In order to help test your decoder (and others), how complicated would it be to add a 'raw' mode (-e r) to the encoder. Where the data string was treated as codewords (ie. already encoded into a valid ASCII/TEXT/C40/X12/EDIFACT sequence). It would probably make more sense if the encoder calculated any required padding (pseudo random stuff) and the error checking so that the user wouldn't have to.
(submitted by mblaughton on behalf of mungewell)