Hacker News Reader

Comment by runeks | Hacker News Reader

This is exactly why I’m hesitant to buy a hardware Bitcoin wallet. Not because I know they are insecure, but simply because I don’t trust them to be careful enough when they have no liability in case they make mistakes. For now, as far as I can see, the best bet on offline Bitcoin wallet security is a WiFi-less Raspberry Pi-like device, with a “USB-stick interface” (for transferring unsigned/signed transactions) and USB auto-run turned off.

We really need a hardware wallet company to partner up with an insurance company, such that the incentives of the users and wallet-producers are aligned. Claiming you have a secure device, while at the same time disclaiming all responsibility for money-stealing bugs, seems disingenuous to me. “Our device is much more secure than anything else out there, but we’re not willing to risk a dime to support this claim”.

View on HN

Same here, raspi works well for me (I taped over the Ethernet port so I don't accidentally connect it at some future date)

I think the only issue comes with the future case where I'm actually trying to transact in crypto, rather than just collect (speculate) long-term assets. I suppose at that point, I'll just use a software wallet on my phone and keep much smaller amounts there (much like savings vs checking account).

I wish it was easy enough to derive a Bitcoin address from a private key using pen and paper so that I could just make up a private key, write it down, work out a few details, and then have a Bitcoin address I could provide to anyone... with only human error as a possible problem or attack vector.

Bitcoin addresses are derived from the public key, which in turn derives from a private key. So you can already do what you propose, as long as your private keys satisfies the strength that Bitcoin requires.

The math is not trivial enough to do with pen and paper.

If it was, how secure would Bitcoin be? How much more vulnerable would it make it to brute-force attacks?

There's nothing stopping you from applying the key derivation algo step by step using a calculator, or an offline computer.

It would make for an insightful, albeit time-consuming, experiment.

https://bitcoin.stackexchange.com/a/29880

> If it was, how secure would Bitcoin be?

It doesn’t matter how easy/simple it is to derive a public key from a private key. What matters is how easy it is to go in the opposite direction. As your link shows, it’s not that hard to do elliptic curve multiplication by hand, and there’s no need for it to be hard/complex.

Why do you prefer physical hardware over plain paper?

The selling point of hardware wallets is that they do the signing for you, in an assumed safe sandbox.

I believe to make an analog with paper, you would:

- Spin up an offline VM with some bitcoin wallet software capable of doing offline signing.

- Import your private key

- Import the transaction data in a controlled matter, e.g QR code

- Sign the transaction

- Transport the signed transaction back to an online device in a controlled manner, e.g QR code

- Publish the transaction with the online device

Ah, that makes sense. However, if you're not interested in spending, just storing for a long period of time, they are more or less equivalent in terms of security, I'm assuming?

Yes, I'd say so.

You would probably not own a hardware wallet without storing the keywords for recovering the private key on paper (or similar). At that point it is more or less how you store the paper that will affect your level of security.

One use case to consider is that your successors in interest are the ones doing the spending. I.e., you are dead. It would be such a shame if they typed the private key into the family's malware-infested Windows PC and then wondered why the balance was zero.

A hardware wallet firewalls your loved ones from that risk.

Reply on news.ycombinator.com