XML: THE NEXT MICROSOFT MONOPOLY?

So far, the noisy debate over Microsoft's growing control of the Internet has focused almost entirely on whether Netscape will be driven out of the browser market. But there's another nearly invisible standards question that could be far more important: Who will end up writing the rulebook for XML?

What's XML and why does it matter? XML ("Extensible Markup Language") can be a tricky concept to describe, but essentially it's a meta-language that provides a way of defining content tags for text. For instance, an XML tag might indicate that a text item represents a product price or an SKU number. Unlike universal HTML tags (which control text formatting), however, XML tags are defined by individual applications. Thus, XML can provide a way for an accounting program to understand data that comes from a Web-based EDI system, but only if both applications share the same tag definitions and XML parser.

Common content tags are a tremendously powerful concept, which--at least in theory--makes it possible for all kinds of hardware systems and applications to talk to each other, regardless of their native file formats. IBM, which supports a bewildering collection of development platforms, middleware, and database environments, is rushing to make XML a company-wide standard. And Adobe, Netscape, Sun, and Microsoft all have XML projects under way in their R&D labs.

But it's Microsoft that looks like the real front-runner in shaping XML standards. The next version of Office will use a hybrid HTML/XML markup language as a "companion" file format for all suite components. (Office will also include a new Vector Markup Language that will make high-quality graphics platform-independent.) Office users will be able to create richly-formatted documents in Word or Excel, which can be opened or even modified with an XML-enabled browser and then sent "round trip" back to Office. As long as a parser is built in, XML documents should render correctly on the Web, in print, on CD-ROM, or even on display platforms that haven't been invented yet.

Moreover, it's possible to write fairly complex XML scripts that drive applications, modify databases, or interact with users. Java developers are already talking about how beautifully XML and Java complement each other, especially for serious Web-based commerce and content applications.

To be sure, these are all future considerations. The World Wide Web Consortium only got around to endorsing XML in February, and no browser currently includes an XML parser. Individual industry groups--including chemists, mathematicians, and graphic artists--have proposed their own specific XML extensions, but so far Microsoft, Netscape, and other Internet heavyweights are still arguing over some key XML fundamentals.

Which means, for better or worse, that Microsoft almost certainly will end up defining de facto standards based on the XML extensions that Office, Internet Explorer, and Microsoft's Web authoring tools will all support. By default, the vast majority of XML-based documents will rely on these extensions, and we suspect users will--as usual--welcome the consistency and predictability that Microsoft standards usually provide.

But it's also worth remembering that Microsoft tends to treat language standards as an opportunity for political leverage. Just as Microsoft "embraced and extended" Java to make its own version run better on Windows machines (and worse on other platforms), we expect that Office-standard XML will quickly evolve into yet another semi-proprietary Microsoft environment. Mysterious glitches will plague users of competing non-Microsoft applications like Navigator and Notes, and in the end XML will become just one more cross-platform standard that ironically reinforces the market dominance of Windows and Office.