Desktop auto-update process conflicts and stuck UI fixes

Fixes on record

Merged to main. A stamp counts the issues a fix is linked to.

• #38158 fix(desktop): stop in-app update from killing its own backend (#37532)

  merged to main

  files: apps/desktop/electron/main.cjs, hermes_cli/main.py, tests/hermes_cli/test_update_stale_dashboard.py

  Show the diff

  • apps/desktop/electron/main.cjs

    @@ -1412,6 +1412,18 @@ async function applyUpdatesPosixInApp(opts = {}) {
         PATH: [extraPath, process.env.PATH].filter(Boolean).join(path.delimiter)
       }
     
    +  // `hermes update` reaps stale `hermes dashboard` backends (a code update
    +  // leaves the running process serving old Python against the freshly-updated
    +  // JS bundle). But OUR backend is one of those processes, and killing it
    +  // mid-update produces the boot→kill→crash loop in #37532 — the desktop
    +  // already restarts its own backend via the rebuild+relaunch below, so the
    +  // reap must spare it. Hand the live backend's PID to the update process;
    +  // _kill_stale_dashboard_processes reads HERMES_DESKTOP_CHILD_PID and excludes
    +  // it while still reaping any genuinely-orphaned dashboards. (#37532)
    +  if (hermesProcess && Number.isInteger(hermesProcess.pid)) {
    +    env.HERMES_DESKTOP_CHILD_PID = String(hermesProcess.pid)
    +  }
    +
       // Branch-pin so a non-main checkout doesn't get switched to main (and self-heal
       // to main when the pinned branch no longer exists on origin).
       let branchArgs = []

  • hermes_cli/main.py

    @@ -7363,7 +7363,10 @@ def cmd_gui(args: argparse.Namespace):
         sys.exit(launch_result.returncode)
     
     
    -def _find_stale_dashboard_pids() -> list[int]:
    +def _find_stale_dashboard_pids(
    +    *,
    +    exclude_pids: set[int] | None = None,
    +) -> list[int]:
         """Return PIDs of ``hermes dashboard`` processes other than ourselves.
     
         ``hermes dashboard`` is a long-lived server process commonly started and
    @@ -7378,6 +7381,15 @@ def _find_stale_dashboard_pids() -> list[int]:
         it.  This helper is just the detection step; see
         ``_kill_stale_dashboard_processes`` for the kill.
     
    +    *exclude_pids* is an optional set of PIDs that must never be returned.
    +    This is used by the Hermes Desktop Electron app to protect its own
    +    backend child process: when the desktop spawns ``hermes dashboard`` as
    +    a backend and triggers an auto-update, the update must not kill the
    +    dashboard that the desktop itself manages.  The desktop sets the
    +    environment variable ``HERMES_DESKTOP_CHILD_PID`` on the spawned
    +    backend process; ``_kill_stale_dashboard_processes`` reads it and
    +    passes it here.  (#37532)
    +
         Returns an empty list on any scan error (missing ps/wmic, timeout, etc.).
         """
         patterns = [
    @@ -7453,6 +7465,8 @@ def _find_stale_dashboard_pids() -> list[int]:
         except (FileNotFoundError, subprocess.TimeoutExpired, OSError):
             return []
     
    +    if exclude_pids:
    +        dashboard_pids = [p for p in dashboard_pids if p not in exclude_pids]
         return dashboard_pids
     
     
    @@ -7604,7 +7618,18 @@ def _kill_stale_dashboard_processes(
         launch args (--host, --port, --insecure, --tui, --no-open).  The user
         restarts it manually; a hint is printed.
         """
    -    pids = _find_stale_dashboard_pids()
    +    # When the Hermes Desktop Electron app spawns this dashboard as a
    +    # backend child, it sets HERMES_DESKTOP_CHILD_PID so that the update
    +    # path can skip killing the desktop-managed process.  (#37532)
    +    exclude: set[int] | None = None
    +    raw_pid = os.environ.get("HERMES_DESKTOP_CHILD_PID")
    +    if raw_pid:
    +        try:
    +            exclude = {int(raw_pid)}
    +        except (ValueError, TypeError):
    +            pass
    +
    +    pids = _find_stale_dashboard_pids(exclude_pids=exclude)
         if not pids:
             return
     

  • tests/hermes_cli/test_update_stale_dashboard.py

    @@ -185,6 +185,48 @@ def test_invalid_pid_lines_skipped(self):
                 pids = _find_stale_dashboard_pids()
             assert pids == [12345]
     
    +    def test_exclude_pids_filters_specified_pids(self):
    +        """exclude_pids removes specific PIDs from the result — used by
    +        the Desktop Electron app to protect its own backend child.  (#37532)
    +        """
    +        with patch("subprocess.run") as mock_run:
    +            mock_run.return_value = MagicMock(
    +                returncode=0,
    +                stdout="\n".join([
    +                    _ps_line(11111, "hermes dashboard --port 9119"),
    +                    _ps_line(22222, "hermes dashboard --port 9120"),
    +                    _ps_line(33333, "hermes dashboard --port 9121"),
    +                ]) + "\n",
    +                stderr="",
    +            )
    +            # Exclude the desktop-managed backend PID
    +            pids = _find_stale_dashboard_pids(exclude_pids={22222})
    +        assert 11111 in pids
    +        assert 22222 not in pids
    +        assert 33333 in pids
    +
    +    def test_exclude_pids_none_is_noop(self):
    +        """Passing exclude_pids=None (the default) changes nothing."""
    +        with patch("subprocess.run") as mock_run:
    +            mock_run.return_value = MagicMock(
    +                returncode=0,
    +                stdout=_ps_line(12345, "hermes dashboard --port 9119") + "\n",
    +                stderr="",
    +            )
    +            pids = _find_stale_dashboard_pids(exclude_pids=None)
    +        assert pids == [12345]
    +
    +    def test_exclude_all_pids_returns_empty(self):
    +        """If all matched PIDs are excluded, the result is empty."""
    +        with patch("subprocess.run") as mock_run:
    +            mock_run.return_value = MagicMock(
    +                returncode=0,
    +                stdout=_ps_line(12345, "hermes dashboard --port 9119") + "\n",
    +                stderr="",
    +            )
    +            pids = _find_stale_dashboard_pids(exclude_pids={12345})
    +        assert pids == []
    +
     
     @pytest.mark.skipif(sys.platform == "win32", reason="POSIX kill semantics")
     class TestKillStaleDashboardPosix:

• #52203 fix(update): announce gateway drain waits so desktop updates don't look hung

  merged to main

  files: hermes_cli/gateway.py, hermes_cli/main.py, tests/hermes_cli/test_gateway_service.py