Inside DITA: Topic-Based Authoring Without the Ceremony

By Erich Champion

Most books about DITA explain the vocabulary. This one is written in it, so every technique it describes can be read in its own source - and where a chapter claims something works, the book itself is the evidence.

What You'll Learn

Topics and maps come first: what makes a piece of content movable, why being self-contained matters more than being short, and how separating the order of a book from its content lets one source produce a printed edition and a help system with neither authored twice.

Then reuse, and the ways it fails. Content references, keys and indirect addressing, conditional processing with DITAVAL - alongside the type-matching rule that produces a misleading error, the filter that silently breaks every reference to a pool, and the reason a reusable sentence fragment that works in English breaks in German.

Extending the language gets a part of its own, worked through a specialization small enough to read in a sitting, covering domains, constraints and generalization. That part also argues that most projects should leave the mechanism alone, and says why.

Publishing gets the most space, because it is where projects stall. Installing and running DITA Open Toolkit, the arguments that cover nearly every build, project files, PDF through Apache FOP, HTML5 and EPUB and Markdown, and how to write a plugin that survives an upgrade. The chapter on print customisation walks the plugin that produced this book's own PDF, including its numbering override and the reinstall step that makes people think their edit did nothing.

The last two parts are about the decisions around the language rather than the language: Lightweight DITA in its three syntaxes, letting contributors write Markdown without giving up structured publishing, adopting DITA without buying anything, the point at which a content management system stops being optional, and where DITA 2.0 actually stands - which is not released, whatever you have read.

Who This Book Is For

Technical writers evaluating DITA who want the costs stated as plainly as the benefits. Documentation teams already committed to it and stuck on publishing. Engineers who have inherited a DITA toolchain and need to understand what it is doing. And anyone who has been quoted a six-figure figure for structured authoring and wants to know what the free parts actually cover.

It assumes you can read XML. It assumes no schema knowledge, no XSLT, and no prior contact with the standard.

What Makes This Book Different

The book is a DITA project, and that constraint was enforced rather than claimed. Its chapters are information-typed, so procedures are tasks with real steps and catalogues are references built from tables. No chapter title contains a chapter number, because numbering is presentation and the stylesheet counts them. Every cross-reference is a key reference, so the toolkit writes the link text and the page number. Every version and date the book cites lives in one file.

It also records where the argument broke down. A chapter once claimed the build published the book two ways when neither generator applied a filter; an audit found thirty-six citations naming specification sections nobody had read; the prose violated the project's own style guide twenty-seven times in one pattern. All of it is fixed, all of it is now checked by a script, and the last chapter describes the episode rather than hiding it. A worked example that shows only the parts that went well teaches very little.

• Format: Paperback
• Language: English
• ISBN: B0HDN1WBPP
• ISBN13: 9798191838342
• Release Date: August 2026
• Publisher: Independently published
• Length: 159 Pages
• Weight: 0.85 lbs.
• Dimensions: 11.0" x 0.4" x 8.5"