WP-34S N-Queens Benchmark Based on Patrice Torchet's BITMAP Solution

[xerxes]

(10-24-2014 11:32 AM)patrice Wrote:  To Xerxes: After studying a few programs from your benchmark, I see that the programs fall into 3 categories:
1) programs that can find all the solutions AND quit nicely.
2) programs that can find all the solutions but fail after the last solution.
3) programs that can find only solutions with the first queen at position 8.

The original program was designed to find the first solution only. Later I was forced to make a structured version,
due to the lack of GOTO in some languages, but it's the same way to the solution. The fact, that the unstructured version
is also able to find all solutions, is a side effect and was not intended.

I've tried to comprehend your observation about category 3 using the FX-603P, but I'm not able to see what you mean exactly.
The algorithm starts with putting the first queen at position 8, without the knowledge that it's not necessary to move
it later for the first solution only. Even if I put the queen at position 8 and start calculating at the second queen, it
doesn't change the execution time significantly.

I've also tried to start the calculation with the first queen in other positions without failing, but of course different execution times.