Reply To: Daily Cipher

[Puzzling_Pelican]

@madness thanks for these examples, from the looks of things, almost all of them are caused by the exact same bug in the code which caused 1x1s to be broken. When the midpoint is perfectly in the middle of 3 letters, some consistent rounding needs to be performed on it. Originally this was to move it vertically to the nearest valid midpoint however my code did not perform this correctly and so ended up rounding diagonally (which is also why 1×1 mappings form alternating diagonals). It turns out that this one bug caused more issues than I thought. I will try to fix it in a bit unless you fix it first.

As for the inconsistent ones you have picked, I can only understand how one of the two are valid, from what I see performing the cipher manually, the midpoint is already considered ‘valid’ and so is not rounded before points are reflected in it.

Finally, if you have a better idea for what happens in the case that the midpoint is perfectly in the middle of 3 letters, other than a vertical move (or whatever the bug in my code did to it), I would love to hear it. This was one of the biggest compromises I made in the ruleset for P3. Only requirements are that it is consistent, simple and self inverse (same to encryption and decrypt).

Thanks again,
Daniel