ScribeSource deosn't handle the zero message request correctly.
When scribed sends zero message request like this:
sent <0> messages to remote scribe server <dwen:1564>
The log receiver of scribe-source will return ResultCode.TRY_LATER.
The return value reultes that scribed falls into infinite loop to
retry to send the zero message request out, and the whole
message channel is blocked.
So I suggest that ResultCode.OK is returned when the size of
LogEntry list is 0.