The Last Page
XML's Achilles Heel
By Dale Dougherty, Editorial Director
While in San Jose at the XTech '99 Conference, I asked IBM's Marie Wieck if customer interest was driving the firm's involvement with XML. She said yes and sighed. "You should see my travel schedule. Last week I was in Munich with customers. Tomorrow I'm in San Diego," she said. "Industry acceptance of XML is faster and broader than Java," she added. Why? "Because everyone has data," she replied. And everyone wants to understand what XML means for their data.
The developers of XML are fond of saying that XML is just syntax. Jon Bosak, chair of the XML Working Group at the W3C, made this point at XTech '99. "XML is just like ASCII," he said. "We don't have ASCII conferences. Maybe in a few years we won't need XML conferences because XML will be so commonplace." Still, people come to these conferences trying to learn what XML is, and if they can fight through the hype, they can also learn what XML isn't.
XML standardizes syntax but goes no further. XML establishes a way to define tags and use them in documents, but it doesn't help you establish what these tags mean. XML doesn't standardize semantics. It doesn't tell you what the contents of a DATE element mean: date of publication, date due, date modified, and so on. In fact, XML doesn't tell you whether the contents of DATE are anything but a string. It's a common misconception that XML solves this problem for data interchange, when in fact XML neatly sidesteps it.<>
