What's New - June, 2006
[Home]
June 23, 2006: The Clock Angles Puzzle, posted a few years ago asks for the number of times that clock hands are at 90° angles in a day. The equations given to calculate the times were accurate but confusing to another viewer recently. I just posted a table of the values which may help clear thing up.
June 21, 2006: I've been busy for the last week or so, but not to much to show for it yet. I have a running version of a "Delayed Column Generation" program which solves the "Cutting Stock" problem - cutting lengths of material of specified lengths from stock material of a specified length. The "columns" in this linear programming problem represent the possible patterns for cutting the material from the stock pieces. The complication is that the number of these patterns increases very rapidly as the number of required part lengths increases. Delayed Column Generation solves the problem by starting with an imperfect, but easy to find, solution and systematically adds "columns" (other possible ways to cut the stock) until no further improvement is found. Still need to document this and perhaps find a way to produce integer only solutions for the number of stock pieces required.
Another program waiting documentation solves the "CoalToDiamonds" puzzle. A 3X3 array of "Coal lumps" is turned to "Diamonds" by passing through 2 intermediate steps. Clicking changes an item to the next step, but it also changes he other items in that row and column to the next phase (including Diamonds which convert back to Coal). Changing an array of 9 coal lumps to 9 diamonds requires 16 steps, almost impossible for mere humans. There are over 264,000 start puzzle configurations of which 16,384 can be solved with 1 to 16 moves required. Again, I just need time to do some documentation for this program before posting it.
I did however post an update to the Countdown timer class so that i could use it to track run times for the various search methods used to solving CoalToDiamonds. The timer now counts up as well as down and is more accurate than version 1 when running on a busy computer.
Oh, and there was a new Project Euler programming problem last week involving "Addition Chains", a set of integers from 1 to some N with the property that each number in the set is sum of 2 other numbers in the set allowing reuse. (1,2,3,5,10,15 is one example.) There are lots of interesting questions about how many ways there are to get from 1 to N, which is the shortest, etc. There are now about 40 people (7 Delphi programmers, including "delphiforfun", me) that have solved all 122 problems. A few years ago I posted the first 10 even numbered Project Euler problems, if you want to try your hand and need help.
We have family visiting next week, so may not get much more programming done until July.
June 12, 2006: We're back from a week in Huntsville where we enjoyed 12 year old granddaughter Sarah in a dance production. What a cutie! And smart enough to find program bugs in word search program Akerue and the tricky hangman program, Hangman2 in her spare time between rehearsing and performing. Akerue somehow would not allow words to be added to the "Added Words" list for valid words not in the dictionary. At the easy level, Hangman2 did not allow 3 letter words and hung the "convict" in 10 moves instead of the 12 advertised. Both programs have been fixed as of today.
June 1, 2006: Two guys, Ted Blumberg and Jeff Bratt have taken care of my spare time this past week by being actual program users. Ted is a designer for First Source Products in Johnson City, TN . He used Cutlist to plan layouts some bed frames he is making for personal use and likes those multi-page printouts (which should now work). Jeff is a programmer/photographer/woodworker who spent time this week playing with and finding bugs in Akerue.
Cutlist 3.2.3 and Akerue 4.1 are the results of their feedback. Thanks guys!