right now, we are doing quite fine at handling text/html. we could perhaps add specialized handling for, e.g., application/pdf (and other EPUB?).