Note that the receive threads "break" when they are processing responses or unsolicited requests from BCS. These "breaks" are necessary since some computational time needs to elapse inside the imservice/EIM-connector in order to handle the event. A future optimization would be to reduce these breaks through traditional mechanisms e.g. fork, etc. However and nonetheless, in the BCS there must always existing some mechanism to queue and/or retry events in the case where the client is unresponsive.
The dashed green lines represent passage of time, and separate different stages for this example: login, followed by unsolicited presence update, followed by send message, followed by unsolicited message.
Note that the receive threads "break" when they are processing responses or unsolicited requests from BCS. These "breaks" are necessary since some computational time needs to elapse inside the imservice/EIM-connector in order to handle the event. A future optimization would be to reduce these breaks through traditional mechanisms e.g. fork, etc. However and nonetheless, in the BCS there must always existing some mechanism to queue and/or retry events in the case where the client is unresponsive.
The dashed green lines represent passage of time, and separate different stages for this example: login, followed by unsolicited presence update, followed by send message, followed by unsolicited message.
การแปล กรุณารอสักครู่..
