Model Atom in terms of RDF without fully committing to striped syntax or requiring understanding of RDF.
RSS 1.0 *almost* works, but "module proposers forget about RDF and design modules that break interpretability as RDF/XML (e.g. http://igargoyle.com/rss/1.0/modules/event/)", as a result, "RSS 1.0 claims to be an RDF vocabulary, but in practice it ends up being an XML schema." Defaulting to XML Literal values ensures cleaner data for RDF processing.
Atom Feeds and Entries are composed entirely of properties and their values. Aside from atom:feed and atom:entry, the core Atom elements represent properties that may be applied to Atom Feeds and/or Atom Entries. In RDF terms, all RDF Triples [RDF Abstract Syntax] denoted by standard Atom elements have either an Atom Feed or Atom Entry as their subject, and their predicates are the child elements' fully-qualified Element names as in RDF/XML.
There are three models for Atom properties defined in this specification:
The content of the element MUST be CDATA, and have no attributes.
The value of the property is an RDF literal.
The Element can contain mixed content, and have attributes.
For core Atom elements, the value of the property is an RDF XML Literal. Extensions may use other mechanisms to signal structural characteristics.
* RDF Resource
The Element MUST be empty and MUST have a single attribute denoted by the tuple
In Atom documents, extension elements appear as child elements of atom:head or atom:entry, where they are considered to be properties of the enclosing Feed or Entry, respectively. The semantics of extension elements appearing in other locations are not defined by this specification.