Monday, March 22, 2010

An Update on DITA Features, Tools, and Best Practices

Tony Self


Tools include Arbortext, FrameMaker, xMetaL, oXygen, and others, but not the DITA Open Toolkit.

Content Management includes Bluestream XDocs, SDL Trisoft, Vasont, and others, but not Author-IT.

Publishing tools include MadCap Flare, Adobe RoboHelp, Antenna House, DITA Open Toolkit, WebWorks ePublisher, and others.

For DITA and HATs, "DITA support' means different things to different vendors. Not clear what "full support" means. Does it mean you can export? Import? Edit in a native DITA environment? None are DITA editors. Some are good DITA publishing tools.

"Haphazard" HATs provide capability to do hybrid publishing. For example, with Adobe Technical Communication Suite, have FrameMaker/RoboHelp integration, and you can import DITA and export to WebHelp.

Delivery formats, the changing landscape

DITA is a storage and authoring format, not a delivery format. DITA Open Toolkit can turn DITA into HTML, HTML Help. Eclipse Help. Commercial tools can turn DITA into WebHelp and just about anything else.

In theory, you can do anything with DITA content. If the DITA Open Toolkit doesn't do what you want, in theory, you can program it to do anything else. (Not easy for a lot of people; you may have to write your own DITA tools.) With "tinkering," using open-source plug-ins you can produce HTML Help, WebHelp, AIR Help.

There's more to life than traditional help formats. Help content may be delivered on phones, tablets, eBooks, GPS displays, heads-up displays, and more. A DITA transform exists for the EPUB (electronic book) format.

DITA Open Toolkit

Many tools use the DITA Open Toolkit (DOT) for publishing. No tools do help better than DOT (with plug-ins). Generic output is "bland," but you can specify CSS, scripting, and more.

With the DOT, DITA to "tri-pane" help is possible. With a bit more effort, you can make it better looking. With a lot of effort, you can get some level on context sensitivity. But pop-ups, drop-downs, and so on are not possible (yet).

DITA 1.2

Coming soon, with more improvements useful for technical communicators. (Remember, DITA is a standard, not a tool.)

Major changes include keyref, improvements in the content reference (conref) mechanisms, glossary and terminology, and learning specifications.

Keyref supports a concept called indirection, which is linking. You define key values in a ditamap, and then use a keyref attribute to redirect links to targets, which is useful for, for example, translation. A flexible way of linking, so it creates new ways to set up and use contextual linking. Removes context from the topic to the ditamap. When the topic is reused, you can have different links.

Conref push forces an element into another topic.

For glossaries, terms can be associated with definitions, and definitions include parts of speech, short forms, and alternative forms. They can result in automatic acronym expansion.

Subject schemes allow you to label items in the ditamap to classify items by subject and relationship.  It's a full taxonomy of information. It's also useful for dynamic custom content.

Learning specializations will offer specialized information types for different industries, and they will be "packaged" with the DITA standard.  The Help Subcommittee is working to provide user assistance specialization for DITA 1.3.

Summary tip: trust the separation of content and form.

You will need to use more than one too. You may need specialized assisstance, such as programmers and graphic designers.

Don't try to abuse the semantics of DITA.

1 comment:

  1. Chuck, thank you for putting up the session summaries. They are very valuable and (short of session podcasts) the next best thing to being there! I find them much more valuable than session tweets which I've seen for other conferences. Hope you enjoy the rest of the conference!