From the Director
DITA and Localization: Improving your cost savings
In September, I spoke at the First XLIFF Symposium in Limerick, Ireland. My purpose was to introduce DITA once again to members of the localization community and to demonstrate, with Bryan Schnabel and Rodolfo Raya, the value of marrying DITA with another OASIS standard, XLIFF. XLIFF stands for XML Localization Interchange File Format, providing an outstanding way to package DITA maps for translation.
Here is OASIS XLIFF Technical Committee chair, Bryan Schnabel’s description of the presentation:
“In Limerick Ireland, at the First International XLIFF Symposium, the larger-than-expected audience witnessed a very interesting, and telling presentation, “DITA and XLIFF, a Perfect Marriage.” Peter Reynolds introduced the audience to the concept of the interoperability between DITA and XLIFF; JoAnn Hackos began the presentation with an introduction to the use case, and talked about the fundamentals of the standards. Bryan Schnabel then demonstrated an open source software solution, including his DITA/XLIFF Roundtrip tool that showed how a complete work cycle, starting with a source language DITA project with many topics, could be packaged and translated via XLIFF. He then re-composed the XLIFF file back into the translated DITA project. Rodolfo Raya then used the same DITA project to demonstrate the same concept, using a commercial tool and highlighting the stresses of a real-world operation.”
I was surprised and pleased to learn at the conference that there is at least one professional degree program for localization professionals in the US. Europe has long had professional programs, in part because many of the “English” degrees at European universities are translation-oriented rather than literature-oriented. I also taught workshops for several years in the outstanding masters’ degree program offered by the linguists at the Copenhagen Business School. In Limerick, I met Keiran Dunne, who teaches in the localization degree program at Kent State University through the Institute for Applied Linguistics. He informed me that their graduates are in great demand, especially their PhD students who are sought after to staff emerging degree programs at other US universities.
Information-development managers now have a resource of people educated in managing and administering localization programs for US corporations. The complexity of these programs and the high costs associated with localization are best served by people who understand the challenges and pitfalls to prepare for. Without training and experience, it is too easy for an organization to pay far more for localization services than need be.
I wrote a chapter on Translation and Localization Management for my 2006 book, Information Development: Managing your Projects, Portfolio, and People. Invaluable was the help I received from respected localization coordinators at several companies. They provided me with insights into the care that coordinators must take in managing their processes. Serving as chair of the DITA Translation Subcommittee also provided an opportunity to learn from many of the leading theorists and practitioners in the field. It’s been a great education. The articles published by the DITA Translation Subcommittee are available on dita.xml.org, in the Knowledge Base Wiki.
Working with the XLIFF transformations has presented another opportunity. As Bryan Schnabel explains in his statement about the conference, we can transform a multitude of DITA files, organized as a DITA map, into a single XLIFF file. That process not only saves the administrative costs of handling hundreds or thousands of separate files, it also accounts for such DITA-specific features as conrefs (content references to externally maintained data in the forms of words, phrases, or other reused content strings). Most translation tools do not handle conrefs effectively.
Since XLIFF understands DITA, we are able to adjust the source content to ignore non-translatable elements like <draft-comment> and to add the XML text embedded in illustrations saved as SVG (Scalable Vector Graphics) to the translation package. Once the XLIFF file is translated, we roundtrip the file back into the topics, conrefs, and images that are referenced by the DITA map.
If you have not had an opportunity to see exactly how XLIFF and DITA work together, I recommend reviewing the recording of the webinars that Bryan, Rodolfo, and I gave in 2008.
OASIS hosted two webinars on DITA and the XML Localisation Interchange File Format (XLIFF).
- ‘DITA to XLIFF and Back—Reducing the Costs and Risks of Translating XML Content’ will present a business perspective on translating topic-oriented, XML-based content. The webinar will include business cases developed by expert panelists.
- ‘DITA to XLIFF and Back—Understanding the Technical Solution’ describes in more technical terms how DITA can be used with XLIFF, which provides an easy-to-use mechanism for packaging DITA files for translation and localization providers. A real world ‘end-to-end’ enterprise implementation will be demonstrated.
For a recording of either of these events, please go to <http://www.oasis-open.org/events/webinars/>
I also recommend that you introduce your localization coordinators to XLIFF. In one of the conference presentations, the localization expert from Medtronic reported that all their files are sent to their LSPs using XLIFF. We heard success stories from Oracle and Microsoft using similar mechanisms. We are always looking for opportunities to reduce translation and localization costs. XLIFF and DITA, working together, provide one of the best mechanisms currently available.
Finally, consider discussing XLIFF with your Content Management System (CMS) provider. At least one Component CMS that we know of, IXIASOFT, already incorporates XLIFF into its translation package. We hope to persuade others to do so soon, with your assistance. If you’d like to learn more, don’t hesitate to send me a note to follow up with more detailed information.