The Mozilla Foundation has fixed 10 security bugs in its open-source Mozilla and Mozilla Firefox browsers and Thunderbird e-mail reader, with the release of new versions of all three products this week. Some of the vulnerabilities could allow attackers to run malicious code on a user's PC via a malicious email, a specially crafted vCard or a malformed graphic on a Web site, project leaders said.

The bug fixes accompany the release of the Firefox 1.0 preview release (PR), a nearly-finished version of the project's next-generation browser. The Mozilla team is hoping recent security concerns about Microsoft's dominant Internet Explorer browser will spur adoption of Firefox 1.0, when it is released, as a safer alternative. Fewer exploits exist for Firefox and Mozilla, partly because the browsers are far less widely used than IE.

The bugs are fixed in Mozilla 1.7.3, Firefox 1.0PR and Thunderbird 0.8, released this week and available from the Mozilla Foundation's Web site, the organisation said. An advisory from Danish security firm Secunia said the flaws were "highly critical", the firm's second-highest rating.

This week's patches, detailed on Mozilla's Web site, are the most serious to affect Firefox so far, according to security researchers. The only other highly critical bug in Firefox so far was a flaw in the browser's libpng component (also patched in Mozilla and Thunderbird), revealed last month, that may have allowed an attacker to take over a system via a malicious graphic. Secunia maintains a list of Firefox 0.x vulnerabilities here.

The worst of the bugs are problems displaying vCards and bitmap image files, and an error involving malformed links. An attacker could cause a buffer overflow by sending an email containing a specially crafted vCard and then potentially execute code, but only if the e-mail were displayed in the preview pane of Mozilla Mail or Thunderbird. An attacker could trigger an integer overflow in the browsers or e-mail readers via an overly wide bitmap image in a Web site or an email, also allowing the execution of code. A link using non-ASCII characters in the hostname could be exploited via a Web site or an e-mail to trigger a buffer overflow and execute code.

Details of the other bugs are available from