One of the more popular posts on this blog is titled DITA is not the answer and, whilst things are certainly moving forward, it’s a little sad that it is still valid.
A recent comment on that post suggested that it’s not just DITA that is lacking, it’s the working realities of single source that is flawed.
Well, that and the fact that I keep referring to single source when I am actually meaning content reuse (for you can have one source for everything but not reuse the content anywhere).
You can read the full comment yourself but the relevant bits are:
I have never seen single sourcing work. Maybe a single author who knows the topics thoroughly enough to reuse, or a tightly knit group of writers synched up at the same level.
The only place we are going to reuse content is in web mashups using semantic markup once the content is in the cloud.
It’s an interesting view and one which touches on something that has been on my mind these past couple of weeks as we are in mid-migration towards our single source solution.
Just how do you coordinate a team of writers, working in discrete areas of the documentation, with a large number (3000+) topics?
There are a number of ways we are tackling this and only time will tell if they are successful. Firstly we spent some time discussing how best to structure the source topics. Do we group them by product area? By topic type? Or some other arbitrary method?
We decided to group at the highest level (the top level folder) by user persona, and below that we grouped topics in accordance to how they are viewed from the product, so development kit wide ‘Events’ are stored in single folder, where as topics for a specific piece of functionality in the development kit are stored in their own folder. Your system will be different, of course, but this method suits our needs.
After that we need some way of knowing both what type of information a topic contains, as well as where that topic is used. We are not authoring in a DITA specific environment but decided to model our topic types on the DITA model to future proof us as much as possible (we are using Author-it which will export to DITA XML should we need it in the future). We have different templates for each type of topic (Concept, Procedure, Reference and so on), primarily to allow us to identify a topic (by default, Author-it shows which template a topic is using).
That leaves the final piece of our puzzle. How do know where a topic is used? This is more than just a list of which deliverables the topic will appear in, it also has to hint at the context of how the topic is being used.
Does any of this mean that we are more likely to reuse content? Not necessarily but it should give us a fighting chance, and once we’ve updated the content plans for all of our deliverables we will start to really see the benefits. Those content plans were the very things that suggested we could reuse content across multiple deliverables and I’m certain that, with a bit more analysis, we’ll get further gains.
Can single source and content reuse work? Of course it can. There are plenty of good examples out there and they all share one thing in common, something that isn’t really broadcast by the vendors; content reuse from a single source takes a lot of hard work.
But it is possible.