 Show Posts
|
|
Pages: « 1 [2]
|
|
Yes BitMaxz, of course I've read your old post.
Thanks for the answers BitMaxz& nc50lc!
|
|
|
|
I'm sorry, but you're actually talking out of your ass here.
Why would you include the cost of seed plates when using a hardware wallet, but not when using a laptop? It makes no sense. Either you factor in the cost with both or with none. I prefer none, because I won't assume how someone secures their seed based on their wallet choice. Sure seed plate is not needed, but that is the recommendations I use to see. Why not when using a latptop? Because with bitcoin core you just backup a wallet.dat You'd advise to do backups; sure, you can do that. But reliability of laptop storage would mean you might be looking at replacing that crappy old laptop HDD short time after setting it up, so having to buy a new one (or upgrade to SSD) will again cost you time and money to install, set up and restore. Lots of hassle. Backups can be made in USB Flash Drive, DVD, SD Cards etc Not 'many of them can be hacked easily'. The hacks that were possible, weren't too simple to perform (took multiple hours & good equipment + knowledge) and they were only on hardware wallets of the 'first generation' (without secure element) and on firmware versions that are ancient by 2022. I believe the Trezor hack utilized a 2016 firmware; that would be 6 years ago now. I also explicitly mentioned that I'd prefer a HW wallet with secure element over a laptop when it comes to physical attacks. The way to exploit the trezor one is the same for the last version, both can not be fixed. That doesn't require a bug in Bitcoin Core. Assuming the same attacker model (access to the device for a reasonable, but equal amount of time), he'll just need to pull the HDD out of the laptop, while he'd have to perform a pretty advanced low-level hardware attack on the hardware wallet and might not pull it off in time for the owner to notice the loss, restore the seed and move the funds. That completley goes out of the point, the bug was in ledger software making to loose funds because of change, so again, I do not imagine this kind of "bugs" happening in the most reviewed client such as Bitcoin Core Source? And even if they did exist, as you say: it would be pretty complicated. Probably buying you enough time to move your funds. https://www.youtube.com/watch?v=s3f1zNpzINYAnd check their blog for more info how technique was improved recently. Well first, you can put in a passphrase on a hardware wallet as well, so that's not an argument. And while you say more attacks will be developed, also better SE chips will be developed. It's a cat-and-mouse game and the attacker is always going to be a step behind. Until now, even the oldest of secure elements used in hardware wallets are secure, to the best of my knowledge. So now we assume that if HW is seized can be compromised, so let's add a passphrase, if you add a secure one entering will be so tedious in some models or nearly impossible to be easy of use, and in other HW you will have to type with the PC which is not a safe practice. Again BS: falling for phishing sites and giving away your seed words isn't due to the wallet being a HW wallet. That can happen with literally any wallet. How can this seriously be considered an argument against hardware wallets?
Or is it more leaning into whataboutism such as: 'Well, attacks that a HW wallet can't protect you against, do exist, so they're useless'?
There are reasons against hardware wallets, but the ones you present are so weak, it's ridiculous.
Falling for scams is not HW fault, what I just said is that someone educated is less likely to fall for it, I do not see how ridiculous is this statement.
To store several copies of the wallet.dat you'll need several airgapped devices. If you store them all on the same device it's as good as having no backup at all.
Yes correct, several devices. Given full physical access and sufficient expertise the seed can be extracted from certain Trezor models, but AFAIK no such attacks have been successfully mounted on Ledger devices. Feel free to bring me up to date in case there's been any successful seed extractions from Ledger devices recently. Recently I am not aware no, old ledger models yes but since nano X not that I am aware. I guess you're referring to paper wallets? Since hardware wallets are commonly deterministic it's technically not possible to lose funds to unknown change addresses (though if you have any more info on that I'd love to hear it). That was on ledger, in version 0.13 if not wrong. Just search ledger lost funds chane address, you will find all information about it. Seeds are also used by Bitcoin Core and other software wallets so that doesn't make a difference. Except, a lot of phishing sites ask for private keys directly rather than the seed, which for a regular user is impossible to obtain from the hardware wallet directly. Pishing sites use to ask mnemonic seed, Bitcoin Core uses bip32 and backup is a wallet.dat, so there is no mnemonic words to backup. * The average Joe will not save 400+GB on his HDD. He doesn't understand pruning, he doesn't want to wait days for the initial sync and doesn't understand why should he download such amount of data.
* The average Joe probably has Windows or mobile phone he wants to use Bitcoin with.
* The average Joe is not tech savvy and he would make mistakes if we would advise him use cold storage. I've seen case when one was claiming he has cold storage which he connects to the internet when he send transactions.
(Yes, he got his money stolen by some malware/exploit).
Bitcoin core is great, but it's not for everybody.
Bitcoin core, if used as hot wallet, is as vulnerable as any hot wallet.. You are right, maybe I was wrong assuming that average Joe would be interested in learning more or could have concerns about delegating security. If computer is compromised yes, it will not be safe, however I was referring in to use bitcoin core offline. You don't see the average PC user do this, right? The average user who gets confused when the internet icon has moved sure can't handle this. And even if they can do it, it's a lot of work for making a transaction. That's okay if you do it once a year, it's not okay for daily use. That was a good one  |
|
|
|
You're right; an airgapped laptop running Bitcoin core will offer similar security against a lot of attacks and it costs nothing if you already have a second computer that you can 'sacrifice' for this. I say 'sacrifice' because if I were to use such a setup, I'd rip out any wireless connectivity options permanently and also never run anything else on it ever again.
There's a few points against it, though. Firstly, if you do not have such a second device 'for free', purchasing a reliable laptop will be more expensive than purchasing a hardware wallet.
Old hardware also has the risk of breaking down more quickly, especially if that device still runs a HDD, which would mean a loss of funds or need to buy another device and restore the seed.
They also don't have secure storage chips unlike most good hardware wallets. This means a HW wallet can withstand a physical attack, when the attacker gets their hands on it, while a laptop in the hands of an attacker can quickly lead to compromised private keys.
In fact, a recent MacBook would be the best choice against a physical attacker, since it's much harder to access the solid state memory on them than on basically any other desktop computer. That would most definitely be more expensive than any hardware wallet, though.
Besides the hardware aspects, there's the issue of user error. What seems trivial to you and me, is going to be super hard for the majority of people out there. The risk of them skipping a step or doing something a bit differently because it seems to them just as secure, is very high. For instance, people are already taking pictures of seed words even though the software tells them not to - just one example. We saw already on this forum that they may believe disconnecting from the Wi-Fi is equivalent to desoldering an antenna (which sounds hilarious to us).
And finally, even if someone is able and has the laptop to spare, it's just more convenient to use a HW wallet with same or higher security. By spending a hundred bucks on a HW wallet, I don't have to modify a laptop, set it up, download the blockchain, go through all those security measures. I also simply may not want to have a huge device (compared to the hardware wallet) and its charger laying around (harder to conceal, takes more space, harder to carry while traveling / moving).
Yes buying a laptop will be more expensive, however when you see how much some hardware wallets cost + seed plates if they buy (it is the setup that I see most often people talking about) price does not go too far from a cheap laptop, also about storage causing lost of funds we may assume that several copies of wallet.dat will be stored, with a strong passphrase it would not be that easy or even possible to crack it in a reasonable time, allowing funds to be moved before they are able to crack. That can not be said if they find the seed or get the HW seeing how many of them can be hacked easily, even worse if firmware is not updated and still have vulnerabilities, and to not mention bugs that may even cause loose of funds, I can't imagine Bitcoin Core having bugs with change addresses to cause loose of funds as we ever seen with HW for example. There are some complicated techniques to even hack SE of HW, I do assume that in future because of more HW adoption those techniques will be developed, being less far secure than we may think, like we see right now some forensic labs unlocking iphones, androids in matter of seconds but brute forcing a strong passphrase will be always more complicated I think. About user error you are right, however someone taking the time to learn and do all the process will be less likely to commit error, we see some users loosing all funds from HW because they fall for pishing sites asking seed etc. |
|
|
|
1. If you want 100% security with bitcoin core or electrum, you need an air-gapped device for generating the wallet and signing the transactions. Not all people have an air-gapped device.
2. Exporting the unsigned transaction and signing it on a air-gapped device may be difficult for many people. Most of the people want to be able to make transactions easily.
3. If you use bitcoin core, you need to download the blockchain. That's why many people prefer SPV wallets.
4. Many people want a multi-currency wallet for holding their altcoins.
5. Most of the people don't know how running their own full node can be helpful to them.
1. I think most of people will have 1 on old computer which will do the job. 2. Yes valid point, that is why I've mentioned coldcard as example since it would be the exact same steps which may seem difficult, instead of using the SD CARD could be a USB Flash Drive, but for other HW like trezor, ledger you are right. 3. Valid point too, many people may prefer it at beginning... but as soon as they start to learn more, they will understand why it is absolutely important to run a full node (which leads to your 5th answer since running a node and not using does not makes too much sense). 4. Yes, if they are interested in altcoins sure. |
|
|
|
|
I think I've found the answer about 3000, since listdescriptors shows 3 kind of addresses and each descriptor has range from 0-999, so 1000, that makes the 3000
|
|
|
|
|
I am asking this question since whenever I see someone asking about how to store bitcoin, all the answers I read is buy a HW, but I rarely see someone advising to just use Bitcoin Core... the most reviewed and secure client? Those two arguments should be enough to convince people isn't it? When digging into bitcoin, we all know how important is to run a full node, and a big part of HW they may run Bitcoin Core as node, but not a lot use them as a wallet, and even worst, some they not even use their node, the one that auto connects with the HW.
I do understand that there is a lot of marketing involved and they want to sell it, but in reality what offers HW vs Bitcoin Core?
Hardware wallets generate the private keys offline, and allow to sign transactions with your keys offline in a friendly way.
So Bitcoin Core allows to do the same thing, most people have an old pc that they do not use, they can just install bitcoin core, generate a wallet, export descriptors to the online PC (the daily use PC), there you create a watch-only wallet, import descriptors and you will have a functional watch wallet (allowing to create new receiving addresses and create unsigned transactions) after a transaction is created you just copy the psbt file in a USB flash drive, bring to the offline PC, sign and bring back to the online PC to broadcast.
At the end what are we doing extra? Some air gaped wallets like coldcard we create the unsigned tx, copy to SD Card, sign with the device, and bring back to the online PC to broadcast. We are doing exactly the same thing or can be even more since if we don't use it directly with Bitcoin Core and instead we use electrum, we will need also need to have a server like electrs which makes the bridge between Bitcoin Core and electrum.
I am not against any HW (just mentioned that one as example since it offers an air gap setup similar as what would do with an offline PC) but comparing to any HW we have on the other side the "official" client, the most reviewed client, secure and free. Any company will review Bitcoin Core since they use it, but that won't be the other way, why would a programmer involved in an open source project such as bitcoin would review for free products of a private company? I am talking about the security of device, firmware updates, etc.
Let me know your thoughts.
Thank you.
Edit: Sorry, wanted to post in bitcoin and not bitcoin tehcnical support (since I do not think it is the correct place) but I can not delete it
|
|
|
|
|
Hi
After creating a wallet with bitcoin core 22.0 and descriptors true, when typing getwalletinfo it shows:
{
"walletname": "wallet",
"walletversion": 169900,
"format": "sqlite",
"balance": 0.00000000,
"unconfirmed_balance": 0.00000000,
"immature_balance": 0.00000000,
"txcount": 0,
"keypoolsize": 3000,
"keypoolsize_hd_internal": 3000,
"unlocked_until": 0,
"paytxfee": 0.00000000,
"private_keys_enabled": true,
"avoid_reuse": false,
"scanning": false,
"descriptors": true
}
I thought the default keypool was 1000, so now the default has changed to 3000? What also surprises me is the size, having a keypool of 3000 and only 20Kb.
Thank you
|
|
|
|
@nc50lc tutorial updated with the step to add the change descriptor and a mention for your help  |
|
|
|
Anyways, since we're at it, I'd like to add a step to that tutorial to include the change (internal) addresses so you wont have to manually add a change address in every transaction that needs change.
To get the "parent descriptor" for the change addresses, use the command listdescriptors.
The one for the change addresses has "internal": true," result | also pick the script type of your choice.
Then import it the same way you've imported the receiving addresses' parent descriptor.
Sure, I am going to update the tutorial. So yeah with listdescriptors I find 2 parent descriptors with the same path 84' one with internal false being used as receiving addresses used in tutorial and another with internal true for change addresses. I will do one more test and update it. Thanks |
|
|
|
Alternatively, you can refill the keypool on-the-fly with the command: keypoolrefill <new keypool size>You can also set your desired keypool size right when you import the descriptor, it should be inputted in the value of \"range\": [start,end]. ( e.g.: \"range\": [0,2999]) Assuming the following setup:
- Dedicated offline PC with bitcoin core, wallet created and export parent_desc
- Daily use PC with bitcoin core synced, parent_desc imported with importdescriptors, it allows me to create receveing addresses and create unsigned tx, so after creating a psbt, bring to the offline pc, sign and bring back to online pc to broadcast.
-snip-
Have you tested the set-up yet? Problem with the setup is: the offline PC wont be able to sign those transactions. Without the blockchain, " signrawtransactionwithwallet" will fail and you'll get the error message: "error": "Input not found or already spent" regardless of the presence of the associated key or descriptor in the wallet; also with " walletprocesspsbt", it wont be able to sign and will return with " false". The workaround is to plug an external drive which contains the blockchain so the offline PC can scan it whenever you need to sign a new transaction. Hi first, of all thanks the tip! Yes I have tested the setup of course. PC offline is able to sign transactions without blockchain since it has all info needed thanks to parent_desc and psbt. I've done a tutorial step by step if you are interested. https://medium.com/@ChangTraca/bitcoin-core-as-cold-storage-sign-transactions-offline-and-generate-receiving-addresses-from-the-f0b2de8436b5 |
|
|
|
|
Hello
Assuming the following setup:
- Dedicated offline PC with bitcoin core, wallet created and export parent_desc
- Daily use PC with bitcoin core synced, parent_desc imported with importdescriptors, it allows me to create receveing addresses and create unsigned tx, so after creating a psbt, bring to the offline pc, sign and bring back to online pc to broadcast.
When importing descriptors, the default keypool is 1000 so I do understand that 1000 addresses are automatically generated by default.
My question is, what would happen after 1000 addresses are used? Since all addresses have been generated using the watch wallet only (with the descriptors imported), offline wallet has no more data than the original 1000 keypool, will it able to find the key to sign? Or since I have past the keypool limit of 1000 I would have to generate more addresses in the offlinewallet in order to be able to sign?
Thank you.
|
|
|
|
|
