Being "able to make those libraries" is a function of two things: (i) knowing LISP, and (ii) knowing the problem domain that the libraries cover.<p>The average programmer with the mission to write urlencode/urldecode functions will probably get a subtle detail wrong. It takes real experience with how the web really works to get this sort of thing right. Many people using languages like LISP use ideas, such as continuation-based webapps, that sound really nice but that will have reliability and scalability problems if you try to use them in production.<p>When it comes to XML parsing, it's a very difficult problem and you might notice that every major revision of a platform often comes with new XML infrastructure, because the infrastructure on the last one sucked. Microsoft finally has something that's halfway decent in Linq-to-XML, but it's only been through trial and error in the style of Thomas Edison.