Pages:
Author

Topic: same private key? (Read 1717 times)

sr. member
Activity: 1297
Merit: 294
''Vincit qui se vincit''
February 10, 2021, 10:25:13 AM
#95
Learned a lot here.
Thanks to those who answered the question from the OP.
This helps a lot honestly since we should be more aware of a situation like this, though I never heard this kind of problem so maybe this wont exist at all. The developer creates a huge number of combination words to ensure its uniqueness but of course if the wallet comes from suspicious site, then this kind of problem might exist so they can scam the users, its hard to say it.
There are lots of situations where you can gain an experience and knowledge through them. As we can see on his issue, wallet address and private keys or even pnemonic/seed phrase are auto generated and cannot be restored on the database of website wallet address generator unless it the website is used for scamming people.
member
Activity: 1021
Merit: 12
February 10, 2021, 08:40:16 AM
#94
hey,
I wonder, even if the probability so small is, if someone else get the same private key as me could he/she spend my Bitcoins and viceversa? would we have the same Bitcoin Adress?


of course, this is the importance for us to keep private keys from people who can not be trusted because it leaks once our assets can be directly spent or transferred to their personal wallets and none of them can help because of the decentralized nature of bitcoin.
full member
Activity: 1708
Merit: 105
February 05, 2021, 08:02:21 AM
#93
This is very difficult or we can say it impossible. Wallet created in such a way that i cannot be duplicate. We especially me sis not hear from anyone that his private key is match to another account.
If we suppose that its happen than answer is ""Yes""

yes true, theoretically it would not be possible to say that there would be a double private key, because the simulation made was so complicated and for a system error to give an existing private key to the other, it is certain that it will not happen.
the most likely one is the duplication of the private key due to a private error that is not stored properly and is taken by fraudsters in many ways that can be done.

the most frequent one is via email that gives a link where the contents of the scam are, now starting to use telegram which is booming through airdrops which are indeed very easy to manipulate so that we can have a computer entered by the fraudster.
So many ways now live we must look after ourselves properly.
member
Activity: 109
Merit: 10
Student Coin
February 03, 2021, 08:50:26 AM
#92
This is very difficult or we can say it impossible. Wallet created in such a way that i cannot be duplicate. We especially me sis not hear from anyone that his private key is match to another account.
If we suppose that its happen than answer is ""Yes""
sr. member
Activity: 2436
Merit: 267
Hire Bitcointalk Camp. Manager @ r7promotions.com
February 03, 2021, 08:41:55 AM
#91
It is true that the private key cannot be the same, the creators of bitcoin do not play in maintaining security, of course with the expensive price of bitcoin the security key is tightly secured, randomized from the numbers 1 to 9 and from the letters A to Z there is definitely nothing the same.
I really agree with you that everyone will make a key with the best possible security and security that can be guaranteed because it uses a very difficult combination and its security will definitely be tested. Obviously, it becomes a priority that is impossible until it is not well controlled. so it will not be possible to have the same private key occur, other than what we really suspect is that private key was taken because did not maintain it properly.
Therefore, if someone says that there are similarities, it is very difficult to prove that with very good security, it can be the same, unless we are indeed hacked by irresponsible people.
sr. member
Activity: 2296
Merit: 360
February 01, 2021, 05:43:03 PM
#90
When you do talk about private key collision then thinking off on what would be the odds is.. It would be good on reading up this thread.

https://www.reddit.com/r/crypto/comments/drxyjq/chance_of_private_key_collision_for_256bit_ecc/
https://www.reddit.com/r/Bitcoin/comments/1ylis1/collision_probability_for_bitcoin_private_keys/

So i wont really be bothering myself on finding some private key with having some balance.
full member
Activity: 1190
Merit: 123
★Bitvest.io★ Play Plinko or Invest!
February 01, 2021, 05:07:29 PM
#89
The chance of this happening is really slim or should I say nearly impossible, since I came to the crypto space have never heard of two people owning the same private key Shocked, if that was possible it would have happened a long time ago cause every day Moore people keep creating new wallets, so if it was possible one of them would have had the private keys to other persons wallet, but its not possible because private keys are combination of letters and numbers so there's absolutely no way two people can own the same combination something must be different between them.
legendary
Activity: 1554
Merit: 1139
February 01, 2021, 01:05:45 PM
#88
In imagining OP's question, I imagined something else and that is, no two persons has the same microchip number with regards to phone numbers although, country codes plays a part in this. The least I've heard of is some dormant numbers been recycled can't say if it's really true about that. But then, the point I'm driving at is, the microchip numbers is just a combination that ranges within 100000000000 or less and majority of the world have phones now and still not a single person has the same microchip number. Cases of redirect calls seems to be the only explanation for receiving calls for which it wasn't meant for You.

With that being that, how then would two persons have the same private key?! Especially with the fact that, not so many has the idea of private keys or seed phrase, not so many are into or enthusiastic of bitcoin and the private key or seed phrase are combined in a way that, you really can't make out anything if it,it's just words or alphanumeric combinations that follows no definite other and as such, it's limitless as to how many combinations it can generate. So, having the same private key or seed phrase as someone else is close to impossible and it would be a disaster should this ever happen.
sr. member
Activity: 1736
Merit: 357
Peace be with you!
February 01, 2021, 06:43:31 AM
#87
hey,
I wonder, even if the probability so small is, if someone else get the same private key as me could he/she spend my Bitcoins and viceversa? would we have the same Bitcoin Adress?
Yes of course. Though just like what you said the probability is so small like grains of sand. It is all the matter of luck because ever since the creation of crypto I have never been heard of something like the topic had happened. I do hope I am the lucky one who can guess the same private key with those "lost forever" Bitcoins out there.
sr. member
Activity: 1120
Merit: 438
https://bitcointalk.org/index.php?topic=5274318.0
September 06, 2020, 03:13:48 AM
#86
I thought this is a simple question until I found this is interesting after reading the comments.
so, a private key / wallet address is not purely generated uniquely.

it just close to unique because of the chance to have the same private key / wallet address close to impossible, because the amount of different keypairs that can be generated is huge.
CMIIW.

I like the coin flipping analogy BlackHatCoiner though.
legendary
Activity: 3668
Merit: 6382
Looking for campaign manager? Contact icopress!
September 06, 2020, 01:49:00 AM
#85
The chance is about the same as hand-picking the only one correct grain of sand from a beach.
An internet search puts the number of grains of sand in the world at around 1020, give or take a few orders of magnitude. 2256 is equivalent to ~1077. So the chance is similar to hand picking the only correct grain of sand in the entire world chosen at random, four times consecutively.

You are most probably correct.
I was expecting I may miss some "orders of magnitude", but I thought that should be not a big problem since the goal was easier understanding, not being 100% correct.
However, the idea of adding the "four times consecutively" is very good and although I don't expect too many who didn't understand the initial explanation gasp the difference, I'll add that. Thank you.
full member
Activity: 616
Merit: 102
September 05, 2020, 08:25:58 AM
#84
For bitcoin private keys it is clearly impossible to be the same, because the private key consists of a unique combination of numbers and letters, try to imagine for yourself if an investor who has a lot of bitcoin assets will definitely feel annoyed if he finds out that his private key is the same as someone else's.
legendary
Activity: 2268
Merit: 18771
September 05, 2020, 03:45:00 AM
#83
We don't need to create an algorithm for reversing hash functions, actually.
That's true, until we start talking about quantum computers, which is the most likely way that bitcoin will be "broken" in the near future. A quantum computer in a few decades could potentially use Shor's algorithm to reverse elliptic curve multiplication and calculate a private key from a public key. It could not, however, reverse a hash. Since the public key is only revealed when a transaction is made, as long as you do not reuse addresses, then your coins remain safe.

The chance is about the same as hand-picking the only one correct grain of sand from a beach.
An internet search puts the number of grains of sand in the world at around 1020, give or take a few orders of magnitude. 2256 is equivalent to ~1077. So the chance is similar to hand picking the only correct grain of sand in the entire world chosen at random, four times consecutively.
legendary
Activity: 3472
Merit: 10611
September 05, 2020, 02:20:49 AM
#82
The amount of data, size, structure are known in this case. For a compressed public key, it is 33 bytes hex-number as input for sha-256, and standart SHA-256 hash as input in RIPEMD160 function. Does it make things easier to crack? I guess not, but at least we have something to consider.

ireversibility of the hashes is not about the data size but about the algorithm that they use. to put it simply we can use a simple mathematical formula like x+y. if we know the result of this formula to be something like 12 you have no way of reversing it back to x and y. that is why it is impossible to reverse a hash function.

that formula above is not far from reality either. your RIPEMD160 result is literary the addition of 2 32-bit numbers for each of its hash state items. hash_state[1] = hash_state[2] + d + e where d and e,etc. are computed by bit rotations.
full member
Activity: 2268
Merit: 121
Enterapp Pre-Sale Live - bit.ly/3UrMCWI
September 05, 2020, 01:44:37 AM
#81
the possibility of using the same private key can happen but whether it can be used maybe this needs to be questioned again. if it can't be used, then finally the problem has been resolved besides it can indeed be used.
legendary
Activity: 3668
Merit: 6382
Looking for campaign manager? Contact icopress!
September 04, 2020, 09:20:49 AM
#80
If it can happen, it will happen, this is how I see it.
~
The input in to SHA256 can be any amount of data up to around 2 exabytes (2 million terabytes)
~

I'll attempt an easier comparison (maybe not as accurate as yours, but still something more real that just numbers).

The chance is about the same as hand-picking the only one correct grain of sand from a beach, 4 times in a row (randomly different grains of sand for each 4 picks)
You can't do it with a machine (the numbers we talk about are so big the machine would not have enough power to only count those), so you'll have to hand pick.

Now think again about the "it will happen" part.


Edit: I improved (corrected) the math based on @o_e_l_e_o's post
sr. member
Activity: 1274
Merit: 259
September 04, 2020, 09:09:42 AM
#79
I am of the opinion that if the private key is the same as yours, it could possibly take your coins, it could happen,
but it is very unlikely to happen, in my opinion the private key is made with very good security standards, so getting the same private key is very difficult to happen.
legendary
Activity: 2464
Merit: 4415
🔐BitcoinMessage.Tools🔑
September 04, 2020, 07:22:47 AM
#78
My point about hashes is quite different though. Reversing a hash is quite literally impossible. Take SHA256 as an example. The input in to SHA256 can be any amount of data up to around 2 exabytes (2 million terabytes), but the output is always a string of 64 hexadecimal characters. Therefore, data from the input is lost during the process which makes it impossible to reverse. Although finding collisions is possible (but difficult), you can't definitely tell me what my input is. Just as if I told you I have two numbers which add up to 1000, you could give me two numbers which fit the criteria but have no way of knowing if they are the same two numbers which I am using.
The amount of data, size, structure are known in this case. For a compressed public key, it is 33 bytes hex-number as input for sha-256, and standart SHA-256 hash as input in RIPEMD160 function. Does it make things easier to crack? I guess not, but at least we have something to consider.

We don't need to create an algorithm for reversing hash functions, actually. Even if it were possible, it wouldn't give us much information anyway. After all, our goal is to find private key not public key, public key calculation and further hashing into bitcoin address is just for verification (whether or not the private key we found can be used to sign a transaction). What we do is generate a private key, calculate corresponding public key and then applying (RIPEMD160(SHA256(public key))) function and base58 encoding. If our value is matching with sought value, we are fine, if not, we continue until death of universe. Personally, if had a huge amount of proccesing power, I would spend it on mining of new bitcoins, it would be more profitable and more rational than finding the same private keys.
legendary
Activity: 2268
Merit: 18771
September 04, 2020, 04:51:54 AM
#77
If it can happen, it will happen, this is how I see it.
Without getting too much off topic here, this logic is incorrect.

Given that entropy is always increasing towards the eventual heat death of the universe, then there is finite amount of time in which things can happen. Therefore, just because something can happen does not guarantee that it will happen. Even if we assume that time is infinite, this also does not guarantee that everything which can happen will happen, as infinite does not necessarily mean everything. There are infinite numbers between 1 and 2, but none of them are 3. The line y = x will continue infinitely, but will only ever visit the value of 1 once.

My point about hashes is quite different though. Reversing a hash is quite literally impossible. Take SHA256 as an example. The input in to SHA256 can be any amount of data up to around 2 exabytes (2 million terabytes), but the output is always a string of 64 hexadecimal characters. Therefore, data from the input is lost during the process which makes it impossible to reverse. Although finding collisions is possible (but difficult), you can't definitely tell me what my input is. Just as if I told you I have two numbers which add up to 1000, you could give me two numbers which fit the criteria but have no way of knowing if they are the same two numbers which I am using.
legendary
Activity: 2156
Merit: 1151
Nil Satis Nisi Optimum
September 04, 2020, 03:06:00 AM
#76
Knowing an address provides exactly zero help when it comes to bruteforcing a public key, and knowing a public key provides exactly zero help when it comes to bruteforcing a private key. An attacker would still have to exhaust half the search space, on average.
Thank you for the corrections, but I personally prefer avoid strong adjectives when it comes to questions about odds, sometimes I fail and use "infeasible" insted of "extremelly unlikely" and "almost impossible". If it can happen, it will happen, this is how I see it. It may be that knowing bitcoin address provides zero help, but at least it gives some information to an attacker: existence of UTXO and amount of bitcoin. Address is now active and therefore more interesting than empty, inactive ones. Still, with the current computation power, it is impossible to crack specific address. But what if an attacker is not interested in specific address, but in all addresses with balance. I have heard of so-called "bitcoin collider" pools where they create a list of known bitcoin addresses with funds and then searching for collisions. What they do is simply a calculating of bitcoin addresses: private key -> public key -> address and comparing them with the list of known addresses. It looks like stupid waste of energy, but sooner or later they may find something.

did not know that this is even possible, but of course is it, and in the end after enough time, one could find some collision, it is not likely event, but i agree, if it is possible, it will eventually happen
just probability that is stuck to zero, it is impossible, when it is not zero, than some number, no matter how small, it leads to something happen after enough time
Pages:
Jump to: