DITA- CMS From OASIS and IBM
Darwin Information Typing Architecture Content & Document Management
Mar 11, 2009
The Darwin Information Typing Architecture was contributed, along with some tools, to the Open Source community, by IBM. It is a way of classifying information, storing it, and then publishing it in a variety of formats without changing the source. This separation of information from presentation makes it ideal for CMS, document systems, web authoring, and a whole host of other applications.
The architecture is also strongly validated against XML DTD (XSD) representations, during the authoring process. This way it is impossible for authors to create a document that does not follow the principles laid down by the DITA experts within the organization. In this respect, as well as various other aspects (like information typing, and individual content units), DITA shares a lot with a predecessors such as IMAP.
DITA vs Information Mapping (IMAP)
The DITA approach to chunking documents and creating a pool of material that can then be used to build up publications is very similar. Whereas Information Mapping has a number of different types of information (infotypes), DITA starts out with just three:
- Concept - explanation oriented
- Task - process oriented
- Reference - fact oriented information
From these three, DITA architects can then specialize to create a number of different infotypes depending on the requirements of the target publishing platform. The DITA Open Toolkit starts out with around 300 objects and attributes that are described using XML. This is a great benefit for information architects and document publishers who want to jump into DITA as quickly as possible.
The Open Source approach of DITA is one of the key differences to IMAP, which is a method protected by copyright. The consequence of this is that while there are a limited number of tools available to manage IMAP projects (all of them reasonable quality), there are a vast number of tools created by a lively DITA community. Not all the Open Source DITA tools match the quality of the DITA Open Toolkit, but this alone gives a competent individual more than enough to get going.
Implementing DITA
DITA can be implemented in one of two ways:
- Slowly, evolving over time
- From the ground up
For those building a new document repository, the ground up approach will mean understanding what information they want to be able to publish, and how it is defined, before being able to implement enough XML to apply the DITA architecture to do something useful. Those adapting an existing document set can choose one part of it to become familiar with DITA before moving the rest to the new information management platform.
At the very least, users will need some kind of DITA editor, or authoring tool, something to manage the collection of publications - a CMS, for example - as well as something to publish the end result. The DITA Open Toolkit is one option for this last, and it is fully anticipated that there will be a DITA aware version of the Open Office word processor in the near future.
DITA for Publishing
The DITA architecture allows the user to separate the information (stored as Topics) from the presentation. This means that the same underlying Topic data can be used to generate all kinds of output documents - PDF, HTML, Help, and so on.
The key, as with IMAP, is in making the correct choices at the time that the information architecture is established, so that each Topic type is correctly specified. From there, it is a case of building the correct containers for the Topics, selecting the right instances of the Topics, and finally generating the output.
The appearance of the output is not governed by the DITA information, but by the style information that is applied to it - and that's the beauty of the DITA method of publishing.
