@dauwhe @robinberjon I'm reminded of the argument over introducing <di> (to wrap <dt>/<dd> pairs).