Even though I don't believe I have anything to worry about, I'd still prefer to add more signers to the multi-sig so I don't have full control anymore. This would make it safer for everyone, I really do not like the fact that users have to trust me. For now this is the only option though, and I will not take any steps in this direction unless I am 100% sure it's done in a safe way. The community would also have to agree with the plan before I set it in motion
The way you are trying to involve individuals from the community and keep talking about community in the main operations with the multi signature addresses and things, I wonder what the three letter agency will feel about it when they will target your project. If you become bigger then today or tomorrow they will come after you and the people with you working in the managerial level holding the keys.
If they get the false sense of understanding that the mixer is running by bitcointalk community then immediately they will come after bitcointalk and destroy it.
The goal of this thread is to have a discussion regarding this issue. If at any point we come to the conclusion that it's riskier to run Whirlwind as a community project then I will simply continue to do it myself. I don't understand your point about Bitcointalk as a forum getting dragged into this since it has nothing to do with Whirlwind
If we didn't have the Notes then I agree, Fast mode would have the same disadvantages as any other mixer. But since outputs from Notes and Fast look exactly the same there is no way for any outside observer to know which mode you used. So Fast mode is as secure as the Notes from a privacy set standpoint. I would still recommend using Notes regardless because they offer the end-user full control over the process.
That's a good point, and one I did not consider. For an external observer using blockchain analysis, then a fast mix appears identical to someone using notes. They can see the deposit being made, but since they don't know if the user is using fast or notes, they are unable to reach any conclusions about the time frame of when the withdrawal will be made or how much will be withdrawn. Both fast and notes users benefit from being in the larger anonymity set provided by the other type of user, and having the different process help to obfuscate what is happening.
Correct
So, as far as I've understood (without giving much emphasis on the details), whirlwind is a mixer that knows the input, but doesn't know the output (i.e., I send 0.01 BTC, but they don't know which 0.01 BTC output I will spend). Is that correct?
I have some questions:
- First of all, what's your setup, as NotATether said? ChipMixer was proved to have poor setup, and even if your service isn't prone to failure due to centralization, your absence would lead to the corruption of the service (at least now that it's brand new).
- How do you plan to select anonymous trustworthy members?
- Who grants us that the authorities will not try to shut down the federation? AFAIK, from what I've read, the trustworthy members will only protect the users in case whirlwind is shutdown, and it protects their privacy using blinded certificates, but it doesn't grant that the service will continue being online after whirlwind (the user) disappears.
Very interesting implementation, I hope it goes well.
-Can you be more specific about this question? What do you mean by what's our setup
-No idea at the moment, we would all have to agree on a "procedure"
-If we manage to implement the multi-sig with multiple trusted members, then even if I go missing 1 hour after that it does not matter. The remaining members can run the service as if it never happened, so the service can continue with or without me
Having said all of the above as far as I'm concerned I am not doing anything illegal. I don't encourage illegal activity and will never promote the service on the darknet or for any illegal purposes, I'm a simple provider of privacy services. There are no statistics regarding % of CEX funds coming from illicit sources so we can't compare to what we know about mixers, but my guess is that the number is very similar if not higher for centralized exchanges. There are bad actors in every industry, you can't just shut down all businesses of one type because of a few bad apples. If the service will start to get seriously abused by bad actors and big pressure will be put on us then I'd much rather shut down the service early and honorably than put users funds and privacy at risk, but for now I still believe there has to be a way to run everything legally.
Thank you for your responses, hope you don't mind if I ask you some deep questions:
1. Do you do something or plan to do something to prevent abuse of your service? I mean to minimize it cause nothing is totally preventable. There are people who care about their privacy and there are people who want to do illegal things, do you have a plan to make your service unlikeable for the people who do illegal things? To get rid of them. Do you think are there any measurements that you can take while keep your service functional for people who care about their privacy?
I know this question can sound strange but it's still an interesting one. More likely I mean, you may be able to get list of addresses that are known to be found in illegal activities and you may include these addresses in your blacklist to not be able to use your service.
2. I think, you understand that doesn't matter how trusted someone is on this forum, there is a chance that any signer can actually be a spy. By the way, what do you think, what's the number of signers that can make you feel safe and get rid of cooperation to steal money? Definitely 3/3 or 7/7 won't work, you need something like 2/3 or 5/7 at least. I think this is a huge challenge.
1.I'm still trying to figure out if there is a way to do this, if I have any ideas I will write about them before I implement anything. Here is an idea I had, but we need the Blind Certificates in order for it to work. It would be possible to prove your funds are not coming from specific addresses without revealing which one your deposit actually is.
With the Blind Certificates I talked about in my previous posts it may be possible for users to prove their funds don't come from specific addresses linked to hacks/ransom/etc., so if that is possible then honest users have a way to prove they are not thieves while retaining privacy, and bad actors are isolated so sending the BTC to whirlwind is pretty much useless if they plan to use centralized services afterwards since they couldn't prove they are not one of the bad actors.
2.Your first point is my biggest concern and something for which I'm not convinced that a solution even exists. We will come to a conclusion together after more discussions, whatever that may be