I have been using the c++ implementation of dancing links a lot during the last decade for fun [1,2] and for research. There were some issues in the original C++ code that are now fixed (providing a empty set of rows was crashing sage, also, we can now reset the search from the start, reduction to sat and reduction to milp, etc). I forgot about the existence of the Python implementation, since I always use the C++ one.
[1] http://www.slabbe.org/blogue/2015/09/there-are-13.366.431.646-solutions-to-the-quantumino-game/ [2] https://ask.sagemath.org/question/81610/test-if-a-graph-has-a-claw-decomposition/ That being said, Franco Saliola from Montreal told me a few years ago that the code written by Knuth in recent years is faster (for some problems that Franco was working on) than the implementation currently in Sage. There are heuristics on the choice of the next column to deal with in the search that can be better suited for some problems. This is something on my TODO list since many years to check this out and maybe provide this in Sage if copyrights allows it. Sincerely, Sébastien -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To view this discussion visit https://groups.google.com/d/msgid/sage-devel/35dda6a6-0c9d-4826-bc21-fa08f1f1b103n%40googlegroups.com.
