Not being sure what exactly to do, here is one A^6 surrounded by 6 of B^3: ec3464-curve-tiles-1-AAAAAA-and-6-BBB.pdf Now that shape might be the same as for A^6 (and B^3 might be that shape as well), I don't know. Even if this is the case, it does not hold for other curves, see wild-curve-tiles-1-AAAAAA-and-6-BBB.pdf This one certainly has a shape different from A^6 and B^3. What I really like about these curves is that instead of one curve being decomposed into smaller copies of itself, now there are two curves decomposed into smaller copies of both curves. See either ec3464-curve-A-decomp.pdf together with ec3464-curve-B-decomp.pdf or wild-curve-A-decomp.pdf together with wild-curve-B-decomp.pdf The grids used before (triangle, square, tri-hex) all had one kind of edge, this one has two. Now guess what I expect to find on grids with k different edges... Hacker question: how do I map the points of the grid to a 2-dim array, preferably using integer coordinates? My best idea so far is to address the points in two steps: address the hexagons by their Eisenstein coordinates, then the six points on them in some (rather arbitrary) way. Another idea would be to distort to the (3^5.6) grid (implicitly dropping some edges there) and use Eisenstein coordinates there, but this feels somewhat of an ugly thing to do. Best regards, jj
[...]