Comment by taeric

Anyone know if there is already a comparison of this to Knuth's DLX algorithm? I see they claim this approach will be superior to any backtracking algorithm. Empirical results would be interesting.

I'm also curious how well this generalizes.