Topic: [ANN][DASH] Dash (dash.org) | First Self-Funding Self-Governing Crypto Currency - page 6420. (Read 9723858 times)

Who can make the community clean up
We hate to everything we do
We should stick together
Believe me

Of course I want something that is as secure as possible. But there are lots of trade offs that need to be made to ensure that most users needs are covered while keeping usability at it's maximum.

Using a PoW like that was one of my first ideas to protecting against a Sybil attack


The problem with PoW type solutions is the NSA and other powerful entities would have cheap access to large amounts of processing power. Plus, making a user do that hinders the usability of the product.

I like collateral transactions because it accomplishes the same thing and they can be increases to a point where attacking the network becomes way too expensive to do efficiently. Plus, if someone was attacking the network we could ban their collateral inputs by tracing the payments back to the source and isolating them individually.  

DarkSend should embrace the community....Those who wish to attack go straight in for the kill, cold.

DarkSend needs to have some economic disincentives.  Fees for participation, penalties for misbehaving, minimum values for election.

Why not choose random nodes that have shown community engagement? Length of service, etc. Allowed to make small transactions when new, then slowly build-up to larger transactions - Almost like a PoS, or posting on bitcointalk - in terms of reputation establishment?

Anoncoin already has IP obfuscation.

Would you accept that 20% of your coins are not anonymous?

If you are trying to hide from an oppressive totalitarian regime where death or jail time waits you if you are discovered, then you want something 1 in million, not 20%.

20% is analogous to pulling the trigger on a 5 round revolver with one bullet pointed at your head, i.e. Russian Roulette.

Try redoing your calculation with 1%, 0.1%, 0.01%, etc.

In a good, user-friendly solution, all that a user should do is just tick a checkbox and be ok, with the client taking care of everything.

IP obfuscation (TOR or something else, I don't know) will have to be integrated next for the same reason as the one I mentioned earlier: It will be a market weakness. Someone else will take DarkSend, integrate IP obfuscation and claim to solve the IP issue that DRK had in order to take the market.

Btw, enjoying the brainstorming between Evan & Anonymint.

Thats HashCash PoW for email to stop spam. In the Satoshi paper. That was my suggestion.  Hey, you. Stop 51% me with a new fork.

The new algorithm will always attract more people's eyes
Get more attention
Hope that the new algorithm can solve many problems
Give us hope
 

Would you feel Darkcoin is threatened if another altcoin has true cpu-only and very strong anonymity?

Any way I am happy to read below you are thinking about how to improve the anonymity. Your prior reply had me worried that you actually wanted to make it weaker on purpose.  Now I see you are open to improving it.



That is 2012 document. Many think NSA has control over most or many of the nodes on Tor. Remember these servers cost a lot of money and who is providing that for free and getting nothing in return?

Warning FAQ: Tor doesn't protect you from a global adversary:
https://tails.boum.org/doc/about/warning/index.en.html#index7h1

http://en.wikipedia.org/wiki/Tor_%28anonymity_network%29#Exit_node_eavesdropping


Agreed it is. I think I had figured that out before when I mentioned it in the CoinJoin thread and dismissed it. I apparently forgot that since.

I have a new idea for you. You could force each input to be accompanied by an anonymous proof-of-work that costs considerable computing time. Then move the collateral payment to accompany outputs stage.

I think most users wouldn't mind expending 5 minutes computing time before they send a mix transaction.

I have another idea as well. On failure only, every input into the mix could reveal which collateral payment they sent in the output stage, so you can isolate the input that was the adversary. Then you blacklist that input. The inputs anonymity isn't destroyed because no mix transaction was completed. But how can you blacklist system wide? How can you trust that node didn't lie just so it could blacklist someone's coins?

Wow 14 Ghash and difficulty at 705, nice!

I am running a permanent node. pm'd you eduffield

ill run one on lotterymining

Both seed nodes were down, I forget to check them. Anyone else want to host a seed node?

- It needs to be configured to 200 connections max
- Must have 9999 open from the outside
- Must be a static IP
- Must be permanently available  

PM me and I'll add them to the source

Tor is not the solution in my opinion. It is not mainstream and may be too much for folks to learn about proxies and routers for non-browser tor usage.

The community should clean up
Let the roles of the community to help us
Support your proposal
Get some action

Feds finding ways to get through TOR

http://online.wsj.com/news/articles/SB10001424052702303949704579461641349857358?mod=WSJ_TechWSJD_NeedToKnow

Roll-up, roll-up, roll-up....expect new users of DarkCoin (and DarkDogeCoin too.  Much $$, Much anonymous, So moon)

Had the same problem, it took like an hour + to connect to a node but it's all good now.  I just left it alone.

The way I see privacy-with-potential-exploits working in practice is similar to how bitcoin deals with the possibility of 51% attacks. They are known vulnerabilities, and the risk can be to some extent quantified and mitigated. To many, this means the risk is acceptable, and provides enough trust to bet their money on. For others, it's an unacceptable risk.

With darksend in the proposed implementation, the risk of one's anonymity being broken in a given transaction is a function of the ratio of snoops to non-snoops, and the number of stages of mixing employed. The number of snoops is unknown, but there are two points salient here:

1. We can quantify the required funds in DRK to mount an attack of a given efficacy, and
2. With 51% attacks, the hashing power available to a hypothetical attacker (like a large government or financial group) is similarly unknown, and yet this hasn't killed trust in Bitcoin.

An example of 1. could be, "To mount an attack that would break the anonymity of 20% of DS transactions, assuming that there currently exist 1000 uncompromised full nodes capable of being elected a master node, and assuming 3 levels of pooling, we would require approx 58.5% of the network (i.e. cube root of 20%), i.e. 1410 Sybil nodes, each requiring at least 1000 DRK, to a total of 1.41M DRK."

So if our assumptions are reasonable, we can come up with scenarios that reasonably define the scope of risk.

Of course, true anonymity is infinitely better. If Anonymint has a solution for this, it'd be great if he could expand on "divide and conquer", since the phrase in this context is meaningless to me.

Higher end CPUs still mine nearly as well as the GPUs do. I don't think anything says "CPU only" anymore.

Tor + Multiple rounds of DarkSend should be nearly perfect anonymity. Even the NSA says they have problems breaking through Tor. (http://www.theguardian.com/world/interactive/2013/oct/04/tor-stinks-nsa-presentation-document)

I was thinking more about Divide and Conquer, I believe it's vulnerable to Sybil attacks. I don't think you can do decentralized DarkSend without collateral, otherwise what would stop 100 nodes from taking up 80% of the spots per DarkSend session and forcing the divide and conquer algorithm to go 20 or 30 levels before filtering them out (EVERY SESSION)?