary>
+/// TASK 32·2 — the reconnecting sign is driven by what YouTube says, not by a timer (spec revised
+/// 2026-11-05: the creator pointed out that the cutoff is not a constant, it depends on variables
+/// "some that are dark to us"). So: the chat-ended marker is a prompt to check, the broadcast status
+/// decides, and a completion ends recovery honestly.
+/// </summary>
+public class ChatEndedSignalTests
+{
+    [Fact]
+    public void ChatEndedEvent_IsNotAChatRow_ButItIsASignal()
+    {
+        const string json = """
+        {
+          "items": [
+            {"id":"e1","snippet":{"type":"chatEndedEvent","publishedAt":"2026-11-05T10:00:00.000Z"}}
+          ]
+        }
+        """;
+
+        // It never becomes a message…
+        Assert.Empty(YouTubeChatService.ParsePage(json).Messages);
+    }
+
+    [Fact]
+    public void ThePollingService_RaisesChatEnded_ForThatMarker()
+    {
+        var service = new YouTubeChatService(new YouTubeAuthService("id", "secret"));
+        var raised = 0;
+        service.ChatEnded += () => raised++;
+
+        service.NotifyChatEnded();
+
+        Assert.Equal(1, raised);
+    }
+
+    [Fact]
+    public async Task BroadcastComplete_EndsRecovery_InsteadOfPretendingToBeLive()
+    {
+        var reported = new List<string>();
+        var recovery = new StreamRecovery(new StreamRetryPolicy(new[] { TimeSpan.Zero }), new StreamRecoverySteps
+        {
+            RestartAudio = _ => Task.CompletedTask,
+            StartPush = _ => Task.FromResult(true),
+            CheckBroadcast = _ => Task.FromResult(BroadcastLiveness.Complete),
+            Report = reported.Add,
+            Log = _ => { },
+        });
+
+        var result = await recovery.RunAsync(CancellationToken.None);
+
+        Assert.Equal(RecoveryOutcome.BroadcastComplete, result.Outcome);
+        Assert.Contains(reported, text => text.Contains("complete") && text.Contains("cannot be resumed"));
+    }
+}