When no processor function is registered on a socket not all the tokens from the received message are inserted into the bufferwave. THere is a code loop to insert all the tokens into the bufferwave, then subsequently call the processor function. At the end of the loop there is a test to see if the processor function exists and is compiled. If there is no processor function, there was a break statement, instead of a continue. This meant that the first token was dealt with, and not the rest. The bug has now been fixed