Fix duplicate HMS error notifications from race condition

Move _notified_hms_errors tracking update before the async notification
send to prevent concurrent status callbacks from both seeing the same
errors as new. This mirrors the existing milestone tracking pattern
which correctly updates before the await.

backend/app/main.py

@@ -410,6 +410,9 @@ async def on_printer_status_change(printer_id: int, state: PrinterState):
         # Find new errors that haven't been notified yet
         new_error_codes = current_error_codes - previously_notified
 
+        # Update tracking immediately to prevent duplicate notifications from concurrent callbacks
+        _notified_hms_errors[printer_id] = current_error_codes
+
         if new_error_codes:
             # Get the actual new errors for the notification
             # Filter to severity >= 2 (skip informational/status messages like H2D sends)
@@ -480,8 +483,6 @@ async def on_printer_status_change(printer_id: int, state: PrinterState):
             except Exception as e:
                 logging.getLogger(__name__).warning(f"HMS error notification failed: {e}")
 
-            # Update tracking with all current errors
-            _notified_hms_errors[printer_id] = current_error_codes
     else:
         # No HMS errors - clear tracking so future errors get notified
         if printer_id in _notified_hms_errors: