This site contains software, essays, and other content by Brandon Thomson.
I enjoy solving programming puzzles. My favorite puzzles are those that could be solved with a very simple brute-force algorithm if computers were 10^10 times faster or had 10^10 times more memory, but require clever reductions in the search space to run on our "slow" machines. I'm not a mathematician so I don't enjoy puzzles that _require_ an advanced knowledge of mathematics and I don't care much for the competitions that pit you against other humans in a race against the clock.
I find that the puzzles on projecteuler.net usually meet my requirements. Project Euler is a set of 158 problems (at the time of this writing) that are designed to be solvable only using the aid of a computer. In practice many of the easiest problems turned out to be solvable using math shortcuts that the author didn't anticipate, but for non-mathematicians the easiest way to solve the problems is almost always by programming a computer.
Naturally a lot of people write about their solutions, so be careful Googling around if you don't want a solution spoiled for you. Using languages like Python or LISP will tend to get you solutions faster than languages like C or Java. Mathematica will save a lot of time by providing access to lists of primes and other nice features, but it almost seems like cheating.
You can view my profile, but you'll need to register first.
A Kakuro puzzle, also known as a cross sum, is often referred to as a mathematical transliteration of the crossword. Essentially, you must fill in blocks such that all the digits in a group of blocks sum to a specific number, both horizontally and vertically. You must also not repeat any digits in a horizontal or vertical box, and can only use the digits 1-9. Wikipedia's page on Kakuro has a lot more info.
This short 'n sweet little program written in Python will solve any of the three hard-coded puzzles, or any others that you enter into the appropriate place in the source file. The pygame library is required to display the output, which looks pretty good. Press 's' to solve when the program is running.
No guarantees against bugs. I haven't looked at the code in a while.
This code is licensed under the GNU GPL v3.0.
Download the Python source code.
Looking for something else to do? You can enjoy a little story about Alpach Software, or visit a few useful links.
Questions? Comments? GRAVIX [at] GMAIL [dot] COM