You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When the SDK initializes, it gives no feedback as to if the start() has completed or not.
This could lead to third party applications considering the SDK as started, while in reality it's still starting.
It could happen, for example, that the SDK has not yet received the agent version from the agent, while trying to call the call_agent() method, resulting in a InvalidAgentVersion error.
The text was updated successfully, but these errors were encountered:
In my use case I solved it by defining two attributes in my class, one for each agent started status and connection status.
So in both the methods on_agent_started_change and on_connection_status_changed I set these two attributes.
In my case I used threading.Event() for the attributes so i can just return agent_started.wait() and connection.wait() in my start function, but it surely can be done without the use of Event()s
I'm leaving the issue open to mark a missing highlight in the use of these methods in the doc.
lupodellasleppa
changed the title
No feedback for successful start
[DOC] Improve connection status and agent status callbacks use cases
Jul 28, 2021
When the SDK initializes, it gives no feedback as to if the start() has completed or not.
This could lead to third party applications considering the SDK as started, while in reality it's still starting.
It could happen, for example, that the SDK has not yet received the agent version from the agent, while trying to call the
call_agent()
method, resulting in aInvalidAgentVersion
error.The text was updated successfully, but these errors were encountered: