CSC153, Class 26: Randomness and Simulation

Overview:
* The random procedure
* Rolling dice
* Lab

Notes:
* Exam 2 almost ready (should be available this p.m.)
  + Due Friday the 14th
  + Four exams total plus optional final
* No, I haven't graded homework 2
* Questions on writeup 3?
  How should we turn in the stuff?  One big Scheme file.

Did anyone figure out how to make Quicksort do without the
same? predicate?

----------------------------------------

Scheme provides an interesting procedure that differs from
many other Scheme procedures:

(random i)
  i is a postiive integer
  returns an integer between 0 and i-1, inclusive

(random i) may return different values when it is called
different times.

> (random 5)
4
> (random 5)
0

Goal: It is "difficult" to predict what value will appear next.
Implies: Values are evenly distributed.

* How is random different from other Scheme procedures?
  Most Scheme procedures are intentionally consistent 
    (and should be)
  random is intentionally inconsistent 
    (and should be)
  read is also
  (Things based on read and random might therefore be
    inconsistent.)

* Why would we want this procedure?
  + Games may require some element of randomness
  + Simulation of events may require some element of
    randomness
* Most computers use an algorithm to generate random
  numbers:  We call those pseudo-random
* We can build devices that generate completely unpredictable
  numbers.

Read the lab.  Wait a few minutes before trying to copy files.

----------------------------------------

Why is it that Linda and Larry's code fails to work well in
DrScheme but works fine on the Web?

(define foo
  (let (...)
    (lambda (..))))

Konquerer is too smart for its own good.