WHEN the Code Red computer worm made its debut on 12 July 2001, it seemed harmless enough. A week later, it transformed into one of the worst attacks the internet has ever known. In the intervening days, someone had fixed a bug in the worm. This allowed it to spread like wildfire, striking computers at random by infiltrating a program called Microsoft Internet Information Services. Once lodged inside, Code Red sent copies of itself to other machines worldwide - it is this ability to replicate that distinguishes worms from ordinary computer viruses. Within 24 hours it had infected over 350,000 machines (see Maps).

The flood of online traffic was so intense it blacked out internet access to thousands of people in many parts of the US, and knocked out servers in scores of companies, including Associated Press and Fedex. What's more, Code Red carried a nasty surprise: each copy ...