Subject: FSBs and mechanized documentation
From: Rich Morin <rdm@cfcl.com>
Date: Sat, 4 Mar 2006 10:05:25 -0800

As discussed in http://www.cfcl.com/rdm/MBD/, I'm interested
in the application of mechanized approaches (e.g., Doxygen)
to the generation of detailed documentation of large systems.
By merging mechanically-harvested information with comments
from developers, these systems can provide a very rich and
easily navigable source of information.

It occurs to me that these approaches might be applicable to
some Open Source projects (and related FSBs):

  *  The raw information is available for inspection.

     Open Source development tools (e.g., Bugzilla, CVS) have
     accessible code and data, so tools can easily extract
     relationship information, etc.  Also, the information is
     free of proprietary restrictions.

  *  There is a need for detailed documentation.

     Open Source projects frequently involve large code bases
     and large numbers of geographically-dispersed developers.
     These developers could benefit from the sort of web-based
     documentation that packages such as Doxygen generate.

Nonetheless, I haven't found much evidence of these tools being
used in the Open Source community.  I would be interested in
hearing about projects that are using these tools, including
feedback on the perceived benefits and problems.

I would also like to start a discussion of ways in which FSBs
could benefit from these approaches, how projects might be able
to collaborate on building up useful infrastructure, etc.

-r
-- 
Technical editing and writing, programming, and web development:
  http://www.cfcl.com/rdm/resume

Contact information: rdm@cfcl.com, +1 650-873-7841