Thanks for the pointer, this page is really cool. It looks like someone here found a different Octopuszle solution: http://forums.xkcd.com/viewtopic.php?t=99760

Thanks for the pointer, this page is really cool. It looks like someone here found a different Octopuszle solution: http://forums.xkcd.com/viewtopic.php?t=99760 http://forums.xkcd.com/viewtopic.php?t=99760 Looking at the four corners of the solutions on mindsports.nl, the pieces are nearly identical, and in particular the diagonals are identical. It makes sense that the two hand-solvers would use this symmetrical configuration. The programmatically found solution has different corners. Still, that amount of constraint is surprising. It would be a nice surprise if that corner symmetry could be proven to determine the diagonal complex.