Should catch the exception and add all events back to the queue, preventing loss of events.
For the record: It seems RescueTime has issues with lost connections as well. My connection died today at 8:57, a few minutes before this issue was opened I noticed it was no longer running. No activity after 9:00 had been reported. Restarting it only resumed logging and did not send the apparently lost events between crash and restart.
Should catch the exception and add all events back to the queue, preventing loss of events.
For the record: It seems RescueTime has issues with lost connections as well. My connection died today at 8:57, a few minutes before this issue was opened I noticed it was no longer running. No activity after 9:00 had been reported. Restarting it only resumed logging and did not send the apparently lost events between crash and restart.