I just happened to type in the address
http://xml.apache.org/fop/fonts (no extension)
in the browser's address bar (IE6), and to my surprise, I ended up looking at the PDF version of the page in question.
Just wanted to know: Is this deliberately done this way?
No. The FOP website is generated offline by forrest, which I doubt generates content matching extensionless URLs.
BTW I can't reproduce this, Mozilla gets the HTML. I guess it has something to do with some server automatism, perhaps content type negotiation.
You can ask on the forrest list whether they deal with the kind of URL you tried, and perhaps an Apache HTTPD guru whether such an effect can be caused by some server configuration settings. Eventually, I can take this to the infrastructure guys.