For the past three years, my annual wrap-up of content management systems has mostly counted the exploding number of branded products, for sale and open source, on the world market—now nearing 3,000.

This year, I want to focus on a handful that are doing things so well that they show the way to the future for all the others. One characteristic these market leaders share is their use of XML, whether in native XML databases or simply XML files. These CMSs take advantage of advanced techniques like XQuery and XPath to include fragments of content in multiple places. Content is dynamically assembled, customized, and personalized on-the-fly to enhance the user experience.

Leading websites have managed online content this way for some time. Now more and more content is “single-sourced,” meaning that it feeds not only the web, but traditional print materials like advertising, market collateral, and documentation: multichannel publishing and in multiple formats.

XML is migrating back to the beginning of the content-creation process. Authors are being asked to write modular content components that can be assembled in many different ways by the dynamic publishing engines. Many decades ago Ted Nelson, who named hypertext, imagined that bits of text could be “transcluded,” pulled in from their original source documents to assemble new documents, complete with a link back to that single source. Now the most sophisticated publishing systems can identify every content component and connect it back to its source, back to an author who could be pictured as working on a writers’ assembly line.

Putting aside visions of Charlie Chaplin’s Modern Times, with writers gumming up the workflow, the metaphor seems more than apt. Content is now considered components—like interchangeable parts that can be plugged into a machine, or electronic parts that can be plugged together in a system. Content is modular—bits of text, images, audio and video media, Flash animations, 3D objects, and complex learning objects with built-in testing. And all these malleable pieces have XML with metadata in common.

How do components get assembled? Imagine an owner’s manual for your new car personalized with your exact options. Then imagine the mechanic who enters your VIN and prints out the customized service manual for your particular car. That is single-sourced modular content.

A couple of decades ago, Vannevar Bush invented the Memex, “a device in which an individual stores all his books, records, and communications, and which is mechanized so that it may be consulted with exceeding speed and flexibility.” Bush imagined we would link pieces of content with “associative indexing…whereby any item may be caused at will to select immediately and automatically another.” It seems to me that folksonomic tagging of content components is getting us very close to the Memex.

So rather than writing long narratives, our assembly line writers now author stand-alone topics that can be assembled by a component CMS, but also post-tagged by us someday to form our own personal knowledgebases.

I see state-of-the-art content creation tools greatly simplifying the task of creating these reusable single-source objects, because tools now create content in a general standard like XML, or even the more particular Darwin Information Typing Architecture (DITA) version created by IBM. The tools themselves work with a single-source component assembly model.

Highly integrated tools like Adobe Creative Suite 3 Production Premium for audio and video media and the new Technical Communication Suite allow users to create a content component in one tool, then insert it in another tool while preserving the original source. For example, eLearning content created in Captivate 3, user assistance created in RoboHelp 7, or a 3D animation can be assembled into FrameMaker 8. If the original is changed, it is updated in the assembling tool. Content can even be published with Acrobat 8 to a PDF for viewing anywhere and still preserve backward connections to the original source.

Will creative writers and rich-media designers rebel at this image of themselves feeding a giant assembly line of content? I don’t think so, once they have tried these beautiful tool suites and see how they integrate all the way from creation to consumption of content.

Maybe a better metaphor is that they are members of an orchestra or a film crew, all contributing to a dynamic performance in which we consumers are sometimes the audience, and sometimes the directors, in our own content creations.