Currently, there is no way to differentiate how an event from "read_next_event" fails (could be channelLen = 0, could be end of log, etc), so the user is left to guess, potentially incorrectly, when they've hit the end of the log or if they've simply hit a mal-formed message.