Ben Younkins
CSC 415
November 2, 1999 -- 11:15PM

==> Note:  All testing was done on Borland C++ v5.02


Program Recipe:
  prg2main.cpp
  ckstroll.cpp

Combinations that do evil: (there's probably more)
  from  |  to  | results
 -------+------+---------
    0   |  11  | crash
    4   |   0  | crash
   11   |   0  | lockup
   15   |   0  | crash
   20   |   0  | lockup
    1   |  10  | crash

To turn on debugging:
  Open ckstroll.cpp
  Scroll down to the "GoStroll" function
  After the variable declaration, (*lines 94 & 95*) set the values to TRUE
  helpless -- Prints out raw information for each step taken.
               The format of the information is:
                 Current Space, Move made to here, Next Space, Move to
                    or
                 Current Space, Move made to here, "No move possible"
  helpful -- Prints out the grid with the moves marked as each step is taken

***********************************************************************

Points of interest:
  The program still suffers from errors upon the entry of certain pairs of points on the chart.  I believe that the errors could be avoided with a complete redesign of the program.  Currently the program searches dumbly through all the board, wandering aimlessly until it collides with the target point or until it paints itself in a corner.  No, it is not supposed to be able to reuse any squares in it's search for the goal.  This feature has not presented any problems.  The prevalent error that occurs is when the program has to back up for more than 3 steps.  Upon the 3rd or 4th step back its trail, it somehow loses all record of it's previous path.  I'm not sure how this occurs. 

  The template stack that I'm utilizing for the program is something that I've reused over several programs and this is the first time that I've encountered an error that could possibly be the fault of the stack.

  The program is essentially in the same state as it was upon it's original completion.  The only that's been added was further error protection.  The latest level which was added tonight, seems to prevent the program from falling into the loop, but it instead causes the whole program to crash and burn.