Unstructured single-sourcing
Discussion started by Bill Swallow , on 23 June 09:57 AM

Hey everyone. Just taking a quick look at practices.

How many of you are single-sourcing with unstructured source content?

What tools are you using for it? What efficiencies have you gained, and what inefficiencies have you discovered?

How do you single-source (multiple audiences/content variations, multiple deliverables, a bit of all the above, other)?

Do you see yourself/your team moving to structured content any time soon? If so, why?

Replies
You will need to be a member of this group first before you can post a reply.
Carla Martinek
Lots of questions in one post, Bill!

Single-sourcing in FrameMaker 10 (just upgraded). Current output is printed pages, and PDF docs intended for printing by end-user.

We use topical files (if it has a Level 1 heading, it starts a new file). Lots of conditional text as content is shared across multiple product lines. Conditional text is used at the SENTENCE level, not individual or small groups of words. For translation purposes, this is the minimum string size to conditionalize.

Efficiencies: turnaround time is vastly reduced. One writer is maintaining User and Quick Ref guides for 7 or 8 product lines, plus OEM versions as well. Replacement kit instructions are incorporated directly into the Maintenance Manual book (previously, info was copied and reused). Less "drift" of content across multiple versions of product documentation with the "update once, use many" method. Reduced editing time. Better consistency across all product documentation.

Translation time is reduced (and correspondingly, costs are significantly reduced). We deliver translations in up to 36 languages at product release instead of 1-2 months afterwards as done 7 years ago.

Inefficiences: some problems with conditional text - can be difficult to make sure stuff is tagged correctly. Text insets are not easy to use with all the conditional text, so we abandoned those in most cases. The inefficiences are vastly outweighed by the efficiences gained.

We are looking at moving to structured/DITA content. Need to be able to share XML content with the company Knowledgebase and with Tech Support, so it's a win-win. Want to be able to easily repurpose content for multiple output options, including mobile and tablet platforms, as well as product interface screen (printer screen).

Thursday, 23 June 2011 10:24
 
Gina Goodson Wadley
I've been single sourcing with RoboHelp (WebHelp and previously compiled HTML help - .chms plus Word/PDFs) for 11 years. Currently, we're trying to get all of the tech writers on the same tool (we're on RoboHelp, Flare, and FrameMaker) with the ability to export to a wiki (preferably Confluence).

We are testing converting our files to FrameMaker 10/DITA XML. We know we will improve the content with structured authoring (which we are not using now). Content will be more streamlined and concise, and we'll be able to save money on localization costs by reusing content. However, the conversion takes a lot of time and effort, even with FrameMaker's conversion tables.

I'm currently attending Tom Aldous's weekly series on unstructured to XML in FrameMaker: http://tmaldous.com/2011/06/14/7-part-e-seminar-series-unstructured-to-xml-framemaker-10-instructional-series/. It's difficult for me as a DITA beginner to comprehend, but the webinars are helping.

Tom Johnson also posted some articles on his blog about enterprise authoring along these lines that is really helpful: Update on the Search for Enterprise Authoring http://idratherbewriting.com/2011/03/08/update-on-the-search-for-enterprise-authoring/.

I have read that it can take a few years to show ROI. It may take longer since the tools we're looking at to achieve this are expensive (which include a CMS).

Thursday, 23 June 2011 10:17