Dogfood
So I’ve been thinking. CollabNet (where I work) is all about tools for distributed development. (Mostly for software, but there are customers in biotech and manufacturing too.) Its main product includes a database component called Project Tracker (PT). The essential concept behind PT is the artifact, which is a generic work item that the user can configure to represent any measurable part of a project, such as an action item, requirement, trouble ticket, test plan, use case, or whatever. The idea is that you can build your own project tracking system out of artifacts that you design to match your very specific needs, while PT handles the back end for you — stuff like change tracking, reporting, and so on. We also provide some basic pre-designed artifacts out of the box, which many customers use until their needs get complex enough to require their own custom artifacts.
What I’m thinking is, what if there were an artifact that represented a DITA topic? This artifact would contain the actual content of the topic, plus the reviewing trail, plus the metadata that would enable you to assemble topics into outputs by running a query. A writer wouldn’t have to know any DITA markup to play in this game — they would just have to fill in fields, such as Shortdesc, Context, Step 1, Step 2, etc., which would be converted to DITA markup using the XML export functionality that PT has in spades. I think there is some precedent for a web-form-based model of topic drafting, although I don’t have any experience with it. This new XDocs thing, for example.
I say “just,” but of course I don’t mean that; there’s still a lot of skill involved, but its exercise is moved back from the text creation and tagging phase to the conceptual phase. Writers could spend most of their time and brainpower designing the document and defining the topics required. Once that’s done, the actual drafting might seem like an afterthought.
Categorised as: Geekery