"The constraint that we deploy things cautiously, incrementally, has meant that OWL was built on top of RDF which was built on top of XML which was built on top of URIs and Unicode. All good in principle, but it can make the finished product hard to present and hard to comprehend. To understand how to read and write OWL documents, you need to know XML level stuff (entities, elements/attribute rules, charset issues), as well as RDF's convention for encoding binary relations in XML, as well as OWL's convention for encoding Description Logic constructs in RDF."

-- Dan Brickley on SWIG-L