During my initial implementation of Mailman 3 support, I made an attempt at decoding quoted-printable encoded email bodies. That implementation isn't working that well. I only took 2 byte encoded UTF-8 characters into account, but we of course need to also handle 3 and 4 byte characters.