(24-10-2023, 04:11 PM)nablator Wrote: You are not allowed to view links. Register or Login to view.Note that You are not allowed to view links. Register or Login to view. and You are not allowed to view links. Register or Login to view. don't have any EVA-e so this would be an unlikely way to count distances. 
Oh well, it's only an example.
Well, if you literally apply the cipher from my article to Voynichese, you don't need external props to count the distances. There are only three distances in my version of the cipher: roughly two characters wide (like kek), roughly four characters wide (keoek) and roughly six characters wide (keoyedk). They are quite unambiguous in practice. It gets much harder if you add the fourth distance, but three shouldn't be a problem.
Btw, it someone wants to try looking for any specific interpretation of Voynichese as a distance based cipher, one possible vector of attack could be long sequences that only contain 3-4 different characters. Things similar to ar.ar..al.olar.al. It could be possible to identify specific matching characters in these sequences by analyzing these patterns. qokeedy.qokeedy is worse for this, because ee could be significant, so we have up to 6 different characters here that can pair up in many possible ways, like q with y and o with d and k with cc, etc.
(24-10-2023, 05:06 PM)oshfdk Wrote: You are not allowed to view links. Register or Login to view.Well, if you literally apply the cipher from my article to Voynichese, you don't need external props to count the distances. There are only three distances in my version of the cipher: roughly two characters wide (like kek), roughly four characters wide (keoek) and roughly six characters wide (keoyedk). They are quite unambiguous in practice. It gets much harder if you add the fourth distance, but three shouldn't be a problem.
You should define "roughly".
From your example:
The two k in key.qok are 6 chars apart, why kk2 and not kk3?
The two k in keedy.qok are 8 chars apart, why kk3 and not kk4?
If dots are not counted, then you will have 1,2,3 apart all at distance 1, is that right?
If there is a way to limit distances to something as low as 3, it would be in itself a very interesting result.
(24-10-2023, 05:38 PM)nablator Wrote: You are not allowed to view links. Register or Login to view.You should define "roughly".
From your example:
The two k in key.qok are 6 chars apart, why kk2 and not kk3?
The two k in keedy.qok are 8 chars apart, why kk3 and not kk4?
I had no intention to make my example for qokeedy coherent in any deep way. It was only to demonstrate that visually repeating sequence could encode a non-repeating plaintext, nothing more.
I'm talking about physical distances on physical page. For Voynichese with average letter width of 3-4 mm, convenient distances between anchor points could be roughly 8 mm, 16 mm and 24 mm, if we apply the specific cipher I designed to Voynichese. The probability of my design coinciding with the Voynich code parameters is very low.
Quote:If there is a way to limit distances to something as low as 3, it would be in itself a very interesting result.
I think I demonstrated just that in my article. You can encode a set of 27 characters with 9 cipher characters and 3 distances. I opted for a 12 character plaintext alphabet in my text only to be able to produce more diverse sequences. On the other hand, Voynichese definitely has more than 9 different glyphs, so 3 distances should be more than enough to encode Latin. Maybe even two distances.
(24-10-2023, 06:20 PM)oshfdk Wrote: You are not allowed to view links. Register or Login to view.I think I demonstrated just that in my article.
I meant with Voynichese.
I deciphered you short exercise. Your bar '|' is at a place where there is no unpaired letter left (there are other places). Not really useful, unless someone wants to skip a part.
How would labels be protected? Without nulls, all labels should have an even length. There isn't room for many nulls anyway, so the nulls would be quickly discovered. A few unpaired glyphs would make things more difficult.
(24-10-2023, 07:18 PM)nablator Wrote: You are not allowed to view links. Register or Login to view. (24-10-2023, 06:20 PM)oshfdk Wrote: You are not allowed to view links. Register or Login to view.I think I demonstrated just that in my article.
I meant with Voynichese.
With Voynichese I tried some analysis by manually annotating points on a few pages, but with no conclusive results. Nothing in the data strongly supported a distance cipher.
However, I was mostly looking for a simple distance cipher as described in Cryptomenytices or in Vignère's work. My motivation with the cipher I designed for my text was exactly to try envisioning something that could work with Voynich irregular distance patterns and using a combination of principles known from documented distance ciphers.
Quote:I deciphered you short exercise. Your bar '|' is at a place where there is no unpaired letter left (there are other places). Not really useful, unless someone wants to skip a part.
It's not useful in this example, but it demonstrates that the cipher can include arbitrary extraneous marks, denoting boundaries of cipher structures or plaintext words, sentences or reordering encoded chunks on the page
(24-10-2023, 08:32 PM)nablator Wrote: You are not allowed to view links. Register or Login to view.How would labels be protected? Without nulls, all labels should have an even length. There isn't room for many nulls anyway, so the nulls would be quickly discovered. A few unpaired glyphs would make things more difficult.
Yes, you are correct, that labels present certain challenges. Not for nulls, if you have a clearly separate label, it is possible to use any unpaired character as a null. E.g., using the qokeedy example scheme: oqokyek has two pairs o-y and k-k, qo is an unpaired null, since no terminator corresponds to it.
Another problem is there are many assumed labels of 3-5 characters with no anagrams (the order or characters appears to be fully fixed). If this cipher encoded text by placing the same terminators at different spacial positions, a lot of apparent anagramming is expected. This could work if short labels are just letter/number references or not labels at all, but some technical marks. Anyway, there is no specific evidence for a distance based cipher, for now I'm only arguing that there is no specific evidence against it either.
Personally, I will focus on one-to-many additive ciphers in general.
I wonder why everything is so complicated.
Imagine using two alphapets.
One normal ABC and one for the combination. That's just because you can't count the letters.
It looks like this. The great thing about it is that I can write the same words differently. In the example of "et", which can also stand alone, "t" can be read as "tum" from the combination "et" with the ending "9 = us/um". The same goes with "EVA ch = er". I can write it together, but also separately. In this case, "e" always remains "e".
[
attachment=7801]
And if I take it to the extreme, I also get a text.
That would be a nice final sentence for a plant.
[
attachment=7802]
There has already been something like that. Here is a slightly more complicated version.
PS: A thank you to Nick
[
attachment=7803]
I'm not sure I understand your reading. Why is it that deol becomes tum and doldaiin becomes tatis, what happens to ol in both words?
To my knowledge so far there have been no convincing attempts of mapping Voynichese to any language plaintext via any kind of multicharacter substitution table. Maybe yours is the first, who knows.