Is there any cryptocurrency with absolute zero collision wallets?

[Ceres_123]

Hello! This is my first post here

I would like to know if a cryptocurrency that solves the hash collision problem of wallets/private keys does exist. I know the hash collision probability is virtually zero, but the possibility of two people getting the same wallet/private key by pure chance still exists, and if it is possible well... technically it can happen . Here I describe a collision:


COLLISION: *by salt i mean the passphrase, or 13th word, or 25th word (whatever you want to call it). And seed is the famous 12/24 word list.

Seed.A + Salt.A = Hash.C -> Wallet.C

Seed.B + Salt.B = Hash.C -> Wallet.C


So here comes my question: has any cryptocurrency solved the collision "problem" up to today? I dont think it should be hard to solve, just adding a secret unique ID service somewhere in the blockchain would do the job. Something like this:


ANTI-COLLISION:

UniqueID.A + Hash.C -> Wallet.A

UniqueID.B + Hash.C -> Wallet.B


If we associate a particular seed + salt combination to only "open" a UniqueID distributed only once by the blockchain, hash collisions turn impossible since the only way to get access to the wallet would be to brute-force the UniqueID (which would mean to find the Hash.C and would take an astronomically big amount of time).

This would be similar to how you need to know both e-mail adress and password to log into your email, but email adresses are unique and you cant make two identical e-mail adresses in the same server (block-chain in the case of cryptocurrency). Just that the UniqueID should be kind of "secret" like the seed is, and not necessary to do the transactions, so it can be more secure and anonymous. Something like an unrepeatable serial number attached to the "wallet", or the like.

I know it still exists the possibility of somebody inputting by pure "chance" your serial number/UniqueID and your correct seed + salt combination but well... that is something that would even be more improbable than the hash collision, that is already virtually 0.

So, is there any blockchain that does this? I know about Worldcoin decentralized ID, but since the ID is stored in a private key (that´s generated pretty much like any other private key/wallet) collision risk still exists, so technically i could create a private key, open it and find another dude identity (though i repeat i know probability is virtually 0). I searched on Google but didnt find anything except Worldcoin and DIDs.

Maybe is there some technical problem im not realizing?

Thanks to people who read

[markm]

Novacadian told me once that he had a wallet created for himself from scratch, a wallet he expected to be empty, but he found something in it.

If he is right (didn't get confused or something) then I have to wonder whether it is likely to be mere centuries or millenia before the same thing happens to someone else or could it even happen-again even sooner?

I know some bad implementations of some secret-key generators or whatever "random numbers" services they use can cause the range of keys it can generate to be far more limited than intended...


-MarkM-

[Stalker22]

Maybe Im wrong, but I think that this very small possibility of collision in generating Bitcoin private keys exists because the number of possible keys, while astronomically large, is still finite. So, your suggestion to add a unique ID somewhere in the mix would not solve the problem since the private keys are still chosen from a specific range.

By the way, I think you misunderstood how the collision actually happens.  In that unlikely scenario, it doesnt mean that someone just randomly guessed your seed phrase and passphrase, but because of the way the hash algorithm works, there is a theoretical possibility that two completely different seeds generate the same private key.  This is simply a limitation of the hash function.

[Ceres_123]

Maybe Im wrong, but I think that this very small possibility of collision in generating Bitcoin private keys exists because the number of possible keys, while astronomically large, is still finite. So, your suggestion to add a unique ID somewhere in the mix would not solve the problem since the private keys are still chosen from a specific range.

By the way, I think you misunderstood how the collision actually happens.  In that unlikely scenario, it doesnt mean that someone just randomly guessed your seed phrase and passphrase, but because of the way the hash algorithm works, there is a theoretical possibility that two completely different seeds generate the same private key.  This is simply a limitation of the hash function.

hello! First, thanks for your reply

What you are saying is exactly what i wrote seed a + salt a = hash X but seed b + salt b also = hash X. Since seed a + salt a always gives the same hash, there are two options: hash collision (two different seed + salt combinaitons give the same hash) and the option of randomly getting exactly the same seed plus salt as other user (which we can call seed collision or whatever, and that would also be a hash collision)

You are right that the collision is virtually 0 but adding an ID would prevent it from happening even more since you would need to know ID for seed+salt working, like you need to know email adress for email password working

[Stalker22]

~
You are right that the collision is virtually 0 but adding an ID would prevent it from happening even more since you would need to know ID for seed+salt working, like you need to know email adress for email password working

No, you wouldnt need to know the ID for the collision to work.

A cryptographic hash function (like SHA-256) maps a huge input space (your seed) to a fixed-size output (the hash).  A collision means two different seeds produce the exact same hash.  So, theoretically, it would still be possible: ID1 + seed A + salt A = hash X and ID2 + seed B + salt B also = hash X

[Ceres_123]

~
You are right that the collision is virtually 0 but adding an ID would prevent it from happening even more since you would need to know ID for seed+salt working, like you need to know email adress for email password working

No, you wouldnt need to know the ID for the collision to work.

A cryptographic hash function (like SHA-256) maps a huge input space (your seed) to a fixed-size output (the hash).  A collision means two different seeds produce the exact same hash.  So, theoretically, it would still be possible: ID1 + seed A + salt A = hash X and ID2 + seed B + salt B also = hash X

Yes. I don´t have too much technical knowledge but i don´t mean "adding" the ID into the hash to be smashed and hashed, but to correlate the hash X to the ID1 or the ID2 so to open ID1 you need to know hash X. Like even knowing password you cant open email adress if you dont know the specific address you want to open. Or are email adresses also smashed with the password?

According this wikipedia article https://en.wikipedia.org/wiki/Salt_(cryptography) SHA-256 only hashes salt and password, but doesn´t hash the "user", if I am not misunderstanding things 

EDIT: that´s why in "anticollision" part i just wrote UniqueID.A + Hash.C -> Wallet.A and not "=" (like i did in collision) like i wasnt thinking it to be hashed into any value