It can be really handy for scraping/archiving websites if they're kind enough to use a guessable id