"We could have continued down the path we were on," said Gavin, again defending the decision to drop XP from the list of operating systems able to run IE9. "We could have added more features to IE, change the UI, blah, blah, blah. We could have made it work across XP, but that's not what's going to push the Web forward.

"We might have been more cautious [by creating a version of IE9 for XP] but you don't get quantum breakthroughs that way," Gavin said.

Microsoft's taking a risk with this strategy, said Al Hilwa, an analyst with IDC who covers browsers for the research firm. (IDC is owned by IDG, the parent company of Computerworld.)

"[XP] users will have to begin to use other browsers to handle [HTML5 content], and that is a risk because they may elect to stay on the other browser and never come back," said Hilwa in an e-mail reply to questions. "It is basic business that when you open such an opportunity for competitors, it is much harder to win them back. This is particularly true in the kind of fast moving disruptive market we are in and the high quality of the competitive browsers."

Gavin was confident that Microsoft could woo back Windows XP users when they eventually upgraded to Windows 7 or its successor.