Ben and I were discussing a feature addition (related to the dragonfly SensorMonitor), where you could name the sensor endpoints identically to their logged endpoint, but the endpoint wouldn't be bound to the exchange creating a naming conflict. Instead the endpoint would be accessed via sensor_monitor.endpoint_name. This would more naturally support get/set functionality of the endpoints.
The to-do list would be:
Modify endpoint.pyto allow an attribute self.bound = False
Modify spimescape.py so that add_endpoint will skip appending such endpoints to self._bindings (then service.py will skip the queue binding).
Modify spimescape.py so that on_request_msg will attempt to be able to divert requests to a child and not ignore the rks after the first . delimiter.
dragonfly get A.B.C.D will route to A, which attempts to handle the request using B.C.D,
failing that, A will check if child B exists and can handle a request using C.D
Ben and I were discussing a feature addition (related to the dragonfly SensorMonitor), where you could name the sensor endpoints identically to their logged endpoint, but the endpoint wouldn't be bound to the exchange creating a naming conflict. Instead the endpoint would be accessed via sensor_monitor.endpoint_name. This would more naturally support get/set functionality of the endpoints.
The to-do list would be:
endpoint.py
to allow an attributeself.bound = False
spimescape.py
so thatadd_endpoint
will skip appending such endpoints toself._bindings
(thenservice.py
will skip the queue binding).spimescape.py
so thaton_request_msg
will attempt to be able to divert requests to a child and not ignore the rks after the first.
delimiter.dragonfly get A.B.C.D
will route to A, which attempts to handle the request using B.C.D,