Lunch with Bob, Wayne, Steve, and David had numerous highpoints.

• Learning Objects. We'll no sooner deal directly with learning objects than we'll order atoms for breakfast. Yeah, they're in there, but they don't appear on the menu. Assemblies, for example scrambled eggs, are the level we call for.
  
• Housing provides a great example of right-sized assemblies or components. Decades ago, modular housing was deemed inevitable because of efficiencies of production, but it never took off as predicted. Assemblies -- entire rooms -- were too large to customize. Today, pre-built components make up 80% of a new house. An assembly might be a door, hinges, doorframe, doorknobs, and lockset. Homebuilders can mix and match to achieve custom results from standard assemblies.
  
• Wayne and Steve hypothesize that granularity (assembly size) is directly proportional to ease of understanding but inversely proportional to reusability. Exponentially. Hence, I can use a nail just about anywhere but by-and-of-itself, a nail doesn't tell me much.
  
• Object orientation's raison d'être is to avoid reinventing the wheel, or, in this case, the object. The effort should go into design and assembly, not rebuilding objects.
  
• "Design is finished when it's fully constrained." Does this mean that designs are never finished? Wayne allayed my fears that the goal was to turn everything over to machines and their algorithms. The overall goal is to augment human capability, not to replace it.
  
• Great parallel from the world of music. Give a score to an orchestra and also feed it into a synthesizer. The human ear prefers the orchestral version. Have the computer isolate the differences in the performances and generate algorithms to describe them. Feed the algorithms back into the synthesizer and get back a more satisfying sound.
  
• Wayne: "In five years, all content will be dynamic." Who needs the standard version when the personalized one costs no more?
  
• Wayne: We've got the objects and the metadata standards. Now we are working on the mapping. (RDF to define relationships among objects.)
  
• Slippery ground: When is it meta-data and when is it data? I'm sure Heisenberg would have something to say about this....
  
• Wayne: "I don't understand the utility of the New York Times Bestseller List. I want to know what my friends are reading."
  
• Wayne: Learning objects are like teenage sex. Everyone says they're doing it. Very few are. And those who are doing it, do it poorly. Maybe 10 of the Fortune 100 are doing it. Poorly.
  
  
• A standard should follow practice and not vice-versa.

Recommendations:

• Nonaka
  
• Foundations of Language: Brain, Meaning, Grammar, Evolution by Ray Jackendoff
  
• The Way We Think: Conceptual Blending and The Mind's Hidden Complexities by Gilles Fauconnier, Mark Turner (Hardcover - April 2002). Review
  
• Architectural Learning? by Phillp Palin? I can't find a reference on Amazon. Source of the hub and authority structure we discussed.
  
  
  
  

  ---

  
  Links and notes from Steve:
  
  Here are a few links to articles about process maturity and content reuse
  from JoAnn Hackos. You may find them interesting.
  
  
  Accidental Reuse
  http://www.infomanagementcenter.com/enewsletter/200205/feature.htm
  
  Are Process Maturity and Content Management Linked?
  http://www.infomanagementcenter.com/enewsletter/200207/feature.htm
  
  Using the Information Process Maturity model for Strategic Planning
  http://www.comtech-serv.com/pdfs/Strategic%20Planning.pdf
  
  
  Location to download Software Engineering Institute's Capability Maturity
  Model
  http://www.sei.cmu.edu/cmm/obtain.cmm.html
  
  Finally, here are some notes that I shared with JoAnn regarding the article
  above, "Are Process Maturity and Content Management Linked?"
  -----------------------------------------------------------------------
  
  
  1. Emphasize content reuse in the more sophisticated management stages.
  You point out how rudimentary content management for single-sourcing
  does not necessarily provide a highly-useful or sophisticated solution. I
  recommend that when describing the more sophisticated stages, you emphasize
  the nature of content reuse. Here is a rough cut of how I would position the
  stages:
  
  
  Stage 1 - Content Management.
  You can find the content in the system.
  
  Stage 2 - Single Sourcing
  You can present the content in multiple formats.
  
  Stage 3 - Content Assembly
  You can assemble content for a single purpose into collections
  targeted at different audiences.
  
  Stage 4 - Content Reuse
  You can assemble content for collections across multiple purposes
  and multiple audiences.
  
  Stage 5 - Automated and Personalized Reuse
  Content is automatically assembled based upon dynamic specification
  of purpose and audience.
  
  My stages 3, 4, and 5 try to blend a heightened awareness of reuse
  with the hierarchy of automation and personalization in your stages 2, 3,
  and 4.
  
  
  2. Parallelism between reusable content objects and the definition of Level
  3 process maturity.
  Successful implementation of reusable content management relies on
  creating Structured and Reusable content. I was struck at how precisely
  matched Structured and Reusable content is to Organized and Repeatable
  process. I think this would be a good parallelism to point out.
  
  3. Information design and granularity.
  I liked your reference to mature organizations developing consistent
  information types. I recommend also mentioning how increasing levels of
  process maturity support managing of content at ever smaller levels of
  granularity. As process maturity increases, an organization moves from
  managing content at the document and chapter level to having the skill to
  manage content at the topic or block level. Note that information types can
  be applied at both the topic and block level.