XML Workflow and the Holy Grail

Michael Jensen | 03/25/2009 | Digitization

At publishing conferences, there has been of late a lot of pressure to shift to an “XML workflow,” pressure mostly exerted by futurists, technologists, and geeks. They imply that a publisher should be embarrassed if they don’t have XML from the very beginning of the publishing process. Usually, they’re also trying to sell XML editorial systems of one kind or another.

Well, as both a futurist, a technologist, and a sometime geek, let me take a few minutes to differ on that score. There are places where XML workflow makes a lot of sense—journals, or newspapers, or other “throughput” publishers. But for book publishers, there may be a different equation.

A few hundred dollars to an offshore vendor can transform a final PDF of a book into an archival-quality TEI-Lite XML form, an .epub ebook format, and almost any other format you’d want. One has to stop and do some math, before going too far down the path of revolutionizing your workflow processes.

At the National Academies Press, we publish about 180 books a year. While we have been revisiting this “XML workflow” question every 18 months or so for about five years, we have yet to come to a conclusion that it makes sense. We currently send our PDFs to an offshore service upon release, and in a few weeks receive back several versions of the book in XML (and HTML) formats.

The full costs of changing to “XML workflow” include the cost of training (not just compositors, but editors) in a new process and workflow; the cost of transformation and disruption (it’s not a switch that can be clicked, but rather a fundamental process change); the cost of frustration (which is never free); not to mention the substantial cost of new software.

The touted benefits of a full XML workflow generally fall into the category of “better chunks” (being able to sell chapters, repurpose content, or license items, more easily); of getting editorial participation via mark-up early in the process; of being able to have quasi-composition ahead of final composition; of intersecting with “content management systems” that allow easier handling of many documents.

But unless you’re a book publisher who produces more than a 300 books a year, or who have a real market in realtime-licensing of content upon release, or who can justify the cost and disruption in other ways, then it’s probably best to just wait for awhile – wait for the XML workflow to make sense financially, and wait a bit for the XML and the epub version of the books.

Two to three weeks of waiting, and a few hundred dollars per title, is not a dramatic penalty to pay for avoiding the pain of a learning curve whose benefits may be small.

Comments 0

RSS Feed