We never needed to parse tag soup. We only needed to say which DTD defines the new elements a document uses, and what a browser should do with valid but unknown

We never needed to parse tag soup. We only needed to say which DTD defines the new elements a document uses, and what a browser should do with valid but unknown elements. The latter could use #FIXED or default attribute values in the DTD, because some new elements have human-readable content and others don’t.