The fake printer counts the agent's close frame in its read handler, but
Beast answers the close before that read completes, so stop() can return
and the test can read the counter before it is incremented. On a loaded
runner the check saw 0.
Wait for the counter like the other fake-printer counters do, then keep
the exact-count check.
The status stream read the websocket synchronously. Beast's
expires_after() only applies to async operations, so when a printer
stopped sending, ws.read() blocked forever, and stop_status_stream()
joined that thread when the app closed. On Windows the stuck process
also froze the next launch through the shared WebView2 browser.
Run the whole stream on one io_context owned by the stream thread, and
have stop_status_stream() post an abort to it. A stop on an open
websocket sends a close frame and waits up to 1 s for the answer.
Keep-alive pings detect a dead printer, and the stream reconnects until
stopped, with the backoff capped at 32 s. When a quiet printer still
answers pings, its last status is sent again. The stream starts under
connect_mutex after a staleness check, and a stop from the stream's own
callback no longer joins its own thread.
The Moonraker tests move into test_moonraker_printer_agent.cpp.
HanifKoh