The implementation of this extension consists of a small parser written in Python[9], that acts as a preprocessor. Python is an object oriented scripting language.
The decision to preprocess a requested document is made by the server by looking at the extension, an extension of .xhtml tells the server to start the parser.
The current context is constructed from information passed via the GET method. The parser reads the whole file and decides which parts are to be presented. The extraneous tags are eliminated and the requesting user-agent only sees a valid HTML document.
This way of extending HTML has two advantages, the extensions remain private to where they were introduced and no special browser is needed to view the results. Allthough all the extensions introduced were designed in a way so that they could, at a later stage barring public discussion with the concerned group, introduced into mainstream and moved from the server side to the user-agent side.