0.2 ₿ puzzle

[ibbonmailru]

Hi, it's me again. After a deep search, I found some clues and inconsistencies regarding this matter. This puzzle is real but unsolvable as it stands, due to the fact that it has a passphrase. I got 2 confirmed positions from someone, which I'm very grateful for, but I don't have the time or the software to try them. I don't know what else to do, just good luck to everyone who is still working on this!

Can you share that posotions?

[iss333]

Sharing my reasoning in case it helps or someone wants to challenge it.

Words I consider certain (9): moon, tower, food, black, this, subject, one, order, real.

1. Length: 12 words. Fewer than 12 I rule out: below 12 falls under the BIP39 standard minimum (128 bits of entropy), not a length a normal wallet uses. More than 12 (15/18/24) I don't see either: there aren't enough clearly-grounded words in the image to fill 15 or more without making them up. I'm betting on 12 words + a passphrase.

2. I rule out Electrum. I checked the words against the Electrum v1 list. Three of the certain ones —this, food and one— don't exist in that list, while breathe IS in Electrum v1 but NOT in BIP39. The two lists are incompatible: you can't have a seed containing both breathe (Electrum only) and this/food/one (BIP39 only) at the same time. Since those words are well-grounded, and Electrum v1 has no passphrase and uses a different derivation, I stick with BIP39.

3. The passphrase is BREATHE. breathe is prominent (Floyd's chest and the statue's neck), but it's not a BIP39 word, so it cannot be one of the 12 seed words. The only way to use it is as the passphrase. So passphrase = BREATHE, uppercase, exactly as written.

4. I don't believe there's any false information. Every deliberately placed element has a purpose. No decoys: runes, numbers, breathe… everything must be used for something.

5. The "reverse" is still unsolved. There's a clear mirror/inversion motif (the reflected seal, the swapped ties between Trump and Biden, ciphered text written mirrored). The reverse means something, but its exact role in the solution remains unsolved.

6. The order is irrelevant in practice. Even if the seed requires a specific order, nowadays permuting the set of words can be brute-forced in minutes with the right hardware. You don't need to guess the order: just find the correct words and let brute force find the valid ordering. The bottleneck is the words, not the order.

7. Position hints (low confidence). The "5HT" on the tablet leads me to presume black might go in position 5, though I give it little weight (ambiguous characters). And "THIS IS THE FIRST PREDICTION" makes me think THIS is the first word. These are hints, not certainties.

8. I consider "one" a word, not order. In the 13th Amendment text, "subject" is underlined and the "1" in "Section 1" is underlined the same way. Since that underline marks seed words (subject is one), I read the "1" as also pointing to a word: that's why I treat "one" as a word, not a position indicator.

9. Weak words. I'll note some weaker candidates. For example: in the junction box at the top right, next to the cameras, the same symbol appears as on the dollar seal —the eye inside the triangle/pyramid. I haven't seen anyone mention this on the forum. The fact that it shows up twice (seal + box) reinforces that it might point to a word: pyramid or eye.

[mohamedabb]

 For anyone still crunching away at the 0.2 BTC BLM puzzle, I have some good news and some bad news.

The good news: The rune cipher is solved. It's based on Cyrillic, not Greek. I've also eliminated 12 billion seed derivations (BIP39/Electrum) so we can narrow the search space significantly.

The bad news: If you've been using the standard community word list, there are 4 errors in it that might be breaking your generation scripts.

I put together a full repository with the Cyrillic mappings, the dead seed ranges, and the corrected word list so you can update your search algorithms.

Full analysis here: https://github.com/rarsn4/blm-0.2btc-analysis

[Cricktor]

Full analysis here: https://github.com/rarsn4/blm-0.2btc-analysis

(separated link from colon so it can appear as clickable link)

Why don't you use proper link BBcode? Like [url]https://...[/url]

I can't vouch for the code provided in the above repository. I only briefly looked over it and didn't spot something nasty. That doesn't mean to say the code is safe to execute, do your own due dilligence before you execute foreign code!

Thanks for openly sharing your findings.

[anaxi]

The amount exactly matches my loan balance. This means I'll be wasting a month trying to solve this. Thank you 

[mohamedabb]

Why don't you use proper link BBcode?

Fixed, thanks — https://github.com/rarsn4/blm-0.2btc-analysis

I can't vouch for the code provided in the above repository.

Entirely reasonable, and worth stating plainly: don't run it on a machine
holding funds. It's CUDA + Python, it needs no network access, and there is
no wallet interaction anywhere in it — but nobody should take my word for that.

If you'd rather not execute anything, the findings stand on their own and are
checkable in a few lines of Python:

Code:

from mnemonic import Mnemonic
wl = set(Mnemonic("english").wordlist)
print("breathe" in wl)                    # False -- cannot be a seed word
print([w for w in ("they","that","time") if w not in wl])   # [] -- all valid

The 24-word warning is the one I'd most like people to check, because it
silently voids results rather than erroring:

Code:

import hashlib
mn = "abandon "*23 + "art"
print(len(mn))        # 187 bytes -- exceeds the 128-byte HMAC block
# RFC 2104: key must become SHA512(key). Zero-padding instead is correct for
# 12 words, passes every 12-word test vector, and derives garbage for 24.
print(hashlib.pbkdf2_hmac('sha512', mn.encode(), b"mnemonic", 2048, 64)[:8].hex())
# 408b285c12383600 -- the canonical value. If your solver gives anything else,
# your 24-word negatives are void.

Canonical check: abandon x23 + art -> 1KBdbBJRVYffWHWWZ1moECfdVBSEnDpLHi at m/44'/0'/0'/0/0