HTTP 404: How Page Not Found Became a Censor, a Canvas, and a Lifeline

In Tunisia before the 2011 revolution, people saw the page not found screen on so many dissident websites that they gave it a name: Ammar 404, the invisible censor. That story frames this episode on the HTTP 404 status code, a three-digit reply built for machines. The first digit, 4, means the client asked for something the server cannot find. The system was borrowed from older protocols such as FTP, and better codes exist, like 301 for moved permanently and 410 for gone, but few webmasters bother to use them.

From there the episode moves through the odd afterlife of an error message. Servers lie with soft 404s that report success, a problem formalized in a 2004 paper by Ziv Bar-Yossef. Internet Explorer once replaced small error pages with its own, which pushed developers to fake their status codes. Governments and internet providers learned to disguise blocking as a missing page. Designers, political parties, and charities for missing children have all claimed the empty space for their own purposes.

  • A DNS error means the house is not there at all. A 404 means the house is there and the door was answered, but the person you wanted is not home.
  • In 2004 the British provider BT Group launched Cleanfeed, which returned a 404 for blocked pages instead of the 403 forbidden code.
  • Internet Explorer and Chrome would swap in their own error page if a site’s custom 404 was smaller than 512 bytes.
  • During the 2015 UK general election, parties turned their 404 pages into attack ads, while the NotFound project lets any site show a missing child poster on its error page.
  • Standard link crawlers miss roughly 83 percent of real 404 errors because most broken links come from outside a site. Server log analysis catches them.

Leave a Reply

Discover more from pplpod

Subscribe now to keep reading and get access to the full archive.

Continue reading