All the standard templates that show how to construct a basic XHTML page include a public identifier of
http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd and often a namespace name of
http://www.w3.org/1999/xhtml. As the blog points out, these are not actually hyperlinks, they only play them on TV. Huge quantities of software are requesting these URLs 24×7, putting a load on their servers. Often times this results from unfortunate defaults in off-the-shelf XML components such as parsers.
But what did you expect?
This is the web equivalent of having a front-desk receptionist hand out a stacks of self-addressed, stamped postcards, then complaining about how much mail the company gets from all around the world.
HTTP URLs are great for identifiers on a technical basis: they are based on DNS names and have the important qualities of uniqueness and persistence. But as far as human factors go, they are a terrible choice (though with a great deal of inertia at this point). -m