Profile avatar
correctcryptotakes.bsky.social
Remember to always MAC-then-Encrypt, for your health
67 posts 26 followers 35 following
Regular Contributor
Active Commenter
Posts 11 Comments 39

I am a: ⚪️man ⚪️woman 🔘deer Looking for a: ⚪️man ⚪️woman 🔘forest of random subsets

submitted 80 days ago • 0 comments

polynomial not for the algebraic aesthetics but to min-max the number of cryptographers who groan when I make a polynomial of such high degree the Schwartz-Zippel lemma doesn't help

submitted 81 days ago • 0 comments

this is not investment advice but if you work in tech and want to hedge against the potential collapse of the tech industry you should look into being a deer and going in the woods

submitted 83 days ago • 0 comments

@correctcryptotakes.bsky.social found this; do you know a mitigation?!?!?!

submitted 83 days ago • 1 comment

In elliptic curve cryptography, supersingular curves with small embedding degree k are weak because their group structure lets the discrete log problem be moved to F_(p^k)*, where subexponential attacks like index calculus can solve it efficiently. This is good. You want this

submitted 84 days ago • 0 comments

Cyclic groups in crypto are trash. Big numbers in a loop? Yawn. Quantum nerds will shred it with Shor. Just use chaos-based systems or hash chains, not this "algebra" shit

submitted 86 days ago • 1 comment

Feistel networks are straight-up bullshit. Split your data in half and juggle it? Please. Just mash the whole block with an AES-style S-box and call it a day—none of this 'Left-Right-XOR-swap' nonsense. It’s only still around cause cryptographers are too nostalgic to ditch DES

submitted 90 days ago • 0 comments

Side-channel attacks? Just yell your private key louder than the power lines hum. The side channel will be the least of your worries

submitted 97 days ago • 0 comments

Real talk: Merkle trees are overrated. Just stack all your hashes in a single linked list and call it a Merkle stick. Same difference, less recursion. Fight me, @vitalik.ca

submitted 97 days ago • 0 comments

Bluesky sync? Just hash the diff with MD5, sign it with your birth year, and pray collisions don't break the Merkle tree apart. Easy

submitted 97 days ago • 0 comments

Perfected zk-SNARKs: CRS is all zeros, setup’s a coin flip I called "heads."

submitted 100 days ago • 1 comment