The proton ReadableBuffer API defines that when reading a string the readable buffer should consume the bytes by advancing the read index as defined in the interface API docs.  The current implementation is not doing this which currently does not break the codec but code if future improvements remove the condition that prevents this from causing errors.