So with the cold wallet hack I am thinking about using core 29 to store.

[nc50lc]

Isn't there an option to prompt your Trezor to enter the passphrase on the HW's screen itself?
Most attack vectors wont work if your use that instead.
~snip~

Is it even possible for such sensitive information to be entered through a user interface?

In Trezor Suite?
It should be enabled in the settings and there should be an option to create Passphrase wallets,
then the passphrase will prompted in the UI or a button to enter it on the device when you need to create/access a wallet with passphrase.

AFAIK, only older models don't have that option to enter the passphrase on the hardware wallet's screen.

[Cricktor]

but don’t lose it or you are fucked.

It certainly sounds like a broken record: if you use one or more mnemonic passphrases, make sure to very carefully and redundantly document them! Because if you loose or can't reproduce a mnemonic passphrase, you instantly loose access to your coins, which you expressed more casually.

It may sound obvious, but some people screw up the basic documentation. Human brains will inevitably forget complex mnemonic passphrases, so proper documentation is mandatory.

also the trezor has its seed my trezor 5 uses 20 words

IIRC, I don't use modern Trezors yet, it's Trezor's SLIP-39 new standard recovery word backup which encodes additional metadata and uses Shamir Secret Sharing (embedded entropy is 128 bits equivalalent to standard 12-word BIP-39 backup).


I hope other hardware wallet companies didn't fuck up their code paths for entropy generation and properly use embedded TRNGs in contrast to Coinkite's Coldcard mess. At least some companies did code audits, but those don't guarantee you find every flaw in your code base.

[philipma1957]

but don’t lose it or you are fucked.

It certainly sounds like a broken record: if you use one or more mnemonic passphrases, make sure to very carefully and redundantly document them! Because if you loose or can't reproduce a mnemonic passphrase, you instantly loose access to your coins, which you expressed more casually.

It may sound obvious, but some people screw up the basic documentation. Human brains will inevitably forget complex mnemonic passphrases, so proper documentation is mandatory.

also the trezor has its seed my trezor 5 uses 20 words

IIRC, I don't use modern Trezors yet, it's Trezor's SLIP-39 new standard recovery word backup which encodes additional metadata and uses Shamir Secret Sharing (embedded entropy is 128 bits equivalalent to standard 12-word BIP-39 backup).


I hope other hardware wallet companies didn't fuck up their code paths for entropy generation and properly use embedded TRNGs in contrast to Coinkite's Coldcard mess. At least some companies did code audits, but those don't guarantee you find every flaw in your code base.

yeah I am going to make 10 passphrases using a set of punches

the punches are
A to z     thus 26  
 0 to 8    thus  9
and *     thus 1

that is a set of 36 different punches.


shake them in a box pull 1 punch
replace it in the box pull 1 punch

so 1/36 x 1/36 until I do  a large passphrase

2 pulls is                       1296    combos
3 pulls is                     46656    combos
4 pulls is                 1679616    combos
5 pulls is.              60466176     combos
6 pulls is           2176782336    combos
7 pulls is         78364164096    combos
8 pulls is 2,821,109,907,456     combos


now I can fit 8 punch marks on a stainless steel washer.    
and 1 washer means     2.82 trillion   which only slows the attacker some what
but 2 washers mean     2.82 trillion squared better
and 3 washers means  2.82 trillion cubed  even better


now if you have a 20 word seed and 3 washer 24 symbol passphrase you have a decent stoppage of a hack.

[PrivacyG]

My opinion is.  For both philipma1957 and everyone else reading.  I understand the Coldcard situation made a lot of people nervous.  Including me.  But I just replied to another situation on the Wasabi Wallet Topic a short while ago (https://bitcointalk.org/index.php?topic=5476197.msg67012172#msg67012172) and I see this trend of people all of a sudden moving their Bitcoin to a 'safer place'.

Please do not over complicate this!  If you have a Coldcard with some Bitcoin on it then hurry up and move them to a safe place but other wise do not try to re invent the wheel and keep in mind that caution is still very important!  The person I linked above to lost 6 Bitcoin because they moved from Coldcard to a fake Wasabi Wallet.  Others are probably going to lose by moving from a safe place like a Trezor to a hot wallet on an infected computer and lose every thing instantly.  Or some will insert their Seed on a random web page with the same consequence when the Bitcoin they had was in a secure place already.

Do not rush this process!

[philipma1957]

My opinion is.  For both philipma1957 and everyone else reading.  I understand the Coldcard situation made a lot of people nervous.  Including me.  But I just replied to another situation on the Wasabi Wallet Topic a short while ago (https://bitcointalk.org/index.php?topic=5476197.msg67012172#msg67012172) and I see this trend of people all of a sudden moving their Bitcoin to a 'safer place'.

Please do not over complicate this!  If you have a Coldcard with some Bitcoin on it then hurry up and move them to a safe place but other wise do not try to re invent the wheel and keep in mind that caution is still very important!  The person I linked above to lost 6 Bitcoin because they moved from Coldcard to a fake Wasabi Wallet.  Others are probably going to lose by moving from a safe place like a Trezor to a hot wallet on an infected computer and lose every thing instantly.  Or some will insert their Seed on a random web page with the same consequence when the Bitcoin they had was in a secure place already.

Do not rush this process!

Yeah be careful do it step by step.

As of today I have trezor and core holdings.

I also cashed in a good amount of my coins as I will not be able to take a huge hit and get the value back as I dont earn a ton mining. Plus at 69 my time line is likely 20 years or less.

[LoyceV]

yeah I am going to make 10 passphrases using a set of punches

the punches are
A to z     thus 26  
 0 to 8    thus  9
and *     thus 1

Can you easily distinguish between the ones that look similar, like O<>0 and I<>1 ? If not, it's better to leave them out of the pool.

[philipma1957]

yeah I am going to make 10 passphrases using a set of punches

the punches are
A to z     thus 26  
 0 to 8    thus  9
and *     thus 1

Can you easily distinguish between the ones that look similar, like O<>0 and I<>1 ? If not, it's better to leave them out of the pool.

one is  1
I     is  l

so they are easy

0 and O may not be easy

I am waiting on a brass plate for fast easy back up sheet {in house}

the washers in bank vault

[LoyceV]

the washers in bank vault

Somehow trusting banks with your Bitcoin doesn't feel right to me.

[nc50lc]

the washers in bank vault

Those only contains the passphrase, right?

I see its advantage over keeping it in the same vicinity where you stored the seed phrase,
But at least put a reliable tamper-proof seal for reassurance that it'll be obvious if they got curious on its contents.

[philipma1957]

the washers in bank vault

Those only contains the passphrase, right?

I see its advantage over keeping it in the same vicinity where you stored the seed phrase,
But at least put a reliable tamper-proof seal for reassurance that it'll be obvious if they got curious on its contents.

Yeah I would want the colder stored btc passphrase on a washer bolt nut setup in a safety box.

I have the parts with the exception of the brass plate.

So the passphrase should be in 2 spots.

The seed in two other spots.

My trezor 5 has a tested 20 seed setup
It has 2 passphrase accounts
Soon to add a third.

[takuma sato]

This is just my personal opinion and it may be controversial but it's my take: If you buy a laptop and install Debian on it and only use it for Bitcoin using Bitcoin Core, and you only use it to sync the blockchain and then copy-paste your wallet when you need to make a transaction, and once done you delete the wallet (after making a copies somewhere offline) that is a step higher than any hardware wallet. I have always thought hardware wallets are just a terrible idea, it identifies you as a Bitcoin target and you are trusting some third parties with so much stuff that the split second you are exposed to the network doing this remains a safer option.

Of course, ideally you want to have 2 laptops, and learn how to use PSBT transactions. Even tho user experience is still annoying and convoluted, that is the most ideal setup.

[Cricktor]

...

In your Debian laptop setup, why do you entertain the nonsense of copy/delete the wallet? What security improvement do you expect from it?

This laptop is online periodically and Bitcoin Core has to go online too, to sync the blockchain.

There could be some vulnerability by which a hacker could somehow infect the system though the chances are low when you just do only what you outlined as usage pattern (requires some discipline of the user).

Your copy/delete wallet playbook has some risk if users screws this up. And I don't think it's necessary or would improve security of the wallet.

If the device got infected somehow, copying back the wallet, opening the wallet, typing the wallet decryption passphrase will expose the wallet and private keys to malware on the device (if the malware is sophisticated, but let's assume exactly this).

A software wallet can't really protect the keys from system level malware on the device when the user enters the credentials required to decrypt the private key(s) in Core.

A hardware wallet on the other hand doesn't give malware on the computer a chance to attack private keys which the hardware wallet holds encapsuled without any API (cough, except for Ledger Recovery crap) to access them from the computer side. This is the main advantage of a well designed and executed hardware wallet: even an infected computer can't make the attached hardware wallet reveal any private key, seed, extended private key to compromise the wallet.

[philipma1957]

if you have core wallet and a long passphrase

you can go on line synch the core

go off line put in the password .

then go online and do a send

there are a few ways to make core wallets with small amounts

which allows for safe paying.

but means 3 or 4 core wallets

[LoyceV]

go off line put in the password .
then go online and do a send

You can do this, and it will probably work just fine, but doing so makes it a hot wallet instead of cold wallet.
When dealing with cold storage, I always assume malware (let's say a key logger) would patiently wait until it gets an internet connection, and upload your data.

[philipma1957]

go off line put in the password .
then go online and do a send

You can do this, and it will probably work just fine, but doing so makes it a hot wallet instead of cold wallet.
When dealing with cold storage, I always assume malware (let's say a key logger) would patiently wait until it gets an internet connection, and upload your data.

yep true.

lets pick a btc amount for holdings say 2.0 btc

put 1.6 btc on a trezor use 3 passphrase wallets and the standard

0.5 on the passphrase wallets
0.1 on the standard

all cold storage

you have 0.4 left

some on exchange some on core 29 or core 28 wallets.

[asUHWEceyc]

Since v17:
https://github.com/bitcoin/bitcoin/blob/master/doc/offline-signing-tutorial.md

For Austin Powers:
https://github.com/bowlarbear/yeti-2.0

[nc50lc]

Since v17:
https://github.com/bitcoin/bitcoin/blob/master/doc/offline-signing-tutorial.md

This is a good suggestion,
But take note that the latest version of that tutorial is designed for the next version of Core which isn't released yet.
philipma can follow it if he can build a binary based from the latest source code though.

I've also suggested this Bitcoin Core Air-Gap setup in my first reply, BTW.
He seems not to be interested so I didn't follow-up with the specific steps without the exportwatchonlywallet command.