Trezor - a hardware bitcoin wallet, now open for pre-orders | Hacker News Reader

Trezor - a hardware bitcoin wallet, now open for pre-orders

bitcointrezor.com

137 pointsrplnt57 comments

Cool idea, it reminds of an IronKey but for Bitcoin; I'm frustrated that there isn't any solid or easily findable technical information about the device though!

When I click on "About Trezor" it takes me to...a brief overview? I would expect that link to be a much more thorough explanation of what Trezor is, how it works, and why I should buy it.

Why would Trezor be better than an IronKey? Is it a push button interface for sending bitcoins? Does it just store my wallet or is it running a bitcoin client? I read it works with most bitcoin clients, what does that mean? What are the buttons for?

If it's just storing my wallet what kind of encryption is the device using? Does the device have hardware-based tamper proofing (IronKeys will self-destruct after X number of failed passphrase entries, it will self-destruct if someone tries to physically tamper with it)?

Here's a short (30min) talk from Bitcoin 2013 conference http://www.youtube.com/watch?v=3t18a-wXBnw

And a bitcointalk thread: https://bitcointalk.org/index.php?topic=122438.0

Hopefully they'll add something to the product site itself.

>Why would Trezor be better than an IronKey?

An IronKey is designed to defend against it being lost or stolen. A Trezor is designed to defend against malware. Malware can submit a transaction to the Trezor, but it can't push the button on it.

>Does it just store my wallet or is it running a bitcoin client?

It just stores your wallet. Your Bitcoin client needs to have Trezor support for this to work.

>When I click on "About Trezor" it takes me to...a very brief overview?

Yeah, didn't they say it was open source?

> An IronKey is designed to defend against it being lost or stolen. A Trezor is designed to defend against malware.

Malware can submit a transaction to the Trezor, but it can't push the button on it.

Okay, that's interesting, thanks.

> When I click on "About Trezor" it takes me to...a very brief overview?

Yeah, didn't they say it was open source?

And? As far as I can tell they are a for-profit venture that is simply open-sourcing their product.

What I get from the video is that it stores the wallet - a set of public and private keys for your addresses, along with a list of transactions to/from those addresses - and signs transactions.

Bitcoin clients will implement support for reading the list of transactions to/from your addresses to figure out how much money you have. They will also implement support for using the device to sign transactions, so the private keys never touch your computer, and therefore a compromised computer cannot steal bitcoins without the device prompting the user to confirm the transaction.

As the device is a single-purpose computer, likely running the absolute minimum code necessary to implement transaction signing, it should be significantly more secure than your PC just there. And you can easily physically secure it in any way you want, whether that's carrying it around with you at all times or locking it in a safe. Anything over that is a bonus, in my eyes.

I do not currently use Bitcoins, at least partially due to security concerns, but with this, I would be very interested.

The transactions are stored by the bitcoin network, so the device probably only stores the public and private keys for your addresses, nothing else.

What happens when it breaks, gets lost, etc?

A paper wallet (copied to several secure locations) seems safer, if less convenient: https://en.bitcoin.it/wiki/Paper_wallet

EDIT: Thanks to rplnt[1] for the Bitcointalk forum post link[2]. According to one of the apparent Trezor creators, "No need for periodic backups, writing down the seed to paper during the device initialization will be enough forever"

[1] https://news.ycombinator.com/item?id=5880849

[2] https://bitcointalk.org/index.php?topic=122438.0

Generally speaking, most hardware wallets let the end-user re-initialize the device based on a seed (a random number). So as long as you save this seed securely (eg. printed on paper and stored in a safe) you can clone the wallet and recover your coins if you lose it.

I don't know if Trezor has this feature or not.

Thanks for your reply, mrb. When you say that the wallet can be cloned, would it require another physical Trezor unit? If so, that would seem to be a potential downside vs. a regular paper wallet.

It would not require another physical wallet. The algorithm creating the Bitcoin addresses from the seed does not need to be made secret, and could be run on a computer, for example, then you could re-import the Bitcoin addresses and the coins into a software wallet.

This is a very interesting idea. However I think the ordering process for this highlights a crucial problem with Bitcoin, instability of value. They aren't planning to deliver the device for four to five months. Using the last four months as prediction of future fluctuation in Bitcoin value (insert usual caveat about past performance predicting future performance), the more expensive version of the device could cost anywhere from $100 to $700 USD when you actually receive the product.

Yeah the value of a BTC could increase, but it could also stay the same or even decrease, somehow. I'm not sure on the economic theory of this, but the people who pay are basically saying, in X months I will value this device more than I value 1 BTC now.

real currency has the same issue as well.

this is dealt with currency futures and forwards.

You honestly think the US Dollar will change in value by 700% in a few months?

Tone down the lack of clue.

Imagine you are buying goods denominated in euros and it takes two weeks for them to be delivered. You don't want to incur any P&L due to the change in exchange rate between now and delivery.

The exchange rate changes all the time. You would use a currency forward to hedge out that change.

I don't understand this. Unless all of your money is already in Bitcoin, just buy exactly 1 bitcoin right now and immediately send it to them, rather than sending 1 bitcoin from your existing holdings (or send from your existing holdings and immediately buy 1 more bitcoin, it's exactly the same effect).

If you'd rather speculate on the price of that 1 bitcoin, then you're free to do so.

If Bitcoin ever gets to the point where it's used for everything the deflation should be much more reasonable and predictable. Then maybe companies will offer a discount for prepaid preorders.

If you pre-order with one Bitcoin today, then you paid the dollar value of one Bitcoin today, plus transaction costs. Future fluctuations in Bitcoin value are irrelevant.

1 Bitcoin today and 1 Bitcoin in six months is not the same at all because you lose all the value you could have add from this Bitcoin during the six months so you have to discount it. That's called present value of money and that's one of the way companies manage their investments.

Theoretically, you have to use the real rate of return of your money and discount it on the time period. Of course, estimating the real rate of return is an art in itself. Considering value of bitcoin has been shown to be highly volatile, doing any semi-long term investment with it is basically gambling.

That's true, but you're talking "just" about time value (interest), sig's comment was about the volatility of BTC in terms of USD, different things. You'd also lose the interest if the product was denominated in USD or another major currency.

I would agree with you if it wasn't a pre-order. You pay, you get your goods, deal closed.

With a pre-order on the other hand, you give them the money and you get nothing for four months. Probably longer.

I'm not sure I see the difference. Once you've commited your 1 Bitcoin at today's value, you have effectively paid that price, irrespective of when you receive the product.

If the value of Bitcoin fluctuates between now and the time you receive your product, then you may "lose" or "win", but that's a chance you take, even with the product in hand. In other words, there is still the possibility that your 1 Bitcoin could have allowed you to purchase something more valuable than the product in hand at some point down the road.

Seems that the only thing you give up with a preorder is the interest potential or use of the money for some period before receiving the value (i.e. product). But, that would be the case with a preorder irrespective of the currency used.

Yes, but in that case you jusr lose the intrest of 4 months on your money which is close to nothing and not something to worry about.

1-3 BTC is way too much. This should cost at most as much as a very cheap mp3 player. At the current USD-BTC exchange rate, this means no more than 0.2-0.4 BTC

I understand they have upfront costs, but they could work out a campaign where the cost per unit goes down as they reach certain sales goals.

I don't doubt the price will drop as they become more popular (it's probably a few dollars worth of raw hardware), but right now it's a fairly niche product.

"a campaign where the cost per unit goes down as they reach certain sales goals."

That's an interesting idea. It gives purchasers an incentive to promote the campaign, as opposed to "early bird" specials which incentivises early purchasers, but almost disincentivises later purchasers.

This has been done on Kickstarter and software bundles like MacHeist in the form of "unlocking" additional rewards, but I don't think I've ever seen them actually lower the price if they hit certain goals.

It seems as though this model will merely deter possible buyers. Why would I pay a premium now when I can wait a few days, let others incur the initial, inflated cost, and then purchase my thing at a substantially lower cost?

This phenomenon, where people wait until the last minute to go in on a product, is already pervasive on Kickstarter. (I'm not entirely sure why, but the most obvious reason is that people want to ensure that the project will definitely be funded or will attain some sort of stretch-reward before buying). Dynamic price-lowering will only exacerbate this trend, discourage initial purchases, and likely produce more failed projects.

You got it backwards. Think of a model where you initially commit to pay full price, but everyone gets a discount that increases as sales targets are met. It would be no different than any "refer-a-friend" discount that you see at gym clubs or satellite subscriptions.

I want one badly, but this is simply too expensive. $100 for that little device in plastic and $300 (3x more) for it in metal? Outrageous.

Think of this pitch as being less like a normal retail product and more like a self-organized kickstarter campaign. You are not paying for the device, you are paying for their living expenses and prototyping costs while they spend the months it will take to get the device ready for production. Hardware engineering is an unbelievably slow and tedious process if you are used to software development. On top of that, the material costs can be substantial. It is not uncommon for the cost of a pre-production prototype to be 50-100x that of the final manufactured version.

I happen to be working on a design for a small electronic device in my free time. It is a much simpler device than the Trezor, and the ultimate manufacturing cost per unit will be around $5, despite its simplicity, because I'll only be making a few dozen of them. Cost for the first prototype will be close to $90, accounting for my time at $0/hr. If I get it wrong, I will have to redesign it and build another one. Hardware is expensive.

>Think of this pitch as being less like a normal retail product and more like a self-organized kickstarter campaign. You are not paying for the device, you are paying for their living expenses and prototyping costs while they spend the months it will take to get the device ready for production.

This is why I'm not as infatuated with crowdfunding as everyone else. Because individual customers are supposed to supplant traditional investors, initial costs are much higher for customers because we have to subsidize research, cost-of-living, etc. Moreover, we also incur all of the risk (which is formidable, considering the volume of failed projects and the ubiquity of scams on many crowdfunding sites).

I don't understand why I, as a consumer, would prefer a system where all of the cost and all of the risk are precipitated onto me, instead of onto the country's investment apparatus.

Better all these guys spend some rich VC's money than spend mine.

It's the Tesla Roadster of Bitcoin wallets. Somebody's gotta buy it for the price to come down. (Heck, I've got nothing better to do with my BTC.)

Convert it to USD and buy shit?

The metal one is CNC machined according to this http://www.reddit.com/r/Bitcoin/comments/1gc5xh/trezor_presa... hence the price.

I dont pre order anything bitcoin related after numerous vaporware and the BFL saga which is still ongoing.

It's a shame that the standard terms in the Bitcoin world are now 100% paid 6 months in advance preorders. I realize that Bitcoin has attracted a lot of get-rich-quick suckers, but it seems like Bitcoin companies are trying to drive away any non-suckers. Or maybe this is just what a race to the bottom looks like.

> It's a shame that the standard terms in the Bitcoin world are now 100% paid 6 months in advance preorders.

It's not "a shame" because it's what individual market participants each chose to tolerate. If you don't like it, offer better terms and steal their business, but don't complain about other people's voluntary acceptance of transaction terms.

I don't think honest businesses can compete on a level playing field with companies that are indistinguishable from scams. Hence stuff like the FTC, although I realize many Bitcoiners would like to try a world without such regulation.

It'd be cool if you could do NFC or some other method, so you could couple it with a mobile device. USB is great, but some less wired approach makes this even cooler. (for v2, obviously.)

Something like this makes me think the future of currency is bitcoin, or something like bitcoin.

I'm working on an NFC card for Bitcoin for my final year engineering project. Aiming for a credit card sized final product.

This is one of the first times i've seen the round number price in BTC rather than USD.

I don't think people need yet another dedicated gadget that is no matter how secure - prone to be lost and potentially misused.

I'd probably steer toward Yubikey as it's more universal and solves the same problem - keyloggers and malware.

How does Yubikey protects you from malware? I'm genuinely curious as from the overview it seems to only store passwords?

As for the misuse, the device can be PIN protected (it's not much but saves you some time to transfer the bitcoins).

Yubikey does not store passwords, it generates them. It generates one time passwords for two factor authorization that expire very quickly. So if your password is stolen, the attacker would also need your Yubikey to generate a one time password that can be verified by Yubikey with your device. The Yubikey plugs in USB and acts like a keyboard typing in the password when you hit the button

I see. So it doesn't protect your private keys.

I can see firmware updates being a problem.

If the malware can hijack the firmware update process, then it copy itself onto the Trezor (making the device worse than useless.)

I assume it will display a fingerprint of the new firmware and the user will need to press the button (like with a BTC transaction.) Still, a few users will probably be caught out by this.

Unless they Tivoize it (then the "open source" claim is rather dubious, but I think this may be the best solution anyway.)

So don't buy a used one.

I'd Tivoize it and have a developer switch like Chromebooks. It goes without saying that switching modes must wipe the device.

What firmware update process are you talking about?

If there is no firmware update process that may be even more reason to avoid buying a used Trezor. No process to authenticate a device of untrusted provenance, yet bad actors could still figure out a way to tamper with the firmware (see the photo on the site of the unit in development with the cover off and an extra cable attached), or replace the board entirely.

A well-documented, well-understood firmware update process with mutual authentication (firmware and device must both be validated) might improve things for secondhand Trezor buyers.

A validation process is a good idea but I'd still feel safer if the device were frozen with no way to update firmware. Many microcontrollers have "e-fuses" which you can blow post-manufacturing to render the program flash read-only, for example. One can also pot the board in epoxy to make tampering more difficult.

What if the current firmware has a bug that allows an attacker to steal your keys?

In case anyone is wondering, trezor translates to vault or safe (or even treasury).

I'm a bitcoin skeptic, but I guess I don't understand why there aren't bitcoin....coins.

Seems to me it wouldn't be terribly difficult to throw some bitcoin info on an NFC disk.

Bitcoins are not an encoded entity, addresses (~"accounts") are. Amounts in addresses are inferred from the whole blockchain (P2P ledger).

https://www.casascius.com/

I completely failed to find the sources for this :(

Have they got any patents on this?

No, and it's going to be open hardware (and software).

Hell yeah, I'll give those boys with the funny accents a bitcoin.

Reply on news.ycombinator.com