Author

Topic: NXT :: descendant of Bitcoin - Updated Information - page 1255. (Read 2761626 times)

legendary
Activity: 1680
Merit: 1001
CEO Bitpanda.com
I am no programmer, but i have a proposal to solve the mixing problem.

How about a pooled approach that is not traceable?

We have at specific times per day (or maybe once a day if traffic is low) a mixing event where everyone can send their coins to an automated mixing address that collects all coins. Once the specific time comes, the logic of the address creates packs of 10000 NXTs and sends them to specific addresses. For example, if you send 100.000 NXT to that address, your coins will be split into 10k pieces on 10 addresses. Those addresses are generated by the mixing algo and the private code is sent to original sender via encrypted address that is only readable if the private key of the original address is known.

Rules:
You can send whatever amount between 1000 and higher
The splitting follows those rules, Balance/10000 --> remainder/1000 --> remainder /100 --> rest is donation(? or maybe remainder/10?)
E.g. If you send 17.200 then you get 1 10k, 7 1k and 2 100 outputs.
The interval of the mixing is depending on the traffic. Only if enough people participate, the mixing will be anonymous.
The passwords are send per AM and are encrypted per the private key of the original account. (i guess that is possible)

What do you think?

This is exactly what i had thought, it's simple and it should work?
But my thought was instead of having a timer for returning funds, i was thinking of reaching certain pooling amount before returning which sometimes will cause a delay when not enough people is using it, but if your desperate for anonymity i think you will have the patient too.


My idea was to limit Capacity of Mixer Pool to 1,000 or 10,000 addresses depending on level of anonymity, all Mixer Amounts only accepts exact 1000nxt+1nxt fee, therefor in the Pool there will be 1,000 addresses of 1,000nxt and a random address will be chosen only after the pool reaches past it's capacity (1,000), and returns your 1,000nxt -1 nxt fee.

Works something like this;
OriginalAddressIn(A)-->MixerAddresses From Pool(A-Z)--->NewAddressOut(A-Z)
Would this NOT work???


Well atleast this approach can be implemented off-chain. Do you guys think there is enough intrest in this to give it a shot?
Without cryptographic proof, source code reviewed, unlikely to be used to any large degree
This is a totally nontrivial problem
Doubtful any ad hoc method has actual annonymyzing power

Yes i know.
My implemention would be a trivial "solution" at best.

Here is how i would go about it:

1. You make an account on the site.
2. Then you specify where your next deposit will come from.
3. You can deposit whatever amount you want.
4. You can request cashouts to whatever adress you want. But the cashout amount will only ever be round 1.000 NXT (or 100 NXT or 10.000NXT depending on the traffic)

How would one be able to connect the deposit with the withdrawal addresses?
member
Activity: 63
Merit: 10
So unfortunately I have to sell off my Nxt.
Is dgex the way to go?
 Also, when depositing to dgex, you are asked for your "ACCOUNT TOKEN", is that just my account number?
newbie
Activity: 56
Merit: 0
We just need the 288 byte zeroknowledge proof mechanism

Yep. Searching various related articles too.

How will the ZeroCoin project benefit from being implemented on top of Nxt? Let's build a compelling case for them to join forces.
Biggest benefit i see for them is that they can skip the long difficult process of getting a secure blockchain with hundreds of servers
Also when TF is doing many TPS, so will zerocash


Great! This is HUGE incentive for ZeroCoin to work with Nxt!
member
Activity: 98
Merit: 10
but how can my transactions help me with forging?

They don't. Anon was talking about other people's (in particular, the other X mil guy's) transactions.
legendary
Activity: 1176
Merit: 1134
I am no programmer, but i have a proposal to solve the mixing problem.

How about a pooled approach that is not traceable?

We have at specific times per day (or maybe once a day if traffic is low) a mixing event where everyone can send their coins to an automated mixing address that collects all coins. Once the specific time comes, the logic of the address creates packs of 10000 NXTs and sends them to specific addresses. For example, if you send 100.000 NXT to that address, your coins will be split into 10k pieces on 10 addresses. Those addresses are generated by the mixing algo and the private code is sent to original sender via encrypted address that is only readable if the private key of the original address is known.

Rules:
You can send whatever amount between 1000 and higher
The splitting follows those rules, Balance/10000 --> remainder/1000 --> remainder /100 --> rest is donation(? or maybe remainder/10?)
E.g. If you send 17.200 then you get 1 10k, 7 1k and 2 100 outputs.
The interval of the mixing is depending on the traffic. Only if enough people participate, the mixing will be anonymous.
The passwords are send per AM and are encrypted per the private key of the original account. (i guess that is possible)

What do you think?

This is exactly what i had thought, it's simple and it should work?
But my thought was instead of having a timer for returning funds, i was thinking of reaching certain pooling amount before returning which sometimes will cause a delay when not enough people is using it, but if your desperate for anonymity i think you will have the patient too.


My idea was to limit Capacity of Mixer Pool to 1,000 or 10,000 addresses depending on level of anonymity, all Mixer Amounts only accepts exact 1000nxt+1nxt fee, therefor in the Pool there will be 1,000 addresses of 1,000nxt and a random address will be chosen only after the pool reaches past it's capacity (1,000), and returns your 1,000nxt -1 nxt fee.

Works something like this;
OriginalAddressIn(A)-->MixerAddresses From Pool(A-Z)--->NewAddressOut(A-Z)
Would this NOT work???


Well atleast this approach can be implemented off-chain. Do you guys think there is enough intrest in this to give it a shot?
Without cryptographic proof, source code reviewed, unlikely to be used to any large degree
This is a totally nontrivial problem
Doubtful any ad hoc method has actual annonymyzing power
legendary
Activity: 1680
Merit: 1001
CEO Bitpanda.com
Are you talking about price going to 0.00001 or 0.0001?

0.00001 seems mission impossible to me.

Impossible? In what time frame I see that easily a target for NXT.

but he said: "despite the dumping" so he means the price will appreciate.
legendary
Activity: 1120
Merit: 1000
Are you talking about price going to 0.00001 or 0.0001?

0.00001 seems mission impossible to me.

Impossible? In what time frame I see that easily a target for NXT.

Are you talking about 0.00001 or 0.0001?
hero member
Activity: 602
Merit: 500
Are you talking about price going to 0.00001 or 0.0001?

0.00001 seems mission impossible to me.

Impossible? In what time frame I see that easily a target for NXT.
legendary
Activity: 1680
Merit: 1001
CEO Bitpanda.com
I am no programmer, but i have a proposal to solve the mixing problem.

How about a pooled approach that is not traceable?

We have at specific times per day (or maybe once a day if traffic is low) a mixing event where everyone can send their coins to an automated mixing address that collects all coins. Once the specific time comes, the logic of the address creates packs of 10000 NXTs and sends them to specific addresses. For example, if you send 100.000 NXT to that address, your coins will be split into 10k pieces on 10 addresses. Those addresses are generated by the mixing algo and the private code is sent to original sender via encrypted address that is only readable if the private key of the original address is known.

Rules:
You can send whatever amount between 1000 and higher
The splitting follows those rules, Balance/10000 --> remainder/1000 --> remainder /100 --> rest is donation(? or maybe remainder/10?)
E.g. If you send 17.200 then you get 1 10k, 7 1k and 2 100 outputs.
The interval of the mixing is depending on the traffic. Only if enough people participate, the mixing will be anonymous.
The passwords are send per AM and are encrypted per the private key of the original account. (i guess that is possible)

What do you think?

This is exactly what i had thought, it's simple and it should work?
But my thought was instead of having a timer for returning funds, i was thinking of reaching certain pooling amount before returning which sometimes will cause a delay when not enough people is using it, but if your desperate for anonymity i think you will have the patient too.


My idea was to limit Capacity of Mixer Pool to 1,000 or 10,000 addresses depending on level of anonymity, all Mixer Amounts only accepts exact 1000nxt+1nxt fee, therefor in the Pool there will be 1,000 addresses of 1,000nxt and a random address will be chosen only after the pool reaches past it's capacity (1,000), and returns your 1,000nxt -1 nxt fee.

Works something like this;
OriginalAddressIn(A)-->MixerAddresses From Pool(A-Z)--->NewAddressOut(A-Z)
Would this NOT work???


Well atleast this approach can be implemented off-chain. Do you guys think there is enough intrest in this to give it a shot?
member
Activity: 98
Merit: 10
It's a good trick if applied in a right place. In Nxt we don't need it if a secret phrase entropy is high enough (~256 bits).

Yeah, but that's the problem! People constantly pick low-entropy passphrases, because they are used to regular passwords.

And then their account gets hacked with something that looks like offline bruteforcing.

So what if instead of doing

MessageDigest.getInstance("SHA-256").digest(secretPhrase.getBytes("UTF-8"))

we add something like PBKDF2 to improve security?

If we did something like this: https://bitcointalksearch.org/topic/m.4821453 then we could easily add Bitcoin-like wallet functionality.  Let the computer use a good random number generator to generate the private key.  Local machine could have its own "weak" password for the local wallet, and the blockchain address would not be able to be brute-forced.

I like nexern's proposal, https://bitcointalksearch.org/topic/m.4774527
with the modifications I suggested: https://bitcointalksearch.org/topic/m.4776220

I would do:

SUFFIX = Hash^n(PASSWORD) XOR SEED , where

SEED is a high-entropy random bitstring of equal length to SUFFIX
Hash() is a commonly implemented strong hash function that returns a bitstring of length equal to SUFFIX
Hash^n() is n iterations of Hash(), and
n is chosen as large as possible such that Hash^n() takes say <0.5 secs to run. This is done when SUFFIX is first generated.

Then n and SEED are stored (unencrypted) to file. SUFFIX is generated from file whenever user logs in with PASSWORD.

Even if the file is compromised, rainbow attacks on weaker passwords fail because different users will have different values of n; an attacker cannot feasibly prepare for all possible n.

(Extremely weak passwords will still get broken by brute force. The n iterations delay brute-forcing for weak, but not too weak pws)

Note: I'm not a crypto pro, I'm coming from just general math. Please review this idea first if you want to use it.
legendary
Activity: 1512
Merit: 1124
Invest in your knowledge
Logic question:

What would happen if the Genesis Account started forging?

Genesis is hungry for nxt
legendary
Activity: 1680
Merit: 1001
CEO Bitpanda.com
Logic question:

What would happen if the Genesis Account started forging?
legendary
Activity: 1176
Merit: 1134
We just need the 288 byte zeroknowledge proof mechanism

Yep. Searching various related articles too.

How will the ZeroCoin project benefit from being implemented on top of Nxt? Let's build a compelling case for them to join forces.
Biggest benefit i see for them is that they can skip the long difficult process of getting a secure blockchain with hundreds of servers
Also when TF is doing many TPS, so will zerocash
legendary
Activity: 1680
Merit: 1001
CEO Bitpanda.com
In tears laughing my ass off at emule, I have a 300k sell order racked at .00001, buying a new nissan 370z  Cool. Keep on selling, the lucky guy that bought your 1 million will buy an island one day because of your stupidity.

ok im lost here...  not that Im all aware on trading in the first place though..  but if you have a sell order for 300K NXT set for .00001 isnt that bad for you for the price to fall?

Well, in the short term yes. However I think it will pass .00001 very shortly, despite Emu's "dumping."

now we get it, you confuse 0.0001 with 0.00001. Dyslexia? hehe
legendary
Activity: 1120
Merit: 1000
Are you talking about price going to 0.00001 or 0.0001?

0.00001 seems mission impossible to me.
hero member
Activity: 490
Merit: 504
but how can my transactions help me with forging?
member
Activity: 98
Merit: 10
I look forward to rich get richer because i think it may be a legit criticism. I would love for you [Salsacz] to make a strong argument to the contrary. here is my line of thinking.

If you think through the forging process logically suppose you have 2 actors forging, one with 2million nxt and one with 1million nxt. The net balance of payments will tend to flow from the account with 2million nxt into the account with 1 million nxt only if the account with 2million nxt engages in greater than twice as many transactions as the account with 1 million nxt. This indicates that its not necessarily about how much nxt you have, how many transactions you make plays a big part. ...

That assumes transaction fees remain at 1 nxt. If it were a non-negligible % of amount sent instead, assuming the 2 mil guy spends ~twice as much as the 1 mil guy, the gap will gradually close.

I'm seeing a lot of arguments about voting from different people assume this fixed tx fee too. Keep in mind that this likely won't remain the case.
newbie
Activity: 56
Merit: 0
In tears laughing my ass off at emule, I have a 300k sell order racked at .00001, buying a new nissan 370z  Cool. Keep on selling, the lucky guy that bought your 1 million will buy an island one day because of your stupidity.

ok im lost here...  not that Im all aware on trading in the first place though..  but if you have a sell order for 300K NXT set for .00001 isnt that bad for you for the price to fall?

Well, in the short term yes. However I think it will pass .00001 very shortly, despite Emu's "dumping."
legendary
Activity: 1680
Merit: 1001
CEO Bitpanda.com
Why does EMunie becoming Emunie ltd  mean they can't run an exchanges within their software?

And what is AEsystem, seen a couple references to it but don't have a clue what it is?

AE = asset exchange = exchange

Because theyy have to comply with UK law. Their law has strict rules on assets, also you can't legally sell drugs and weapons in the UK, which noone can restrict you doing on a decentralized asset exchange...
hero member
Activity: 527
Merit: 503
Why does EMunie becoming Emunie ltd  mean they can't run an exchanges within their software?

And what is AEsystem, seen a couple references to it but don't have a clue what it is?
Jump to: