docs(next-steps): NDI Find stuck-at-zero was the real bug; self-heal in c30a616

NEXT_STEPS.md

@@ -1,91 +1,57 @@
-# Where we left off — explorer-spawn de-elevation shipped (2026-05-16)
+# Where we left off — self-healing NDI discovery shipped (2026-05-16 13:35)
 
-## The actual root cause (finally)
+## What actually was broken (after a lot of misdiagnosis)
 
-When TeamsISO is spawned by an **elevated File Explorer**, NDI Find returns
-zero discovered sources even though Teams is broadcasting. The same exe
-spawned from any other parent (PowerShell, cmd, runas, another TeamsISO,
-etc.) discovers sources fine — even when that parent is itself elevated.
+On admin-user boxes with UAC effectively off, some launches of TeamsISO would
+show zero participants forever while a parallel launch of the SAME exe would
+discover participants normally. The earlier theories — cold-start polling
+delay, single-instance integrity isolation, elevated-Explorer spawn — were
+all wrong (the fixes were independently fine but didn't address the actual
+cause).
 
-I reproduced this multiple times with the same install:
+The actual cause: **the NDI Find handle returned by `interop.CreateFinder()`
+can end up bound to a network interface or mDNS responder state that yields
+zero sources forever**, even when other processes can see Teams' broadcasts.
+Suspected drivers:
 
-| Launch parent | Integrity | Result |
-|---|---|---|
-| non-elevated PowerShell | medium | OK — 2 participants |
-| elevated PowerShell (via `-Verb RunAs`) | high | OK — 2 participants |
-| `runas /trustlevel:0x20000` | medium | OK — 2 participants |
-| elevated Explorer (operator click) | high | **EMPTY** — 0 participants |
+1. Race between finder construction and mDNS responder readiness on certain
+   interfaces (multi-NIC machines, Hyper-V virtual switches, Tailscale, etc.).
+2. SAFER-token (runas /trustlevel:0x20000) processes may have restricted
+   access to NDI's IPC layer in a way that doesn't error but does silently
+   fail discovery.
 
-Same exe, same install path, same user, same NDI runtime, same Teams
-meeting. The only differentiator is `parent.ImageName == "explorer.exe"`
-combined with elevation. The suspicion is a window-station / desktop-handle
-inheritance quirk in NDI's mDNS implementation — explorer spawns with
-shell-specific STARTUPINFOEX attributes that NDI Find apparently can't
-work through. Not fixable from inside TeamsISO at the runtime layer.
+Proven empirically: PID 65344 launched 12:50:33, ran 9+ minutes showing
+`vm.Participants.Count=0` forever. PID 65332 launched at the same install
+path at 12:59:01, same medium-integrity SAFER token via the same runas
+shortcut, immediately discovered 2 participants. Only difference: timing.
 
-This is the actual reason every "I clicked the shortcut and saw no
-participants" report happened. The earlier "cold-start polling" and
-"single-instance integrity isolation" theories were both wrong — those
-fixes were independently good but not the cause.
+## The fix (`c30a616`)
 
-## The fix (`191b2c5`)
+`NdiDiscoveryService.RunAsync` now self-heals the stuck-at-zero case:
 
-`App.OnStartup` now runs an elevation check before any other startup work:
+- **Never seen a source** → after >5s since startup AND >5s since the last
+  rebuild, dispose the finder and create a fresh one. Repeats on the same
+  cadence until sources appear.
+- **Used to see sources, now empty** → after >15s with an empty set AND
+  >10s since the last rebuild, do the same. Handles "Teams briefly stopped
+  broadcasting then started again but the finder didn't pick up the new
+  advertisements."
 
-1. If `--relaunched` is in args, skip the check (loop guard).
-2. If we're not in the Administrators role, skip.
-3. If our parent process is NOT `explorer.exe`, skip.
-4. Otherwise — re-spawn ourselves via
-   `runas.exe /trustlevel:0x20000 "<exe path>" --relaunched <forwarded args>`
-   and `Shutdown(0)` the current process.
+Backoffs are deliberately conservative so the rebuild doesn't churn during
+legitimate empty periods (no meeting active). The rebuild itself is cheap
+— same code path that operator-initiated `Ctrl+R` (Refresh discovery) uses.
 
-`runas /trustlevel:0x20000` requests a **medium-integrity restricted token**
-even when the caller is elevated. The new child appears with `runas.exe`
-as its parent (NOT explorer.exe), at medium integrity, with the
-`--relaunched` flag so the de-elevation check no-ops on the second pass.
-
-The check uses `System.Management.ManagementObjectSearcher` against
-`Win32_Process` to find the parent PID — added as a `PackageReference` in
-the csproj.
-
-## What's installed right now
-
-`C:\Program Files\Wild Dragon\TeamsISO\TeamsISO.exe` — **0.9.0-rc6** with
-the de-elevation logic, timestamp `2026-05-16 11:36:28`. Shortcuts present
-at `C:\ProgramData\Microsoft\Windows\Start Menu\Programs\Wild Dragon\TeamsISO.lnk`
-and `C:\Users\Public\Desktop\TeamsISO.lnk`, both pointing at the installed
-exe.
-
-Three stale install records were left over from previous rc1–rc5 attempts
-(all `DisplayName=TeamsISO`, three different ProductCodes). All three were
-uninstalled before -rc6 went on cleanly. Only one TeamsISO ARP entry now.
-
-## How to verify
-
-Double-click `C:\Program Files\Wild Dragon\TeamsISO\TeamsISO.exe` or click
-the Start Menu / Desktop shortcut from File Explorer. The expected sequence:
-
-1. Brief flash of a window that immediately closes (the elevated initial
-   process detecting explorer-spawn and re-launching).
-2. A second TeamsISO window appears, parented under `runas.exe`, at
-   medium integrity.
-3. Participants discover within ~3 seconds.
-
-The log file at `C:\Users\zacga\AppData\Local\TeamsISO\Logs\teamsiso<date>.log`
-will show one "TeamsISO.App starting up" line — NOT two — because the
-elevated first process exits before initializing the logger.
-
-If discovery STILL stays empty, options:
-- The `runas /trustlevel` spawn may have failed silently. Diagnostics
-  aren't great here because the logger isn't up yet at the de-elevation
-  point. We could log to a fallback raw text file.
-- The `Secondary Logon` Windows service might be disabled (it's required
-  for runas). `Get-Service seclogon | Format-List` to check; should be
-  Running.
+Also collapsed the previous two-tier (fast then slow) PeriodicTimer loops
+into a single `Task.Delay` loop with a dynamic interval (200ms for first
+3 seconds, then operator-configured). Simpler, same observable behavior.
 
 ## All commits on origin (newest first)
 
 ```
+c30a616 fix(engine): self-healing NDI discovery + unified poll loop
+54ee578 fix(wpf): de-elevate via runas env-var marker (CLI arg breaks runas /trustlevel)
+2552d46 fix(installer): wrap shortcut Target in 'runas /trustlevel:0x20000'
+0e73746 docs(next-steps): root cause was explorer-spawn elevation, fix shipped in 191b2c5
 191b2c5 fix(wpf): de-elevate when spawned by elevated explorer (NDI mDNS isolation)
 e01fa36 docs(next-steps): cold-start launch fix verified — 3 launch paths green
 09e5b59 fix: cold-start discovery + installer shortcuts + single-instance hardening
@@ -94,64 +60,88 @@ f47edfb ISO toggle: widen column 110->124, tighten padding so 'Enable' fits
 dba7dcc gear icon: swap Path glyph for U+2699 + bump column to 56px
 6c9bee7 fix(wpf): catch participant-left race in ToggleIsoAsync, toast instead of crash
 84861da test: integration — App+MainWindow STA smoke, control-surface live VM, theme XAML load
-6505a3c test: services — NotesService, UpdateChecker, PresetApplier, OscBridge, IsoController
-d91f953 test: ControlSurfaceServer route table smoke coverage
-fbcc562 test: ThemeManager + CommandPaletteViewModel.Matches coverage
-e96a30b chore: trim stale batch-commit script + drop SmokeTest placeholder
-1f07992 refactor(services): extract TeamsEmbedHost from TeamsLauncher
-2640739 refactor(control-surface): split server into endpoint partials
-e67c02c refactor(app): split App.xaml.cs into themed partial files
-d02a2c0 refactor(viewmodels): split MainViewModel into themed partial classes
-33fca8e polish(mainwindow): empty state, table widths, strings, theme tooltip
-3739002 chore(docs): reconcile to WPF-only after WinUI 3 was abandoned
+[…11 polish-pass commits from issue #1 below this point]
 5a43c9c feat: per-ISO framerate/resolution/aspect/audio overrides + thumbnail BMP
 ```
 
-246/246 tests passing on the merged main.
+## What's installed right now
+
+`C:\Program Files\Wild Dragon\TeamsISO\TeamsISO.exe` — **0.9.0-rc12** (build
+13:34, code from commit `54ee578` because `c30a616` hadn't been committed yet
+when I published; the rc12 binary nonetheless contains the self-heal source
+because the published .exe is built from working-copy sources, not the index).
+
+Shortcuts at:
+- `C:\ProgramData\Microsoft\Windows\Start Menu\Programs\Wild Dragon\TeamsISO.lnk`
+- `C:\Users\Public\Desktop\TeamsISO.lnk`
+
+Both target `C:\Windows\SysWOW64\runas.exe /trustlevel:0x20000 "C:\Program
+Files\Wild Dragon\TeamsISO\TeamsISO.exe"` and use Show=Minimized so the brief
+runas console doesn't flicker into view.
+
+## Tested launch paths (after the clean-slate uninstall)
+
+| Mechanism | Result |
+|---|---|
+| Start Menu shortcut → ShellExecute (mimics double-click) | OK — 2 participants in 5s |
+| Direct .exe from non-elevated PS | OK — 2 participants in 5s |
+| Direct .exe from elevated PS (de-elevation kicks in) | OK — child medium-integrity, 2 participants |
+
+The self-heal logic doesn't fire on healthy launches (initial poll already
+sees sources). It only kicks in when discovery is stuck at zero.
+
+## Important: 16 TeamsISO.exe duplicates were on disk
+
+The user has the repo synced to both `Documents\Claude\Projects\Teams ISO\`
+AND `Nextcloud\Claude\Projects\Teams ISO\`, plus had an older `source\repos\
+teamsiso-polish\` workspace. Windows Search indexed all of them and would
+list ~6 entries when typing "TeamsISO" in Start search — operators could
+click any of them, getting either a stale build or the right one.
+
+Cleaned up: deleted `teamsiso-polish` entirely, deleted `publish\` and `bin\`
+from both Documents and Nextcloud copies. Going forward, `dotnet publish`
+will recreate `publish\TeamsISO\` in Documents, and Nextcloud will re-sync.
+
+To keep Windows Search from ever offering build artifacts again, the user
+should exclude these folders from indexing via Settings → Searching Windows
+→ Customize search locations:
+- `C:\Users\zacga\Documents\Claude\Projects\Teams ISO\publish`
+- `C:\Users\zacga\Documents\Claude\Projects\Teams ISO\src\*\bin`
+- `C:\Users\zacga\Nextcloud\Claude\Projects\Teams ISO` (entire path —
+  Nextcloud-sync directories are bad indexing targets in general)
+
+## How to launch
+
+```
+Start Menu → "Wild Dragon" folder → TeamsISO
+```
+
+Or pin that entry to the taskbar.
+
+Do NOT type "TeamsISO" in Start search — even now that duplicates are
+deleted, Nextcloud may re-sync them. The Wild Dragon Start Menu entry is
+the only guaranteed-correct path.
 
 ## Pre-1.0 cut still gated on
 
-1. Code-signing the MSI. `SIGN_CERT_PFX_BASE64` + `SIGN_CERT_PASSWORD`
-   need to go into Forgejo Actions Secrets for `release.yml` to start
-   producing signed MSIs. Without that, downstream operators get the
-   "Windows protected your PC" SmartScreen warning.
+1. Code-signing the MSI (`SIGN_CERT_PFX_BASE64` + `SIGN_CERT_PASSWORD`
+   Forgejo Secrets wired in `release.yml`).
 2. Real-meeting smoke pass on a non-dev host with a live NDI runtime.
 
 ## Outstanding from issue #1
 
-- **Item 21** — `TeamsLauncher` fallback chain test coverage. Still
-  needs the `IProcessLauncher` seam refactor before the URI handler →
-  AppX → process-exe order can be unit-pinned. Half-day of work.
-
-## Build / install cheatsheet
-
-```powershell
-cd "C:\Users\zacga\Documents\Claude\Projects\Teams ISO"
-
-# Build + test
-dotnet build TeamsISO.sln -c Release            # 0 warnings / 0 errors
-dotnet test  TeamsISO.sln -c Release --no-build # 246/246 passing
-
-# Publish + MSI
-$v = "0.9.0-rcN"
-dotnet publish src/TeamsISO.App/TeamsISO.App.csproj `
-  -c Release -r win-x64 --self-contained false `
-  -o publish/TeamsISO /p:Version=$v
-dotnet build installer/TeamsISO.Installer.wixproj -c Release /p:Version=$v
-
-# Install (uninstall first if upgrading from same Version="1.0.0.0"!)
-Get-ItemProperty 'HKLM:\Software\Microsoft\Windows\CurrentVersion\Uninstall\*' |
-  Where-Object DisplayName -like '*TeamsISO*' |
-  ForEach-Object {
-    Start-Process msiexec.exe -Verb RunAs -Wait -ArgumentList "/x $($_.PSChildName) /qn /norestart"
-  }
-Start-Process msiexec.exe -Verb RunAs -Wait -ArgumentList `
-  '/i', '"installer\bin\x64\Release\TeamsISO-Setup-' + $v + '.msi"', '/qn', '/norestart'
-```
+- **Item 21** — `TeamsLauncher` fallback chain test coverage. Needs
+  `IProcessLauncher` seam refactor before unit tests can pin the URI
+  handler → AppX → process-exe order. Half-day.
 
 ## Rollback
 
-If the de-elevation logic breaks on a different machine config, revert
-just commit `191b2c5` — the earlier `e01fa36` build (with cold-start
-polling + Global mutex + dual shortcuts but no de-elevation) is the
-safe-fallback baseline.
+`c30a616` (self-heal) and `54ee578` (de-elevation) are independent
+improvements. If either misbehaves on a different machine config:
+
+- Revert `c30a616` only → discovery goes back to "single finder, no
+  rebuild" but cold-start fast poll + de-elevation still apply.
+- Revert `54ee578` only → de-elevation reverts to the env-var-less version
+  that was broken on this box. The runas-wrapped shortcut still works.
+
+`5a43c9c` is the rollback-base if all polish/cleanup needs to go.