The anonymity features are new. Nobody has really reviewed them. There are almost no documented attacks on them. But most of them do contain implementation bugs or are plainly stupid ideas, sorry. So you can't and shouldn't trust any feature. Because in 99% of cases you won't get any additional privacy. In fact you may even lose some because of poor implementation.
So I think that the development of any anonymity or security feature must be done in transparent and open way. Otherwise you really can't trust any of the solutions.
I couldn't agree more with you on this. If you checked my posts, I stated that I'm against using new untested methods for production use. Thank you for this
How do we know that mindfox is not NSA or FBI employee and that he won't plant backdoor in his solution? But he doesn't have to be NSA or FBI employee to plant backdoor. Most backdoors in security software are not intentional and are simply consequences of incompetence and/or stupidity (sorry for expression) of software engineers or are consequences of new types of attacks (for example, timing attacks in cryptography).
Again we are in the same pace here.
The only way that you can have any anonymity feature that you can (sort of) trust is to have the whole process completely transparent from the beginning. This includes the "whitepaper" phase where the whitepaper should be reviewed at least by technically inclined members community (ideally they would be reviewed by costly professionals/experts).
Believe me, this is how I usually go. Many users know that I do not hide info and I'm very talkative in general. Please look at this from my side: Have you seen anyone in here that has the technical skills and is willing to offer any productive criticism? Or one that isn't posting only to serve his own purpose?
I don't understand how the CryptCoin community couldn't find such obvious holes in whitepaper and it had to be reminded about them by some anonymous fudster on Reddit.
I wonder how someone can audit something that does not exist yet. Allow me to remind that this whitepaper is actually a draft giving a very general idea on how it is going to work.
This will pose additional risk for investors because they can be victims of any random fudster on Reddit at any time. It really doesn't matter if the arguments of this fudster are valid or not. As long as they look reasonable they will have effect on the price. If the arguments hold water even after they are reviewed by more technically inclined members of community, the price will drop much more.
So the CryptCoin community has still the chance to change it's development methodology to a more open and transparent. This will also lower the risk for investors and probably attract some fresh money, including my money.
The task I was assigned is not to attract more money in the coin but to deliver a certain functionality.
People say that I play with words or semantics. But sometimes the difference between words can be huge.
Just to sum up: I do not answer to anonymous posts because I'm not obligated to. Anyone can come forward and discuss with me with his real username just I am using my real username.
Or do you think that I couldn't change my username since it looks like it's common practice and appear like I'm an alien intelligence came to spread knowledge to poor humans?
I'm not that type though and I will never be. I'm here and I will do what I'm supposed to do.
Will it have flaws?
I expect it to. That's why we will keep implementing till it meets certain QA requirements.
Did I ever claimed my solution will be resistant to any kind of attack present or future?
Of course not, on the contrary. Just read my previous posts. Anyone who claims that is a fool, the least.
Why did I chose whatever method I chose?
Because they are tested methods that do not deviate from the primary protocol (bitcoin).
I stated many times that I would never trust my coins going to an intermediate node for further processing. That's why there are no nodes that will get any coins from transactions. Plain and simple.
I strongly believe that these methods are very hard to trace. Besides, there are even more ideas as next milestones.
Why don't I announce them?
Because I do not want to be accused of trying to manipulate prices (which is something I do not want to have anything to do with). I will announce next milestones as soon as we have a successful test.
And as a ps:
1) If I say more info about current and future plans, I will be accused of being manipulative (and scammer)
2) If I don't say anything to avoid no1, I'm accused of being "shady" and even non-existant.
There is no way one can avoid rumors, good or bad. The best thing one can do, is let them say what they have to say. The best response (imho) is by presenting work when it is done, than posting in threads that tend to be manipulative (no matter if it's positive or negative).