waitForAckNack in mesh_interface uses a default instance of Timeout(), which defaults to 20 seconds. It's called directly by node.py and _main_.py.

It looks like it's set properly in node.py, and passed to mesh_interface.py, but ultimately doesn't change the _timeout instance in mesh_interface.

I think it might be the cause of #827 and the cause of my local mesh being unable to do remote node administration over many hops. Everything works, it's just not waiting long enough to get the reply (ie: it can be seen coming through another node a few seconds after the cli gives up). I don't have logs because I'm a third party.