1) In the second shift signal, you no longer have the hams distanced far enough so that an area covers either only positive or only negative cells. So if an area contains both positive and negative cells does it carry a positive or negative signal? Or if it matters just the polarity of the cell exactly near the cut, why did you distance the hams so much when explaining the tracks?

2) To get a “pizza grid” from the 1 in 3 sat graph, once the nodes and edges are transformed, is the area in between filled with no-hams cells?

3) How would you prove the construction can be done in polynomial time?

Anyway, cool proof! It’s the only hashcode problem for which I found a proof of np-completeness. ]]>

Just wondering, how did you calculate the average distribution or the probability for each sum ?

]]>I tried to invert simply the two pins 1 and 12.

Cannot find a solution quickly, please post this string in a comment below to help me improve the solver: [12,6,5,4,3,2,11,7,9,8,10,1]

]]>7,298,706,025,000,999,840 possible iterations versus

479,001,600 pin arrangements. Correct Icosaku solutions being mere fraction of total possible tile arrangements. ]]>