Hacker News Reader

I prefer torrents inside I2P [0], because otherwise my IP will be shown to everyone. I wonder why there are so few people there. It's of course slow, but it's worth it. There are already quite a few torrent trackers.

[0] https://geti2p.net

What kind of speeds do you get? Because last time I tried it getting above a few megabits per second was a struggle. When we're talking about even an Ubuntu ISO, it's a couple gigs. It's too slow to want me to keep using it.

Indeed, few Mbps is typical for the current number of users. It can take days to download relatively large files. Still, is speed more important than anonymity? Definitely not always, perhaps never.

I hope it gets better with more users, as it happened with Tor.

> I wonder why there are so few people

Maybe because people understand that it's security model is flawed: the peers in the mesh treat all the other peers as not trusted, that's okay; but the network design doesn't seem to take into account potential hostility of the physical medium itself.

Say, your ISP can tarpit packets or shape traffic, or shut down the power or cell tower in your block temporarily, and then measure how the mesh congestion changed. At the same time the mesh relies on building hop chains with TTL of 10 minutes iirc, which I think makes a peer a sitting duck. It is (was) all documented/leaked.

AFAIK garlic routing is still harder to crack than Tor: https://geti2p.net/en/comparison/tor. The main advantage of Tor is the number of users.

The network security people I know stopped bothering publishing attacks on I2P's directory service--which was(/is) in fact subject to eclipse attacks--because I2P didn't seem to care to fix the issues. Here's a great paper.

https://sites.cs.ucsb.edu/~chris/research/doc/raid13_i2p.pdf

>The network security people I know stopped bothering publishing attacks on I2P's directory service

Maybe those you know stopped, but others didn't: https://geti2p.net/cs/papers/.

Great... are the issues getting fixed? The point I was making about I2P working so poorly that some people have given up trying to help--in addition to specifically noting that eclipse attacks do, in fact, affect the design of I2P--doesn't seem affected by "there exists people who still spend time demonstrating I2P doesn't work well" ;P.

I doubt the number of nodes is important when you can split a large group of them into a few subgroups and analyze them or alter how they interact with each other, then repeat subdivision; because again, them overlay networks are like smaller sandboxes inside the big one.

Don’t forget that you have to do all this within 10 minutes, because after that time the network configuration will be totally different.

Yes, overlay network's configuration will be somewhat different, but physically it will be the same. It's like you're rearranging apple bits in a pie to save them; the pie will be sliced and eaten anyways.

> the pie will be sliced and eaten anyways.

The difference is how long it takes and how much resources. There is no perfect security. Security is not a "yes/no" thing.

All correct. It's just important to remember who's the attacker:

  Alice -- ISP1 -- 1000 hops -- ISP2 -- Bob

Alice and Bob's identities are protected from each other, but not from the govt because ISPs collect metadata.

> because ISPs collect metadata

Which metadata? ISPs only know your next hop. Check out how garlic routing works.

> There's only very few projects (that I know of) that are really addressing resilient, organized, curated content dissemination, and they're essentially divorced from the pirate scene. Too bad I guess.

Why would someone trying to build something like PeerTube Or IPFS want to explicitly support piracy? All that does is bring negative pressure down on them and we see in this very discussion, the MAFIAA is more than happy to abuse laws they wrote to bring pressure. Just because someone is interested in or supportive of peer to peer communication doesn't automatically mean they support piracy.

I'm not trying to say piracy is bad, it is just illegal. As shitty as those laws might be they're still on the books and people face penalties for breaking them.

I suspect that it's easiest way to get much users.

In a strange twist of irony... This was what copyright was meant to do: promote a diverse body of works while protecting existing ones.

It should come to no surprise that notable torrenting sites have many low-effort imitators. If anyone could copy stuff already out there, why spend the extra effort to innovate?

I'm all for open-source, but I recognize the value of copyright (and how it protects FOSS-licensed material) and healthy competition and innovation.

"Copycat" in this context didn't mean literally copying TPB (although that did happen and was actually made possible by them dumping their DB at some point IIRC).

Most torrent sites do seem to have an own code base. It's not like they run the same software (or at least they have different themes, I never actually checked).

It is also neither surprising nor bad that the sites still have the same goals/design basics. They all solve the same problem, so they will end up looking similar overall. No problem with that.

What is being criticized here (by me, I don't wanna put words into Warg's mouth since it's an old, probably misremembered interview) is that the way that they deliver that "product" (in this context: the torrent website itself, not the hosted torrents' contents!) in an unimaginative way without taking into account the changing battlefield.

Fully i2p torrent sites? Niche.

Local DHT scraping with an overlay net of torrent index exchange à la yacy or magnetico? Niche.

And apparently: hosting anywhere that is not immediately susceptible to direct US influence, i.e. NOT on namecheap+cloudflare? Still kind of niche. Wtf?

Thanks for the clarification -- I am in agreement with your points.

I recently noticed that LibGen has started using Cloudflare as one of its download options. I was somewhat surprised that Cloudflare would allow use like this.

well if they didn't notice before, they will now, that their engineers can read it on HN :D