Tuesday, November 27, 2007

Drools solver @ Javapolis

I 'll hold a Drools Solver BOF at Javapolis 2007 on Tuesday December 11th at 20:00. You're all invited :)
Take a look at the schedule here.
Take a look the contents of the BOF here.

The drools solver manual has also been expanded with more info about the examples:
Take a look at the updated manual here.

The manual now contains some insight about the problem size of the examples. Did you know that the traveling tournament example nl16 finds a feasible solution out of 2,45064610271441678267620602e+259 possible solutions?

And of course I 've added some more eye candy:



  1. so how long did it take to solve 64 queens?

  2. On my pc (a 2 year old mid-end home pc) on windows XP SP2 with jre 1.6, the second run (the first run takes 4s longer to hotspot compile), it takes:

    INFO Solved in 79 steps and 50391 time millis spend.
    So about 50 seconds to find 1 feasible solution for 64 queens.

    That's with tabu search (solution tabu) without a scalable selector (which I am working on).
    Without a scalable selector, tabu search isn't really useful for real-life size problems, such as Examination.

  3. Congratulations Geoffrey, I will try drools solver asap. By the way, good luck at Javapolis 2007 :)