Need for new paradigm/toolset (Was: Impact of info expl...)

[Emily Skarzenski <eskarzenski -at- DTTUS -dot- COM>]

  Chet Ensign wrote about the need for a new, non-linear paradigm in
  information development. As he said, we need a tool "that merges the best
  of wp/dtp technology with the best of client/server database technology."

  Chet is right on target.

  I see a trend in software development toward "customized" systems that
  exist somewhere between an off-the-shelf and a completely custom
  solution. These products are developed as an 80% solution, where the
  other 20% is client-specific customization, which is either done by the
  vendor, an outside consultant, or technical professionals at the client.
  This allows software vendors to market their wares as customized, but
  reduce the costs involved in developing a completely custom solution.

  Regardless, customers want to end up with manuals and online help that
  are 100% specific to their product.

  The problem this poses for documentation is something like this:

  -- The end-product doc sets have a good deal of content in common.
  -- It's impossible to predict exactly which content will be shared.
  -- It's impossible to predict how the customized parts will be
  customized.

  To me, the solution begins with the ability to modularize information down
  to the topic level and thread together the appropriate pieces on demand.
  We also need to be able to save the "recipe" for version control. And of
  course, the output format should be publication-ready.

  SGML does some of this, I know -- but requires a lot of setup and
  configuration. FrameMaker's conditional text feature indicates a
  recognition of the problem, but is far too limited. Why, I wonder, isn't
  there a shrink-wrapped solution for a documentation scenario that is
  becoming increasingly common?


  Emily M. Skarzenski
  Deloitte & Touche/ICS - Chadds Ford, PA
  eskarzenski -at- dttus -dot- com