Chunking information into topics

Documents can be broken down into smaller chunks of content that can be treated as modular building blocks. Chunks should be organised into separate topic files.

With topic-based documentation, the way in which information is broken down and chunked into topics is very important. A user guide with one very long topic organised into sections is poorly chunked. Topics should be small, self-contained, granular units of content. In general, the smaller the topic size, the greater the chance that topic can be re-used in another deliverable document.

The process of chunking involves analysing the components of the information to be documented, identifying the blocks of content that can stand alone as information units, and then splitting those units into separate chunks.

In DITA, standalone chunks can be organised into topics, sections within topics, or topics within topics. You should aim to have separate topics for each chunk, and to avoid sections and composite topics.

Information in a DITA source topic doesn't have to be delivered at the same granularity. For example, a family of DITA topics may be output as a single HTML topic. (The chunking attribute in the topicref element in the ditamap is used for this purpose.)