Malicious computers caught snooping on Tor-anonymized Dark Web sites

Yes clearly the client needs to store a cookie somewhere (config files, db, browser) but the server needs to parse it in order to authenticate the user.

Thanks for the link.

> This sort of attack on the Tor network by large groups of malicious nodes is inevitable in a borderless community where there is no procedure for vetting or ID verifying new node operators. Tor assumes ideological loyalty from its node operators, that's a fundamental security assumption, but at the same time, a part of the Tor ideology is that there's no vetting procedure for nodes, nor are there any real rules for running them beyond a few trivial ones, and anyway without the former there's no way to enforce the latter. Tor is built on a giant contradiction, essentially.

Yep, exactly. How exactly do you build a strong anonymizing service all the while assuming benevolent nodes? Doesn't make any sense to me.

Too much naivete in the programming profession, I feel. It makes for pretty hilarious fails periodically.

EDIT: For the downvoters: I apologize if the language is too blunt. But those of us who aren't actively developing Tor have the benefit of being slightly more objective towards it -- that's how I feel. I might be harsh by expressing it, in which case I am sorry; I don't aim to be ungrateful, only skeptical of its current technical design philosophy.

> I appreciate your response. The fact that Tor, or any other service that attempts to anonymize its user base, is hiding behind the "we can't promise you full anonymity" legal safety net, isn't helping any of us who want to be anonymous anyway. I understand they have legal issues to deal with and must use evasive language so as to be not eliminated easily by anyone disagreeing with them. Still, it doesn't help us the users either.

"We can't promise you full anonymity" isn't a legal safety net, it's just the truth. There is not and will never be a system that can let me use the internet but prevent me from posting "My name is David Kerkeslager" on Hacker News. And while that may be a contrived example, note that if I signed into this hacker news account over Tor or sent my address in cleartext to a seller on the Silk Road, that would expose my name as well. And Tor can't prevent programs on your computer, such as your browser, from posting information about you, either, because those programs need all the capabilities that a user has.

Tor anonymizes your IP address and that's pretty much all it does. To fully anonymize yourself on the internet, you need to use a browser that doesn't leak other information about you and to not give out information that can de-anonymize you. That doesn't mean Tor isn't useful; you need to anonymize your IP in order to be anonymous on the internet, and Tor is by far the most thoroughly-tested tool for doing that.

> As I pointed out, IPFS indeed has limitations, but same as Tor, if it doesn't get wider adoption, it won't ever become better. Being a programmer, I believe I can properly time the moment we should stop trying to repair something that looks like it's already working against us, and just start over. As a human being, I recognize I might be wrong and I agree with your counterpoint.

I looked up IPFS and I'm still confused by you proposing it as an alternative to Tor. I don't see anything in my brief look that indicates that IPFS is even trying to provide anonymity--they're solving different problems. There are alternatives to Tor, like I2P or Freenet.

The problem with rewrites, especially rewrites by people who don't understand the previous program, is that they make all the same mistakes that the previous program made. With Tor, thousands of highly-qualified people have tested and sifted through the source code and fixed many (hopefully, all) of those mistakes. If we discard Tor and start writing something from scratch, that's what will happen. And while we might be able to fix those mistakes again,

> I have no evidence that Tor is compromised, but if we're to believe what we're told via the official news, what another conclusion would you draw? It's an honest question, not an attempt to shift the burden of proof. Objectively speaking, I have no evidence or proof. If I had any, I think there's a reasonable chance I would be in prison because the FBI doesn't like citizens knowing how do they penetrate their cyber targets.

I think it's wise to acknowledge that there's some chance actors with extraordinary capabilities can de-anonymize Tor. However, I don't think we can draw any definite conclusion. And as I said earlier, there are still plenty of actors who almost definitely can't compromise Tor.

> Again, I agree with you. I know many possible malicious actors don't have the tech and/or the talent to deanonymize Tor users. But IMO the whole point is to make every single malicious actor powerless. In that regard, it seems Tor is failing a lot in the last 1 year.

I don't agree that the goal is to make every single malicious actor powerless. That's an entirely unrealistic goal. I believe the goal of Tor is something like, "Allow people to connect to computers over the internet without exposing their IP address to the computers they're connecting to." And as far as achieving that goal, I think Tor has a very strong track record.