ICE Performance Contest
- Michael Forney
- Sam Whitlock
Run make, then ./ice start.pbm end.pbm
Our solution will usually answer quickly (within 10 seconds), but every once in a while will take much longer. We were unable to solve this problem by the deadline, but we think it's because it is getting falsely mislead down incorrect paths by our calculate_score function.
We ended up using a hack to get the threads to consistantly terminate by just calling exit. It seems that cancellation points are not created with Macs in the necessary places, and we figured consistancy was better than quality in this case.
Our tests no longer pass because of changes we made to the core of our program and the changes necessary to make this consistent.