Wow, been no comments last 24 hours, and only 3 the last 48 (3 full days on the other thread)
Are we running out of ideas, or people off for the weekend?
2 Likes
Today I dreamed that I solved it. The key was: sweets
5 Likes
Alright, just to focus on “death” and Tofugu, a number of related articles appear. Even if they are irrelevant clues, all are an interesting read:
Tsujigiri
The number superstition which correlates with several numbers which can translate to levels: 24, 42, 43 25, 49
Notably: the joked around 42+ and level ender - “42 (the answer to life, the universe, and everything!) can become shini , aka “death” or “to death” ( shini 死しに)”
Japanese Funeral
Japan population decline
Sanpaku
Seizo Fukumoto
Japanese superstitions
Japanese horror movies
Karoshi
4 Likes
I’m pretty sure we’re out of ideas.
I spent some time last night mulling over the base64 idea that people had the first time we discovered the four codes. Basically, to explain how it works, base64 works by converting text into binary (via ASCII), then splitting that binary number into clumps of six bits (i.e. a base 64 number) rather than eight, then converting those back into letters by some other lookup table.
The issue here is that the base64 lookup table has 64 symbols, but our codes are lacking lower-case letters, so even assuming all upper-case letters are possible, there only seem to be 38 symbols in our encoding. Plus, although the base64 lookup table is frequently the upper-case letters, lower-case letters, numbers, and + and \ in that order, there’s no particular reason why it has to be that.
There are common alternate base64 lookup tables, but the only one which uses ~ as a symbol uses it in place of +, which doesn’t work for us. Another idea that briefly struck me was uuencode, because it doesn’t use lower-case letters - it specifically uses the first 64 entries in the ASCII table as its symbols (and it’s also quite obsolete, i.e. as dated as Shift-JIS), but the issue here is that (a) twenty-six of the first 64 entries are punctuation marks, so we should expect to see a lot more of them in the code, and (b) none of them are ~.
We could try coming up with our own base38 enocoding (i.e. A-Z, 0-9, +, ~) and see if that works, but aside from the fact that I’m not entirely sure how it works when it’s not a power of 2, we can’t be sure of the order. And why base38? Maybe it’s base42 (though that makes the ordering problem even worse, because now we have to add four extra symbols somewhere).
So yeah, I’m not sure this is going anywhere.
8 Likes
I dreamed it was Stars (among others, Have dreamed of it several times 
)
4 Likes
It really felt like the keyboard thing was going in the right direction…but maybe I just want that to be the right way…
5 Likes
I’m still intrigued by the fact that the map of keys used is almost perfeclty symmetrical. And it looks kinda like a ū…
Though, I’m also intrigued by the fact that the length of each code fits the form 6n+2.
6 Likes
I still have a feeling it has to do with typewriters…
3 Likes
Back to typewriter/keyboard ideas:
The “space” between the selected “keys” which also starts from death “4” in “descending” order. This also assumes that four lines does not give 4 words; I subscribe to that notation whether this is a correct clue or not.
4う, 8ゆ, -ほ
Rす, Tか, Yん, Uな
け,} む
Zつ, Xさ,Vひ,Nみ <ね,>る,?め、ろ
The keyboard share also reminds of some emoji shapes, perhaps upside down frowning. I’ve seen Koichi use this a couple times and it popped up in the Tsujigiri article from above. Just an observation, nothing concrete.
@spamneko’s keyboard:
3 Likes
I’m out of ideas. My mind keeps going back to the results found in the SPOILER thread from the book cipher (i.e. the three (not four) words that were found), the repetition of words (‘haha’, “Chop!Chop!”, “Durt!Durt!”) and that lightning gif that @a-regular-durtle posted in the last clue/not-clue. But no inspiration pops out.
I think we’re down to randomly guessing four word strings.
Except we don’t even know if the password is in romaji or kanji.
1 Like
Given the relatively small sample of coded text to work with, the frequency of each character does match the frequency of English characters reasonably well.
If someone has the time, it might be worth writing some code to generate the top N likely character substitution permutations – only populating the top 7 or 8 characters – and see if any of the results look like they might be readable.
… though Koichi’s talk of creative/unique ciphers makes me think that may not be the right track. And there’s still the unlikelihood of the double-characters.
2 Likes
I did that with an online substitution cipher solver. Nothing even remotely interesting, but it only used a single alphabet. A polyalphabetic cipher would be beyond the capability of the solver.
2 Likes
So in the next (or part of this one, who knows) step, part of the trick was recognizing ISBN format, not decrypting. What if the four lines are some other standard identifier format? I can’t think of what though. UPC, etc. It is curious that they’re 6-element multiples in length and a very narrow character set (only caps, numbers, and a couple of odd characters)
3 Likes
Yeah, spent a bit of time trying to break it up into a header character, footer character, and six-character chunks in between, but I wasn’t entirely sure what to do with that afterwards. Though, someone did notice further up somewhere that the pairs formed by taking the last character of one code and the first character of the following code can all be found within the first code.
2 Likes
Can’t stop thinking about how absurdly symmetrical the pattern of keys used is. I guess there must be some Japanese keyboard arrangement that puts the most common kana in that pattern, maybe? Seems too symmetrical to be a coincidence. But, if the actual pattern of keys is important, why repeat letters?
I did the marking over again with a US-style keyboard, including separately for each of the four codes, but… I don’t think I’ve learnt anything from it.
All codes:
Code 1:
Code 2:
Code 3:
Code 4:
3 Likes
That actually sounds like a poetry format. There are forms where you repeat elements of previous lines in a structured way…but I’m not sure that’s all that useful here.
3 Likes
Koichi’s reward for the one who cracks the code
9 Likes
I didn’t notice it, but maybe it is a clue pointing us to マ ?
1 Like
Here are the Keys that all Codes have in common:
4 Likes
9 Likes