Eventually I want to change the way encodings and byte order are done to be like BitOrder. There should be a EncodingChangeParser (and Unparser) which is inserted when a change may be needed. This "parser" or unparser would just do that change. Since the vast majority of data formats will use a single encoding throughout all their textual data, this will reduce overhead since there is a fair amount of inspection and such we have to do each time we set the encoding.