The folks at DITAToo have been running a survey asking people how they convert their legacy content over to DITA. The preliminary results are interesting if not terribly surprising: the vast majority of users are converting their content over from Unstructured FrameMaker, and after that, from Word:
From the structure of the survey questions it looks as though people can answer to more than one answer, which only makes sense given that Technical Writers often don’t get their material from just one source. Still, I find it interesting to see is how much content is being converted over from FrameMaker – if you include Structured FrameMaker content it jumps up to almost three-quarters of all content being converted over to DITA XML. No wonder Adobe is keen to add significant DITA support to their venerable DTP software, otherwise they would be seeing a mass migration away from their tool.
The number of people who also work with Word content is also significant, and the numbers suggest that there is indeed a market out there for people looking for Word-based DITA editors. While I am guessing that the Technical Writers who are getting their content in Word used to be converting that content over for use within a DTP program like FrameMaker, the size of the response tells me that Word is still being used entirely for Documentation production at some companies.
For what it is worth, my heart goes out to those technical writers who only have PDF content to convert from, which at almost 27% is not an insignificant number. PDFs have to come from somewhere (usually Word or FrameMaker) so what that means is that the original source material is lost and the PDF file is the only thing the technical writers have to work with. Not a pretty situation; and I should know as I have been there myself. This is where having some sort of Content Management System makes a lot of sense, as it becomes much easier to keep track of the original source material along with what it gets converted into.
My only qualm about this survey is that the number of respondents has, so far, not been given. Without knowing how many have taken the survey it is hard to know how representative these numbers are, though in my experience they make sense.
The survey is not over so if you have not already done so you can still add your own experiences on converting legacy content to DITA.