Casino dice are probably not even necessary. But, for about 15$, you can get better dice with the holes filled in so they're better balanced. For the extra peace of mind for my family's future, I'm happy to do it.

You don't need Coleman's tool. Just use paper : https://github.com/veebch/Bip39-Dice

Casino dice. And you don't need Coleman's tool. Just use paper : https://github.com/veebch/Bip39-Dice

Yes!!! 110% this. Thaaaaank you!

I hear this "dice are imperfect" thing, and I think to myself, "ok, by how much? and that leaves me how much entropy at the end of the day?". And it's way more than enough.

Now compare that to the catastrophic risks of a bad RNG or bad implementation around it.

Plus, why have just one set of 24 words? Who can't spare a couple of hours to roll a bunch of them? So that when the "imperfect dice" fail you, you're still orders of magnitude better off than with the badly implemented RNG.

If people tell me they can't spare a few hours to roll dice, they better also tell me they never watch netflix or sports.

I'm looking forward to it. I've become too uncomfortable with the direction Core has taken. I'm willing to be on a chain which gets a block every few days if need be. No rush. I think we'll look back on that fondly in ten years.

? So... trust me bro, "dice rolls are imperfect" ?

Rolling good dice is the best option we have. Remember the debian ssh bug? And the random number generator problem in Tails? And now coldcard? For kicks, just ask Kimi K2.6 to list bugs related to RNG that have happened in Tails and in the linux kernel.

A printed bip-39 wordlist. Rolling casino dice, odd number = 1, even number = 0 ; get 256 bits like that. Seems pretty good. Last word calculated on a seedsigner, a jade or an offline live-dvd computer.

There's also the pairing method : rolling twice the same die, discard the result if the two numbers are the same, if the first roll is higher than the second, that's a 1; if the first time is lower than the second time, that's a 0. That seems really good too.

I disagree.

I'm just writing this message as an answer to someone else bashing dice but I'll repost it here :

Rolling good dice is the best option we have. Remember the debian ssh bug? And the random number generator problem in Tails? And now coldcard? For kicks, just ask Kimi K2.6 to list bugs related to RNG that have happened in Tails and in the linux kernel.

A printed bip-39 wordlist. Rolling casino dice, odd number = 1, even number = 0 ; get 256 bits like that. Seems pretty good. Last word calculated on a seedsigner, a jade or an offline live-dvd computer.

There's also the pairing method : rolling twice the same die, discard the result if the two numbers are the same, if the first roll is higher than the second, that's a 1; if the first time is lower than the second time, that's a 0. That seems really good too.

Why would rolling dice be nonsense?

Oh, I see, ok. I'm out then. I had found all the parts and was about to order online. Oh well.

FIPS is very interesting but I was looking at that and find this on their website : https://learn.fips.network/lessons/11-threat-model

- **No anonymity**. Direct peers learn each other's npub through the FMP Noise IK handshake. There is no mixing layer, no cover traffic, no tunnel rotation. Tor and I2P solve a different problem.

- **No traffic analysis resistance**. A global observer with vantage points on multiple transports can correlate flow timing and size patterns across the mesh to infer who is talking to whom. FIPS does not pad, mix, or batch to defeat this. Confidentiality is protected; the fact that a conversation is happening is not.

I look at trees while drinking my coffee in the morning. It calms me down.

Are you sure? I see him re-twatting BIP110 people too, knots people too. Seems more neutral than anything.