Of course, because of "link memory", we have to preserve the old links, of the form "yyyy/mm/dd/hhMMsssssssss", like the article you are now reading, whose link id is "2015/05/18/2085000000000". The old Pebble used that string with .html stuck on the end of it, even though the blog articles were actually saved in XML in individual physical files on disk. The blog text was crude HTML inside a CDATA (ugh). My new implementation keeps all the blog posts in a relational database (PostgreSQL). The Java EE @WebServlet annotation makes it easy to map from /*.html to the corresponding JSF page with a request parameter (blogId) containing the date stamp string. You probably don't care about any of this, gentle reader, unless you're a Java or Web Developer, but what it means is: any old links you may find (like old Google search results, old Bookmarks/Favorites, and so on) that refer to pages on the old Theories blog, should still "just work"!