Talk:Solve a Hidato puzzle: Difference between revisions

I suspect that the reason people can solve that problem easily is that they switch strategies easily. Coming up with a way to characterize how to switch is the hard part. (The Tcl code is much more optimized for the case where it's dealing with path sections that are fairly short and which mostly stretch between end points; this is a useful strategy with the majority of Hidato puzzles published in the press, but is far less useful with these evil edge cases.) –[[User:Dkf|Donal Fellows]] 08:07, 3 May 2012 (UTC)