A flaw allowing JavaScript code to appear in tweets was plugged Wednesday morning by engineers at Twitter.com. The far-ranging cross-site scripting attack, which reportedly originated as a harmless, proof-of-concept attack by an Australian teenager, was estimated to affect more than 500,000 users of the social network site. The attacks exploited a common JavaScript feature, onmouseover, that allows programmers to create discrete actions when users move their mouse cursor over an area of a web page. The worm allowed attackers to infect a user's account, as well as those of their followers. – GM