Today was the last “official” day of the conference. Tomorrow (Friday) is a post-conference day of workshops. Unfortunately, I need to leave early Friday, so I’ll miss the workshops.
Today started with 3 keynote talks. The first was from Joshua Duhl of Quark and titled, Once Content is in XML, Now What? He introduced the idea of dynamic publishing. While the talk was mostly a pitch for Quark, it was interesting because Quark is not normally thought of as an XML company. I think it’s encouraging that companies that have mostly been in other parts of the technical documentation world are seriously working with XML.
The second keynote was Document Engineering in User Experience Design, by Robert Glushko of the University of California, Berkeley. Bob has a new book titled, Document Engineering: Analyzing and Designing Documents for Business Informatics and Web Services.
He spoke about the idea of Document Engineering as a methodology that is a synthesis of information and system analysis, and business process engineering. In an e-commerce environment, customer satisfaction is more than the User Interface. We tend to think of the customer experience as being a “front stage” function; that is, the immediately visible parts of a system. But, in fact, the “back stage” functions, the plumbing that makes everything happen matters as much to customer satisfaction.
Glushko says that we need designs that cut across back and front stage operations. He described the “Moment of Truth,” which is the point where a problem or other critical event becomes clear to the customer. Moments of Truth are often the intrusion of a back stage problem (supply chain problems or other failures in plumbing) onto the front stage. The design needs to encompass both the front and back stage to be fully successful.
The third keynote was Social Media 101, presented by Darren Barefoot of Capulet Communications. Darren was probably the most skillful presenter I saw this week in the use of presentation materials. He used very few words; large portions of the talk were accompanied by pictures that alluded to what he was discussing, but rarely, if ever, repeated what he was saying.
He discussed characteristics of social media, which include conversation, collaboration, sharing, broadening of scope, community, transparency, and authenticity. To me the critical point is that all manner of web communications have become multi-way. There are few areas where a pure, one way broadcast of information makes sense. For example, users are now part of the product support cycle, having input into every part of the support cycle.
The second critical message is that resistance is not only futile, it’s counterproductive. Companies will gain more by giving up some control and taking full advantage of social media. This means empowering your most passionate users, giving users the tools they need to help each other, and going to the places your users go. If they use youtube, you need to be there; if they use twitter, go there.
The first regular session of the day I attended was DocBook vs DITA: Will the Real Standard Please Stand Up, presented by Teresa Mulvihill. Since I’m on the DocBook TC, I was particularly interested in this talk. In general, I think Mulvihill’s views are close to mine (see my article titled, Choosing an XML Schema: DocBook or DITA). However, as someone who has worked with customers on both DocBook and DITA, she sees more difference between the two tool chains than I’ve seen (I’ve done extensive work with DocBook, but much less with DITA, so I don’t have her perspective). She strongly favors DocBook in smaller, turnkey environments; a sentiment I share, though possibly not as strongly.
We were fortunate to have people who use both DocBook and DITA in the audience and had a good discussion after the formal talk.
The next talk was a short talk by David Ashton, of SDL (the company that offers the TRADOS translation support tool), titled 24 Ways to Shut Down the Application and other Apocryphal Stories. He gave a nice high level introduction to some of the trials and tribulations of translation, tying it into the publication process.
I then snuck into the second half of Joe Gollner’s last talk of the conference, Extreme Content Makeover: Migrating Content to DITA. While the talk focused on DITA, the basic principles would apply to any data conversion. In addition to the expected exhortations to plan carefully, Joe covered some less obvious, but important points, including:
- Establish Control Collections, which means to look for groupings of files that have similar features, and therefore are likely to need the same kinds of conversions.
- Define the target end state, which goes beyond simply selecting a target schema; that’s just the bare minimum. You also need to consider linking, metadata, and other details of the target.
- Prepare a conversion specification, which defines the end state, naming conventions, mappings, and so forth.
- Establish a representative example set, which is a group of files that cover the range of the control collections and gives you a set of content for testing.
- Build a conversion plan, which is a project plan for the conversion itself.
Overall, Gollner provided an excellent overview of what it takes to do a successful conversion, whether it be a conversion to DITA, or a conversion to some other XML format.
The last regular session I saw was Using Task Modeler to Streamline DITA Content Development, by Mark Wallis of IBM. This was a demonstration of IBM’s Task Modeler. In addition to demonstrating the software, Wallis described the methodology his team uses to structure content. They design “Task Support Clusters,” which provide conceptual, task, and reference information as a self-contained cluster of DITA topics. They use a minimalist approach, but try to keep each cluster independent, to the extent of avoiding links outside the cluster.
The software helps with the mechanics of setting up a cluster by helping authors visually build ditamaps, develop initial topics, build relation tables, and build a skeleton cluster. The interface looked straightforward, and I’ll be trying it out.
The closing keynote was titled Living Multiple Lives: The New Technical Communicator, and was presented by Noz Urbina. He talked about findings from techdoc evaluations his company has done for several companies, but the main focus was on distilling those ideas into suggestions on “What to do on Monday”? In other words, what can participants do when they get back to work next week to improve their environments.
He did what a closing keynote speaker should do; he summed up the main points of the conference, which I take as the following:
- Moving to XML can gain significant productivity. He cites the potential for 15-30% reductions in localization costs, and another 15-30% savings from reuse.
- You don’t get these benefits by buying some tool; in fact, selecting tools should be one of the last things you do.
- You need to understand your needs and use cases first.
- The job of technical communicators is rapidly changing, mostly for the better. But, to take advantage of those changes, communicators need break out of the “manual box,” establish a business-driven strategy, and work closely with other parts of their organizations.
- Agile methodologies can be a good thing for communicators because they pull communicators into the process early on, and the methodology takes documentation seriously.
Overall, I found the conference to be useful, both through the talks and through interactions with other attendees. There was a manageable number of people, and I had the good fortune to speak with many of them, including several of the speakers. There was a casual atmosphere that encouraged discussion with the speakers and other attendees. The facility was excellent and the logistics were handled cleanly and efficiently. Overall an excellent conference.