Re: functional programming is great, but why lists?

[Bill Richter <richter@math.northwestern.edu>]

Thanks, Matthias!  I've been quite impressed with SICP's functional
definition of cons, which is also explained in HTDP.  In fact for a
while I thought the HTDP definition of cons was through such a
functional definition, but later I realized you were using a
constructor, but you explain that the semantics of your constructor
pretty much follows the functional definition of cons.

   Mike, SICP is messed up. To demonstrate why, please ask Hal and Gerry to
   fill in the gaps:
   
    (define (cons? x) ...)
   
   I had a student go thru the exercise once, and that's when I
   discovered that checks and errors counted in a system and not the
   good behavior. The DrScheme experience reinforced this motto at
   several level above that piece of work.

Cool!  Definitely something I wasn't thinking of.  Shriram mentioned
error checking as well.
   
   If you want some theory, see the paper on the Expressive Power of
   lanuages. 

I think you mean:

Felleisen (Science of Computer Programming) 
On the Expressive Power of Programming Languages

Will do.  It's 45 pages, might take me a while.