DITA 1.2

By Tony Self

Overview

Almost three years of work of the OASIS DITA Technical Committee has gone into the development of version 1.2 of the DITA standard. DITA 1.2 introduces a number of innovations that open the door to powerful new features in authoring and publishing tools. Amongst the new 1.2 features are keyref (allowing link targets to be specific to the ditamap), extensions to the way in which content references (conref) can be used, glossary and terminology processing, new learning information types, and constrained information types to simplify the authoring learning curve.

The improvements in DITA 1.2 will make new functionality possible, but it will be up to authoring and publishing tool vendors to take advantage of the new standard. Some functionality improvements will arise out of new features, while others are made possible by changes to existing features.

Indirection and the Keyref Feature

Cross-references and other forms of links are quite simple: the link references the location of the target of the link. However, there are cases where the target of the link within a topic will be different depending on the document or publication in which the topic is used. DITA 1.2 introduces indirect linking, where a link in a topic only references a key, and the target of the link is defined in a keys section within the ditamap. In this way, the target of a link can be defined in the ditamap (at the document level) rather than in the topic.

For example, a link in a topic may reference a key named "local_law". When the topic is used in a ditamap for the German market, the "local_law" key could be specified in the ditamap to reference a German government Web site. When the same topic is used in a different ditamap for the Austrian market, the key in that ditamap could reference an Austrian government site.

Changes to the Way Conref Works

The content reference (conref) mechanism in DITA is the main enabler of content re-use. An element of nearly any type can be re-used, or "transcluded", from its source location to another location. DITA 1.2 has extended the possibilities for transclusion by allowing a group of sibling elements (such as a group of steps) to be re-used in one simple process. This change has been implemented by the addition of a conrefend attribute, which specifies the id of the last sibling element in the group to be transcluded.

Another interesting addition in DITA 1.2 is "conref push", where information in one topic can be inserted in another, without the target topic having to be modified at all. For example, a wholesaler may produce topics that are regularly updated and provided to retailers. A retailer wanting to use the topics in their own publication might choose to "push" some customised content into the wholesaler's topics without actually editing them.

Glossary and Terminology Management and Processing

At the heart of the new terminology management features in DITA 1.2 is a new information type of glossentry. Glossary terms and their definitions are defined individually in simple, glossentry topics with rich semantics. As well as the term and definition, the author can define short forms of the term, synonyms, alternative forms, and even parts of speech. The glossary entry topics can be collected together in a ditamap using a new glossref alternative to the standard topicref element.

A new keyref attribute in the term element will make it possible for terms in within the body text of topics to be automatically linked to the corresponding glossary definition during the publishing process.

Learning Specialisation Information Types

One of the interesting new developments is that some specialised information types designed by subcommittees of the DITA Technical Committee are being incorporated into the base DITA standard. The new Learning specialisations, devised by the Learning and Training Subcommittee, provide information structures to suit training materials. Topic types such as learningPlan, learningOverview, LearningContent, learningSummary, and learningAssessment, make it easier to produce learning objects. New elements for interaction semantics, such as those for true/false, multiple choice, and sequencing questions, make it easy to use DITA as assessment content.

The General Task Information Type

One of the criticisms of DITA's content model has been that it does not specifically provide for upper level processes, or informal procedures. That issue has been addressed by the addition of a general task information type, which is much less strict than the task information type (which is now referred to as "strict task").

Other Features

Conclusion

Although it has been a very long time coming, the improvements embodied in DITA 1.2 make an even stronger case for migration to DITA. As software vendors embrace the opportunities to add greater functionality to their products, DITA authors will enjoy a richer authoring environment, and readers will benefit from better information products.