this is the way

You can always take it to the next level by using ideas from von Neumann to avoid potential bias in your dice: Take any even‑length list of D6 outcomes (e.g. a 120‑length vector) and pair them sequentially (N0, N1), (N2, N3), etc. They emit one bit per non‑tie: 0 if the first roll < second roll. 1 if the first roll > second roll. Discard pairs where the two rolls are equal. Repeat until you have sufficient bits for desired key entropy. You probably don't need to do this, but it was a fun rabbit hole for me. I'm almost certain that e.g. seed signer et al do not use this approach and rather just ingest the raw die roll results.

What is this combo

Its for generating your seed without a PC and then stamping it to metal

can't even trust the dice haha

I'm light of recent event consider the following.... View quoted note →

Are those dice evenly-weighted? 🎲

Asking the important questions.

Hear me out, but there are at least one more level of paranoia you’re missing: If you put the dice in the machine to produce the seed, you’re still trusting the dice-to-seed mechanism. How do you know that those rolls go to THAT seed?

Cross compare air gapped wallets

I hope so