What do we gain by having the post-processing steps available from the unix shell command line? Can you think of an example that adds to the efficiency of the process?