(18-10-2023, 05:07 PM)nablator wrote: You are not allowed to view links. Register or Login to view.In a cipher like mine there are many possible paths, the scribe could have used the p and f coordinates preferentially on the first line without difficulty. Why is a different question, and there is maybe no answer other than "it looks nice".
I'm not sure I understand this point. There are many possible paths, but there are only a few possible starting points.
Unless the table has some fields that encode empty strings. That is, when decoding and you get to this field, you just proceed without adding anything to the plaintext. Null fields as opposed to null characters. This will make the whole graph appear much more manageable. I was thinking about finding a limit on the efficiency of this cipher by computing the optimal table arrangements that would allow using the smallest possible number of target glyphs to be able to jump between arbitrary source characters, but quickly realized that introducing null strings would allow to use arbitrarily small secret alphabet.