# What's New -  March 2004

### [Home]

March 29, 2004: While proving to James the other day that there were really 293 ways to make change for a dollar, I needed to resort my output grid in descending order to match his format.  Here is an update to the Grid Quicksort program which will sort string grid columns in ascending or descending sequence.  (By the way, he had missed the 4 ways to make change that have more than 2 quarters.)

March 28, 2004: I posted version 1 of a Drag Colors Puzzle today.  The idea is to guess a secret color pattern generated by the program in the fewest number of guesses.  The program is classified at the Beginners level since it has less than 50 lines of user written code.   It's analogous to Mastermind, except the only feedback given is the number of colors in the correct position.   It seems that this makes the game more difficult to analyze, at least I haven't been successful so far.  I suspect that similar pattern guessing games must have been analyzed and documented under other names.  Let me know if you locate or develop an optimum guessing strategy just to put me out of my misery.    (This version has pattern  of 3 colors chosen from a palette of 4 colors  - an analysis would derive the best strategy for patterns of length M chosen from N colors.  Seemingly not a trivial exercise.)

March 26, 2004:   A viewer today raised some question about the  Making Change program , which solves the number of ways to make change for \$1 and some other making change problems.   (I think he thinks there are only 289 ways, but it wasn't entirely clear.)  In any event I ran version 2 of the program and couldn't even understand my  description of the "Minimum coin set"  problem.    I improved the problem description and changed the  program to list a sample solution for all of the smaller sets of coins that can make change.

March 23, 2004:  We're home!  When Amy thanked us for helping out with the new baby,  I thanked her back for providing us one more chance to experience miracle of new life.  Trying writing a program to do that!

I received an email last week from a beginning Delphi programmer asking for help with a program in which the user dragged colors to make  guesses of a secret pattern.   It triggered a question about how simple the drag/drop part could be.  The answer is in  this  Simple Drag Drop program.  OK, OK, to kill the suspense -  the answer is just two lines of user written code!   Surprised even me.     We may finish the rest of the puzzle this week - just for fun.

March13, 2004:  Grandma won't let me have the baby so I fixed a small problem in Intersecting Lines demo over in Math Topics.  Viewer Phil Hackett had pointed out earlier this month that testing parallel lines caused a "divide by zero" exception instead of a "Lines do not intersect" message - now fixed.  And Grandma's turn is up.  Thanks Phil.

March 10, 2004:  It's a boy!  Just to let DFF viewers know that programming will take a back seat for the next week or two.   Somebody has to hold the new grandson  and, once in a while, it is even my turn!   I'll get back to e-mail responses and projects when the excitement dies down.

March 6, 2004:  Sharp-eyed viewer, C. Doumar, spotted a bug in the auto-solve triangular puzzle code. It now finds quick  "1 peg remaining" solutions for the 6-pegs-per-side cases.  I reposted Peg Solitaire today with that fix.  7-pegs per-side  still not so quick.

March 4,  2004:   My old business partner asked recently if I could do a  version of the triangular peg puzzle similar to those for in Cracker Barrel restaurants.   Getting down to one peg left is driving his wife crazy!  I re-posted a Peg Solitaire version today which includes it.  I didn't not redo  the drawing routines to draw an equilateral triangle, but you can play a right angle triangle version with 5, 6, or 7 holes per side. (If you can solve these, just mentally rotate the 5-per-side puzzle by 45 degrees and Cracker Barrel will be labeling you as a "Genius". )   Note that "Auto-solve" has not yet solved the "one peg left" 7-per-side version so I don't even know if a solution exits.  If you let your PC run it long enough to find out, let me know.