Those who do not understand history are doomed to repeat it ;) In the beginning, there were tens of text processors with rich text encoding syntaxes (troff, Te

Those who do not understand history are doomed to repeat it ;) In the beginning, there were tens of text processors with rich text encoding syntaxes (troff, TeX, IBM pre-GML, DEC stuff, WordPerfect, etc.) Then came SGML with customizable meta syntax and vocabulary to unify those. A large part of HTML is actually contained in ISO 8879, as "folklore" elements <p>, <h1>-<h2>, and so on. Then came HTML as an SGML application, and later XML as an SGML subset. Then people forgot the aspects of SGML not in XML, and invented ad-hoc syntax such as Markdown and MediaWiki syntax (both of which can be partly represented as SGML SHORTREF syntax), and Ian Hickson (WHATWG) started to specify HTML parsing rules in an informal ad-hoc way, based on the tag omission/inference rules of W3C's HTML4, and what was called "real-world" HTML (nevertheless also inventing new elements <main>, <article>, <aside>). Then, due to presentation as prose and mere enumeration of element names, the basic construction of HTML as a formal DTD grammar was lost, and WHATWG added elements and rules nilly-willy, without being informed by a DTD grammar or other formal mechanism to ensure parse-ability.