What's New -  November 2004




November 30, 2004:   

Panels - Start Shapes - Mid-game StringGrid - Blue wins!

A  young lady recently asked for help in creating a game board so I thought it would be a good chance to explore a few of the choices.  Needing something to test the boards, I coded logic for the game of Reversi (Othello®) to respond to user clicks.  No computer play in this version - we'll save that for another day.   There is more strategy to the game than I would have suspected with a number of web sites dedicated to news, strategy, tournaments, etc.    In any event, here is GameBoards - Reversi over in the Delphi-Techniques section of the site as an introduction.   Three independent boards and games on one page!

November 23, 2004:  I posted a correction for our Tic-Tac-Toe machine today.  This is a program that simulates a Matchbox "machine" which uses colored beads to learn to play.  It gets smarter through trial and error by playing games against a smart human or a dumb computer.  A viewer recently found that the machine would occasionally cheat by overwriting one of your "O"s with his "X" in order to win a game.   I don't believe that the program is smart enough to cheat intentionally, so  I  bypassed that problem today.  It now actually  seems to learn a little faster.   

November 15, 2004:  Here's the first of several small programs that have been accumulating here.  It a Latitude-Longitude Distance Calculator.  Guess that pretty much describes what it does.  There are buttons to estimate distances assuming that the earth is perfectly spherical and one which considers the effect of earth's bulge around the equator.  Neither is original, except perhaps for the fact that they are Delphi implementations. 

November 11, 2004:  I've been working on Project Euler programs for the past month and now have most of them solved  (79 down, 5 to go).  See the October 13th  "What's New"  item for more information about Project Euler.  The higher ranked problems definitely require insight to solve - brute force will never find "the  minimal path sum of an 80x80 array of numbers  from the top left to the bottom right by only moving right and down".   Much fun!

I did post  an update to Tangram2 today.  Tangrams are derived from the ancient Chinese puzzle requiring  seven geometric shapes to be arranged into a square. Extensions require that a set of pieces to be arranged to cover a given pattern.  Viewer Max is converting it to Visual Basic and spotted a bug that let pieces be dropped into a "notch" in the target pattern that happen  to exactly match the piece being dropped.  It's fixed now.   

Muzzle loader hunting season is open here now.  A different kind of time-consuming challenge altogether, but I'll try to get a few more of the backlogged programs posted during breaks this month.