Excellent metamodel news

DITA as contributed to OASIS in 2004 actually had an implicit metamodel, parts of which were described in the original developerWorks articles of May 2001, and others in elevator notes and intro presentations given during the first several years (including the original OASIS Deep Dive series by Michael Priestley). I'm now pulling those various notes and presentations into a more concise story of the architectural rationale in DITA's design for generality and extensibility. I'll ask Michael Priestley to work with me so that we can get some of that historical context, originally sequestered in IBM mail logs and databases, back into the public discussion.

The original design team worked with many factors and influences: emerging business patterns, XML and content usage in the Web context, promotion of best features of SGML lessons learned, considerations of application architectures and the role of top-down and bottom-up design principles, role of metadata, communities of use, standard stylesheet and processing behaviors, root structures and content models, encapsulation of writing best practices, and more. My aims for this retrospective review of the base metamodel are 1) to clarify how it aligns with other metamodels of the day, including those in Information Mapping, OpenDoc, IBMIDDoc, and others, and 2) to make it easier for Subcommittees and implementers to create conforming extensions for community-specific requirements as they become known. I'd also like to write about the design process itself and honor some of the previously unsung contributors, but that'll come later.

I'm sure everyone wants that document asap, and I'm giving it extra focus this week. I might have to get it out in installments since some of the internal provenance may take some time to find and get cleared.

dita-busdocs message