Will Firefox repeat Netscape’s mistakes?

As a recent column noted, the spate of security problems that have plagued Internet Explorer could well mark a turning point in the fortune of one of its main rivals, Mozilla Firefox. But there is another side to this story, for it is important to note that many of Microsoft's woes in this area are self-inflicted, a result of its dogged determination to integrate the Web browser with its operating system.

This formed the heart of its defence in the US anti-trust trial, where it justified the bundling of a Web browser with Windows on the grounds that the two were inextricably mixed. While that was hardly the case at the beginning – Windows 95 originally came without a Web browser, which was available separately as the Internet Jumpstart Kit in the Windows Plus! Pack – Microsoft took great pains to make it so afterwards. But a consequence of this is that flaws in the browser are so deeply plumbed into the operating system that they give almost unlimited power to anyone able to exploit them.

Against this background, the news that Mozilla will be working with Adobe, Apple, Macromedia and Sun to develop an open, scriptable plugin model is worrying. The logic behind this move seems to be that in order to capitalise on users' increasing willingness to consider alternatives to Internet Explorer, Firefox needs to match it in all areas, including plugins. What is particularly ironic about this move is that it represents an eerily exact rerun of an earlier - failed - strategy.

Firefox is the latest iteration of Mozilla, which was born out of the decision of Netscape to make the code for its next-generation browser suite open source. At the time Netscape Navigator first appeared on the scene, users were impressed above all by its speed and stability – even as a beta. This is also true of Firefox, which is sleek, speedy and already stable at version 0.9.

When Microsoft finally woke up to the Internet – and the threat posed by Netscape – one of its responses was a refashioning of its OLE controls as ActiveX: self-contained chunks of code that could be embedded in a Web page to add advanced functionality. Rather too much functionality as it turned out, since they proved – and continue to prove - to be an enormous security risk.

Nonetheless, Netscape felt compelled to match Microsoft with a plugin architecture in version 2.0 of its browser – in fact, precisely the plugin architecture that Mozilla proposes to build on for its own. Netscape gradually added more and more bells and whistles until it became the slow, bloated code that ultimately led to the company's takeover by AOL. By then, Internet Explorer had matched it in functionality, was more stable – and came free with Windows.

As well as negating the very benefits of Firefox that are causing many now to turn to it, bolstering plugin support will also undermine Internet standards – another area where Firefox curently excels.

Two of the main supporters of the new Mozilla initiative are Adobe and Macromedia. Their respective plugins are for Acrobat and Flash. The latter turns the active Web into a passive multimedia show – a kind of online television – supplanting the W3C's Scalable Vector Graphics, while Acrobat's PDF format is a rival to HTML itself, especially its more advanced features like Cascading Style Sheets. Similarly, Firefox would be more likely to regain Netscape's browser throne if it chose not to repeat its past mistakes by emulating Internet Explorer and adding unnecessary and potentially insecure features.

Glyn Moody welcomes your comments.