Configuring a Dynabook (is less: Off topic)

Michael Stevens mjstevens at mindspring.com
Sat Dec 12 16:06:14 UTC 1998


At 08:55 AM 12/12/98 -0800, Jonathan Smith wrote:
>
> [...]
>1. Every time I've tried to write down the theory of evolution as a code
>fragment or class definition I've always ended up either (a) providing a
>much too general (and not very useful) definition, or (b) adding too much
>detail and making too many specific assumptions about the representation of
>individuals, genes, etc.  What I would like would be to program a useful and
>complete description of evolution that makes no more of a commitment to the
>particular details than a natural language explanation for evolution.
> [...]

Reading this reminded me of Douglas Hofstadter's _Fluid Concepts and Creative
Analogies_, an overview of a decade's progress by Hofstadter and his graduate 
students toward useful automated problem-solving and knowledge-building.  The 
descriptions are concrete enough that I was tempted to explore the ideas in 
Smalltalk, but I didn't allocate the time. :-(  I seem to recall observing
that some of the ideas lend themselves fairly well to a distributed-objects
implementation.

_Fluid Concepts_ should be available via your local bookseller.  For good 
measure I checked Amazon just now; they want US$16 plus shipping, and they 
claim it ships within 24 hours.

Happy hunting,
--Michael Stevens





More information about the Squeak-dev mailing list