In an exception-based system, it is possible for a task to both write a point value in the datahub, and receive exceptions on that point. This creates a potential positive feedback situation where a task will essentially start talking to itself forever through the DataHub. Even in cases where the client task chooses to ignore point values that have not changed since the last message, it is still possible to get more than one value for a point into the communication "pipeline" and generate the same infinite loop. The Cascade DataHub solves this by marking exceptions as either a normal exception, or as an echo. An echo is a point exception which is being returned to the task which originally transmitted the value change to the DataHub. Most user tasks should ignore or treat echo messages differently from regular exception messages.