The promise of XML publishing

ScriptoriumTech / Opinion6 Comments

by David Kelly

Being a technological optimist, I always watch for ideas that provide more, better, faster, cheaper, healthier, and throw in a heap of wow factor while you’re at it. On the other hand, after observing the progress of technological ideas over a few decades, I am beginning to formulate what I call “The Principle of Minimum Adequacy.”

The principle goes like this: Most new technological ideas reach their optimum implementation in five to ten years, and afterwards merely undergo cosmetic changes with few new ideas to transform them. Examples: cars are still a frame with four wheels and an engine; airplanes still have wings, ailerons, and an engine; and the Internet uses http protocol over a network and displays text and graphics in a browser. The basics of these technologies have been in place from the beginning, and little about those basics has changed since their introductions.

This is not to say these technologies don’t have room for improvement. Maglev cars with minimal friction, personal jetpacks, and neurological implants for Internet connection are just around the corner. Plan on it!

The Principle of Minimum Adequacy comes into play because most successful technologies fulfill their functions well to begin with, which is why they are successful. Major changes are not needed. If changes are needed (and possible), people make them rapidly. Occasionally you get the jet engine thirty or forty years after the first flight, and the technology gets a big boost. The remaining changes are mostly refinement, adaption to taste, and marketing hype.

The same principle can be applied to the use of XML for document publishing. Text is structured in tags, style sheets transform the tagged content to format-specific markup, then a rendering device such as a browser or FO processor creates the formatted output. The results are satisfyingly fast, consistent, and reasonably cheap (well, sometimes…). And it has been that way for about twenty years.

But, for me, the adoption of XML for publishing has not yet reached a jet engine type of transformation, something that gives it value far beyond its basic function of getting people (or content) from here to there. It seems to me that one of the great powers of XML is to free information from being merely text on a page, and to give it other kinds of roles. And this is precisely where I see the adoption of XML for publishing to have slowed down, per the Principle of Minimum Adequacy. XML fulfills its role adequately as a vehicle for publishing text, so the information remains text on a page or browser window.

But I do see the occasional example where XML intended for traditional publishing straps on a jetpack and takes on greater value. Here are a few examples or ideas I am aware of:

  • Integration with software and testing tools: XML for a command line syntax not only gets published in requirements and user documentation, but is used to form command line test cases for testing the software.
  • Accessibility to Web 2.0/feedback: Chunks of text are structured in a way that allows the audience to add responses to specific pieces of information.
  • Dynamic publishing: A field troubleshooting procedure includes logic for conditional flows and structures for the user to indicate procedural choices. The text changes based on choices made during execution of the procedure.
  • Automated generation of graphs and flowcharts based on text: Using SVG, well-structured XML provides source for graphics or even animations.
  • Customized, self-documenting software and hardware: Software code includes comments; these comments are harvested by publishing software, and documents are assembled for only the software modules that are present at a given installation.

These are not original ideas – but they are ideas I don’t see discussed often in publication scenarios. I’m looking for reinforcement of my technological optimism here.  I am curious how much people are leveraging their investments in XML, and what kinds of ideas are out there for freeing information from its traditional role as text on a page.

How about you? Have you heard of any interesting applications for XML-based text, or do you have thoughts for unusual applications? I’d love to hear about it.

About the Author


6 Comments on “The promise of XML publishing”

  1. Great post! I think that the growth of systems like the open source eXist native XML database are going to have a huge impact on dynamic publishers. The software is free and XQuery with typeswitch is much easier then XSL. eXist and XRX combined are still in the innovation/early adoption phase except in TEI where it is starting to dominate.

    Looking forward to future posts on the topics of dynamic publishing.

  2. Very interesting post. I agree that “native XML database are going to have a huge impact” not only on publishers, but content in general. We have become such a content consuming society, and the web technologies have had a huge impact on what we have come to expect. The former corporate model of content stored in proprietary formats can’t support these expectations. I see the XML technology reaching far beyond my little technical communcations corner of the world. I think that native XMl databases have the potential to transform our content in the way that relational databases have transformed out data. The tools for this are starting to emerge, but it will also require a mind-shift by everyone who creates content of any kind. Unfortunately, I think this will take more time 🙁

  3. I agree that the problem is generally the speed of cultural uptake rather than the availability of technical means. I’m not sure what can be done about cultural slowness — possibly the example of a persuasive “killer app.” I had hoped to hear of some examples – and XML databases are certainly one kind of example.

    I see the XML database as an enabler that can extend XML solutions into much more complex applications. One application I have seen is the use of an XML database for an XML-compatible component content management system. However, I’m not convinced that the application differentiated itself from standard DB-based CMSs in a compelling way. It’s definitely not “killer app” material, at any rate.

  4. Well, in the realm of killer app, I remember a demo from a years ago that blew me away. It was mentioned in an article called DITA for Kids? ( by Michael Priestley and it takes the concept of dynamic publishing a step further, to the dynamic content creation stage. The killer part would be to extend this concept and be able to pull content from any number of sources, and the ability to pull by copy or by reference. The possibilities for this type of app are endless, especially in the very common corporate practice of cut & paste. This, to me, is so powerful, I can’t for the life of me figure out why it hasn’t hit the mainstream yet. Perhaps a tool like this is the missing link to the XML jetpack -a tool to put the power of XML into the hands of the people – without them ever needing to know that they’re using XML of course. When any web user can do what we as publishers can do with XML, they will drive the requirements for those killer apps.

  5. Great post 🙂 Much food for thought.

    Reg. dynamic publishing, I’ve used two models and both in extremely rudimentary forms.
    (1) Store data in a database, use PHP to query the data and extract the result set, and then transform the result set to XHTML for viewing in a browser window.
    (2) Store content in DITA modules, use an HTML form to create a query from the options selected on the form, and return the DITA module that corresponds to the selected query set.
    Both could’ve been refined further but for case #1, I didn’t have the time to and it served my purpose, and for case #2, I am still trying to improve it so it can incorporate more complicated choicebranches.

  6. So you are leveraging the structure of the content to return information that is important to the user, which improves the value of the content. Sounds pretty worthwhile to me, and much in the direction of what I was thinking. I’m curious, also, whether there might be opportunities to retrieve the data for uses other than reading. At any rate, good luck on your improvements — #2 in particular is something we have talked about needing to do for some of our own customers, so I appreciate hearing what you have done.

Leave a Reply

Your email address will not be published. Required fields are marked *