Skip to main content
Working with XML

Recognizing that XML offers solutions to Web site architecture problems does not by itself answer questions of how to work with XML or whether the use of XML is even realistic today. We all live in the real world, and when dealing with Web sites, need to deal with real issues of multiple browsers and display devices and connections. For today's world (2002), this means that much of the XML processing must be handled on the Web server, not on the desktop or client device. The server "translates" the XML files and sends the appropriate HTML pages to the client browser.

Although newer web browsers, such as Internet Explorer 6 and Netscape 7, offer some XML/XSL support, the consistency and depth of support for the various XML specifications within client browsers is unpredictable and unreliable. Therefore, we need to perform our XML processing on the server. In addition, our Web site, like most Web sites today, also has database and programming elements so we wanted to integrate these elements within the XML structure. For these reasons we adopted the Apache Cocoon XML publishing framework for developing and delivering our XML files. (For more information on Cocoon, which is an open-source, server-side environment for XML processing that supports full integration of database and programming logic, visit

On the development side, we encountered a similar situation in that XML software tools are rather primitive compared to HTML tools. Some XML editors, such as XML Spy ( offer useful tools for developers, but are not really designed for other users such as content owners and subject matter experts. In other words, a Web developer could use it to create XML and XSL files, but content owners could not easily use it now for editing their original document and saving it as XML. However, these software packages are heading in that direction and becoming more user-friendly with time.