Confronting the future of Web standards

As technology consumers, we accept that the Web is built on open standards. We are told that an entire standards body, the W3C, exists to maintain them. Dig a little further, however, and we find that the history and the issues behind the standards are far more complex than they appear to be. As any long-suffering Web developer can tell you, the process of drafting W3C standards is as slow and laborious as any i

As technology consumers, we accept that the Web is built on open standards. We are told that an entire standards body, the W3C, exists to maintain them. Dig a little further, however, and we find that the history and the issues behind the standards are far more complex than they appear to be.

As any long-suffering Web developer can tell you, the process of drafting W3C standards is as slow and laborious as any in the industry. As a result, standardization alone can't account for the rapid growth of content and applications that we've seen on the Web -- and it won't be the only force guiding its future.

At the recent Google I/O developer conference in San Francisco, Alex Russell, the project lead for the Dojo Toolkit, gave a fascinating presentation on the potential future of the browser. According to Russell, if you plot the market share of the various Web browsers over the last 15 years, the resulting graph shows a clear cycle of natural monopolies.

A single browser -- first Netscape, then Internet Explorer -- gained market share until it became the clear leader, commanding in excess of 80 percent of the market. Only after it had achieved peak saturation did the leader's share begin trickling off, making way for the next natural monopoly.

Given this data, you could argue that the success of the Web owes less to the fact that HTML is an open standard than to the fact that a few major vendors -- primarily Microsoft and Netscape -- made HTML and its related languages (JavaScript and CSS) into de facto standards. Ubiquity gave those vendors the power to push the features that they wanted while the W3C lumbered along at its usual pace; the rise of XMLHttpRequest is but one example.

How important is this process? "De facto standards are the only ones that matter," writes Sun Microsystems CEO Jonathan Schwartz in a recent blog post. "Well-intentioned standards bodies and departments of justice can do their best, but at the end of the day, volume deployment is the only setter of standards. Ubiquity trumps policy, just about every time."

Of course, de facto standards don't have to be proprietary; in fact, Schwartz paints an optimistic picture. As computing spreads within emerging markets such as Brazil, Russia, India, China, and Africa, he argues, open standards and free software will be what makes it affordable.

As Schwartz writes, "Where is OpenOffice.org deployed in the greatest numbers? In places where saving $300 per desktop is meaningful." If you take this view, as emerging markets become an increasingly important customer base for the technology industry, open standards such as the ODF (Open Document Format) used by OpenOffice.org will almost certainly become de facto standards, as well.

But it is the OpenOffice.org software that makes ODF viable, not the other way around. Similarly, as the Web continues to evolve by leaps and bounds, ubiquity will continue to be the foremost driver of Web standards.

Coding with plain HTML and JavaScript is painful for developers of RIAs (rich Internet applications). As RIAs proliferate -- both in the browser and beyond -- competitive pressures force developers to adopt alternative technologies, such as Adobe's Flash/Flex and Microsoft's Silverlight. AJAX is another alternative, but remember that even XMLHttpRequest has yet to be finalized as a W3C standard.

It seems that for Web 2.0, in the absence of clear direction from the public standards bodies, we're faced with a choice between several de facto standards. It's a situation we've seen before on the Web. The question that remains is what we should do about it -- if, indeed, we should do anything at all.

Maybe the lesson we should take away from the last 15 years is that Web browsers are simply prone to natural monopolies. As Russell asked in his Google I/O presentation, "How would our attitudes toward them change if we accepted that they were?"

Or, more broadly, how would our attitude toward monopolies change if we accepted that natural monopolies, rather than open standards, might sometimes be the best way to push technology markets forward?