Voynichese is a numeric cipher?

[ololololo]

but how much time will it take to find the correct decoding?

Well, I just guessed the words I saw. Of course, the phrase doesn't make sense on its own, but the words were still guessable. My goal was to solve the anagram...
Plus, I did it at speed to see how quickly I could see the words. 
Either it's a problem with the algorithm, or I'm just not paying attention. I don't know what else can be made out of VROE, for example.
However, I wouldn't say it was difficult (at least I quickly guessed the words, even if they were wrong).
Neither the first nor the second stage of decryption is particularly difficult, but it does take some time.
It's worth trying to "torture" other users in this way to understand how it works for them, not just for me.
I don't think my "efforts" will really prove anything to you (at least it will always be a little easier for me than for others )

P.S. It would also be more reasonable for you to try to see if you can use this cipher yourself (i.e., if you can encrypt text, find letters, assemble words, anagrams, etc.). It's just that if I solve the anagrams incorrectly, I don't know what it will prove. If I don't solve them, perhaps another English speaker will solve them correctly...

[oshfdk]

I don't know if there is an error in the script, but some of the letter values look wrong. In any case, thank you for participating in this experiment. To me it's overwhelmingly clear that this cipher is too complex to read, which doesn't align well with my expectations about what style of a cipher it should be.

[ololololo]

To me it's overwhelmingly clear that this cipher is too complex to read, which doesn't align well with my expectations about what style of a cipher it should be.

Well, it's relative. The most important thing about it is simply to solve the anagrams, which can be more time-consuming than deciphering the letters.
In my opinion, this cipher is simple in itself, because at the moment it's just a slightly more complex substitution. All you really need to do is decrypt the letters and get all the anagrams, and you'll have the plaintext.
It's worth considering making it a bit more challenging...

In any case, thank you for participating in this experiment.

And thank you very much!  The experiment proves that I'm not suitable for such experiments...

[ololololo]

The word cheody in my mini-code can be obtained by gluing o o o dy (o o o f). The first two o's combine to make che, and then we add ody. 
Similarly, you can get some other words, such as chedy. They appear when you encrypt the text (well, as che o dy and che dy, respectively).
The following issues are currently unresolved:
Also, some garbage elements can be inserted directly INSIDE a number. You will still understand that it is null, for example dy - dyty (found on f19r). The combination of yt makes no sense.


Two units (in) can also be changed using garbage gibbets, for example, in - yty (by removing the gibbet, you get yy, which is equivalent to in).

There is plenty of room for "Voynichization"... However, there are no specific rules for secondary encryption that can provide the desired statistics...

However, there are currently no other issues with the cipher. Or I don't see them...

[oshfdk]

Maybe I misunderstood something, but you couldn't decode the message I created using the encoder script, is this correct? Which means the cipher simply doesn't work?

[JoeyB]

Maybe I misunderstood something, but you couldn't decode the message I created using the encoder script, is this correct? Which means the cipher simply doesn't work?

I interpeted it as he had trouble doing it by hand since machine decode ostensibly takes seconds: FIRE IS THE PURE AND VOLATILE PRINCIPLE THAT CONSUMES MATTER TO RELEASE LIGHT AND HEAT WHILE TRANSMUTING BASE SUBSTANCE INTO ASH AND SPIRIT

[oshfdk]

I interpeted it as he had trouble doing it by hand since machine decode ostensibly takes seconds: FIRE IS THE PURE AND VOLATILE PRINCIPLE THAT CONSUMES MATTER TO RELEASE LIGHT AND HEAT WHILE TRANSMUTING BASE SUBSTANCE INTO ASH AND SPIRIT

This is exactly the source phrase.

So, it is decodable, but the question remains how quickly a 15 century reader using tools and technology of their time would be able to decode it. From what I've seen so far, this would be a slow process, certainly not faster than one sentence in a few minutes, in edge cases (several possible anagrams for a few words) could take much longer or multiple viable interpretations could co-exist.

[JoeyB]

I interpeted it as he had trouble doing it by hand since machine decode ostensibly takes seconds: FIRE IS THE PURE AND VOLATILE PRINCIPLE THAT CONSUMES MATTER TO RELEASE LIGHT AND HEAT WHILE TRANSMUTING BASE SUBSTANCE INTO ASH AND SPIRIT

This is exactly the source phrase.

So, it is decodable, but the question remains how quickly a 15 century reader using tools and technology of their time would be able to decode it. From what I've seen so far, this would be a slow process, certainly not faster than one sentence in a few minutes, in edge cases (several possible anagrams for a few words) could take much longer or multiple viable interpretations could co-exist.

Agreed, the scribe doing that for 200+ pages is going to have a major headache. And the longer the word, the harder to unscramble the anagram so the headache scales.

[ololololo]

You are all actually absolutely right. However, there are two points:
1). No one prevents us from leaving a hint for a complex anagram, such as rearranging a letter in the correct position or leaving certain letters in their original order. Or you can mark them with a null, for example.
2). We don't know yet how much this applies to Voynichese. There's a chance that the author was more creative than us in this regard... 
The word chodaiin appears several times in Botany and once in the Zodiac. The letters are aligned correctly from a Voynichese perspective - che at the beginning, iin at the end. However, there's also the word ararchodaiin, which represents the root. There are two options: either the author put arar at the beginning as a hint, or... this is the original form of the anagram, and the author did not change anything (which is also possible, because what Latin word starts with two identical letters?).

In general, problems of this kind can be solved manually. But yes, it is expensive, and if you continuously encrypt some text, it will be faster to simply encrypt it and spread out the nulls.

And, @oshfdk, I apologize for being a very poor anagram solver Apparently, I'm not paying enough attention. I think you should contact other, more attentive people next time and evaluate their results.

---

Maybe I misunderstood something, but you couldn't decode the message I created using the encoder script, is this correct? Which means the cipher simply doesn't work?

Maybe... Anyway, the code was generated by a neural network, so it's quite expected. Or, again, I'm just being careless...
I would like to ask @JoeyB how he deciphered that phrase - intuitively or by directly working with the encrypted text? Maybe I could have made a mistake myself.

[ololololo]

For convenience, you can now set some rules for yk/yt nulls. They are inserted in the same positions as ch, but with one difference: it can also be placed before ch/sh. This will help you avoid excessive ch nulls in the text.
In Botany, yt can be used to separate two repeating letters from each other (olol - olytol, chchy - chytchy), which can be misleading and blur the boundaries between words (daiin chy - daiin ytchy). In other cases, it is equal to null ch.