XMR: thoughts on I2P integration into CryptoNote

Discussion of code updates and pull requests proposed by the community across all CryptoNote currencies.

XMR: thoughts on I2P integration into CryptoNote

Postby Lost my passwords » Mon Jun 09, 2014 4:13 am

Currently there's work being done on adding an I2P c++ router into CryptoNote technology. I am wondering if there are thoughts on this in general, and would welcome any type of feedback that anyone has.

I'll open: What would be the difference between using a newly developed c++ router versus the current implementation of I2P in Anoncoin?
Lost my passwords
 
Posts: 14
Joined: Mon Jun 09, 2014 4:08 am

Re: Thoughts on I2P integration into CryptoNote

Postby Lost my passwords » Tue Jun 10, 2014 11:00 am

It's cool I'm all about talking to myself. Aminorex at bitcointalk already made an awesome point.

Apparently the router implementation will be working within the Cryptonote protocol, rather than as a separate program like the current Java application. So it will be taking what would have been two programs and making them one. This yields an overall more simple usage and hopefully a much more secure one.

The security mainly comes from less local network traffic, which can be picked up by malware.

All good things to consider when aiming for privacy/anonymity.

Has an I2P style integration been considered before in CryptoNote? Or was there some technical flaw with the concept that it was considered unnecessary?
Lost my passwords
 
Posts: 14
Joined: Mon Jun 09, 2014 4:08 am

Re: Thoughts on I2P integration into CryptoNote

Postby Hawking_Br » Tue Jun 10, 2014 5:41 pm

I2P is just a network transfer technology. It doesn’t depend on CryptoNote, Bytecoin, Bitcoin or any other coin it can be implemented into. This implies that while CryptoNote is a virtual currency rights transfer protocol, i2p is the data transfer protocol.

The i2p idea is beneficial from one perspective; network nodes are hidden from each other. It indeed gives more anonymity to the whole network but it doesn't seem to add anything to the transaction anonymity. From this perspective i2p brings nothing new to the ring signature scheme that is in charge of CryptoNote's untraceability.

We tried to make CN as cross-platform available as possible. That is why it was divided into several modules. Currency layer is responsible for the transaction anonymity, while p2p layer is responsible for the network connection. It’s possible to build various projects based on our p2p layer, as it is designed to be independent from other layers.

Apart from hiding network’s nodes i2p may cause certain extra risks. If there is a modified daemon attack (DDoS, flood, etc.) it would be difficult to identify the abuser. Some network nodes may even be isolated from the rest of the network by a malefactor, which would be difficult to identify.

Even though we welcome i2p integration as an important initiative, we can’t see it as a necessary addition from the core protocol point of view.
Hawking_Br
 
Posts: 1
Joined: Wed Mar 26, 2014 3:39 pm

Re: Thoughts on I2P integration into CryptoNote

Postby Lost my passwords » Wed Jun 11, 2014 1:54 am

Hawking_Br wrote:I2P is just a network transfer technology. It doesn’t depend on CryptoNote, Bytecoin, Bitcoin or any other coin it can be implemented into. This implies that while CryptoNote is a virtual currency rights transfer protocol, i2p is the data transfer protocol.
...
Currency layer is responsible for the transaction anonymity, while p2p layer is responsible for the network connection. It’s possible to build various projects based on our p2p layer, as it is designed to be independent from other layers.


Maybe my discussion on that would be better off in the philosophy section .. but I'll try to break this down. If the p2p layer wasn't developed for currencies, would a refinement like i2p becoming part of it be a possible better definition of a p2p layer targeted specifically for currency usage? Making a reach here .. my pennies don't have records of their previous logical locations in a network of cash in my town. Why would they need that, when their innate value is simply $.01? People know I use pennies when that's what I use to pay for them .. so why would they need to know where that penny has been or most likely has been? Actually I just realized that may be off-topic because I've seen someone mention that you guys are not currency developers.

Apart from that, the way you're presenting this is that a currency layer is just one amongst many that may be conceived to work with your p2p layer. What other layers have you considered? Anything like bitmessage? Sorry, again kinda off topic.

Hawking_Br wrote:We tried to make CN as cross-platform available as possible. That is why it was divided into several modules.

What logical layers do you see in the software? Just curious .. like when you think of different parts of the software how what do you classify them as? I think we see this as two different things.

Hawking_Br wrote:The i2p idea is beneficial from one perspective; network nodes are hidden from each other. It indeed gives more anonymity to the whole network but it doesn't seem to add anything to the transaction anonymity. From this perspective i2p brings nothing new to the ring signature scheme that is in charge of CryptoNote's untraceability.

I was under the opinion that the transaction anonymity (I've generally referred to this as just "privacy", where the ability to obscure the fact from a third party that someone performed a transaction was anonymity -- what are your definitions?) was pretty top notch. I've seen people ask why it's not broken down into powers of two .. and instead by the digit to a power of ten .. would something like that be worthwhile to pursue?

Hawking_Br wrote:Apart from hiding network’s nodes i2p may cause certain extra risks. If there is a modified daemon attack (DDoS, flood, etc.) it would be difficult to identify the abuser. Some network nodes may even be isolated from the rest of the network by a malefactor, which would be difficult to identify.

Wouldn't increased latency in an i2p network itself far outpace the ability of an attacker to DDoS or flood a node? Though it would indeed make it difficult to easily identify an attacker .. wouldn't it require significantly more resources in order to mount the attack in the first place?

Hawking_Br wrote:Even though we welcome i2p integration as an important initiative, we can’t see it as a necessary addition from the core protocol point of view.

It will be a significant marketing challenge to market not only one, but two technologies whose time has come and are ready for prime time. My hope is to help as many people understand, or feel like they understand, as possible. That will require definitions that are easy to digest.
Lost my passwords
 
Posts: 14
Joined: Mon Jun 09, 2014 4:08 am

Re: Thoughts on I2P integration into CryptoNote

Postby Pliskov » Mon Jun 16, 2014 4:41 pm

Lost my passwords wrote:
I've generally referred to this as just "privacy", where the ability to obscure the fact from a third party that someone performed a transaction was anonymity -- what are your definitions?


The way we define anonymity in this particular case is an inability by a third party to analyze the blockchain to get the details of the users. In other words, you noone is able to tell who is engaged in the transaction and what sum was invovled simply by tracing it in the ledger. The way the anonymity protection works in the p2p network is not covered in this paper. The way the nodes are supposed to interact with each other to hide the data flow is yet to be analyzed.


Lost my passwords wrote:Wouldn't increased latency in an i2p network itself far outpace the ability of an attacker to DDoS or flood a node? Though it would indeed make it difficult to easily identify an attacker .. wouldn't it require significantly more resources in order to mount the attack in the first place?


It really depends on the types of the attacks. DDoS has nothing to do with the channels’ carrying capacity. All the nodes in a p2p network have more or less similar characteristics. The way an attack becomes possible is through an overwhelming volume of traffic directed at a particular node.
Pliskov
 
Posts: 16
Joined: Wed Mar 26, 2014 3:37 pm

Re: XMR: thoughts on I2P integration into CryptoNote

Postby ReC » Thu Nov 13, 2014 10:31 am

Hi fellows

I was thinking about p2p layer, cpu mining, botnets, etc.

Is it possible then to limit the amount of mining work by node address in order to keep the egalitarian proof of work ? what it would means to relay and network scalability ?
reference:
"https://wiki.bytecoin.org/wiki/Protocol_rules"

TX GET

check if tx is new
check if tx is valid for relay; reject if failed
here additional checks can be performed in order to protect network from DDOS. These verifications are to be defined by local node's policy.
User avatar
ReC
 
Posts: 15
Joined: Mon Sep 22, 2014 9:43 pm

Re: XMR: thoughts on I2P integration into CryptoNote

Postby Werner_Albert » Tue Nov 18, 2014 3:25 pm

Is it possible then to limit the amount of mining work by node address in order to keep the egalitarian proof of work ? what it would means to relay and network scalability ?


It is not possible to limit the amount of mining work by node address. The bigger nodes have an intrinsic capability of distributing its hash rate through multiple IPs. Due to anonymous nature of CN addresses, the transactions that are executed by the miners can not be bound together.
Werner_Albert
 
Posts: 56
Joined: Wed Mar 26, 2014 3:23 pm

Re: XMR: thoughts on I2P integration into CryptoNote

Postby ReC » Fri Nov 21, 2014 3:06 pm

Werner_Albert wrote:
Is it possible then to limit the amount of mining work by node address in order to keep the egalitarian proof of work ? what it would means to relay and network scalability ?


It is not possible to limit the amount of mining work by node address. The bigger nodes have an intrinsic capability of distributing its hash rate through multiple IPs. Due to anonymous nature of CN addresses, the transactions that are executed by the miners can not be bound together.


I think I understand in part, but I still not fully understand the entire processes though, may because I have networking bias and think too much about pipes (network cards) than I should. I will keep my research on point "Time-Lock Puzzles in the Random Oracle Model" (Needle in a haystack problem).

Thanks,
I'm learning a lot since I get involved with cryptonote. ;)
User avatar
ReC
 
Posts: 15
Joined: Mon Sep 22, 2014 9:43 pm

Re: XMR: thoughts on I2P integration into CryptoNote

Postby lucygp » Wed Jan 21, 2015 1:59 am

The way the anonymity protection works in the p2p network is not covered in this paper.









_______________________________
Fut coins
lucygp
 
Posts: 3
Joined: Wed Jan 21, 2015 1:55 am


Return to Features & code updates discussion

Who is online

Users browsing this forum: No registered users and 0 guests