In those days I was trying to enhance parsing capabilities of Atom .NET and since I based all the API around XPath I tried to gain more advantages from .NET implementation. The only limit I recognized is the lack of the capability to “serialize” the content of a visiting node (something like InnerXml or OuterXml properties in the DOM classes or ReadInnerXml(), ReadOuterXml() in the pull parser classes). So I asked around and Oleg Tkachenko (MS MVP) pointed me out the MVP-Xml library, a MVP’s collaborative project to enhance XML support in .NET. Cool!