ZeroNet – Decentralized websites using Bitcoin crypto and BitTorrent network | Hacker News Reader

ZeroNet – Decentralized websites using Bitcoin crypto and BitTorrent network

github.com

115 pointsSyrup-tan25 comments

Anyone interested in ZeroNet's architecture may want to take a look at their presentation: https://docs.google.com/presentation/d/1_2qK1IuOKJ51pgBvllZ9...

While I like seeing that this uses relatively strong technologies like Tor, BitTorrent, and Bitcoin, does this beat Freenet for any use case?

Looks like very slick and easy to use. And there is play on it (https://torrentfreak.com/play-p2p-impossible-shutdown-160301... on hn top 10 at the moment)

Yes, it comes with a p2p SQL server and readable namecoin domains. Due to BitTorrent it's also much faster than Freenet. While it doesn't require Java it's still crossplatform.

Sites are essentially bitcoin addresses (public keys) with ownership and modification rights controlled by corresponding private key. The same crypto provides passwordless logins and authentication!

Since it supports Tor, users can also use it anonymously.

The system is impressively elegant.

it would be nice if the docs would mention how to get/set up those .bit domains.

They do: https://zeronet.readthedocs.org/en/latest/faq/#how-can-i-reg...

Depends - it should win out for speed and ease of use. But anonymity, no, in order for a user to possess certain content, they have to download it explicitly and know they're sharing it. It's trivial to know who visits a given website and, for example, send DMCAs to all of them. Freenet prevents this in multiple ways - proxying your connections through others and using encryption to prevent those without the keys from reading the data or even knowing what it is. Freenet probably provides the best anonymity of any current network as it's done in such a way that once a file is on the network, there's just no way to trace it back to its source.

The people using Play on it might find themselves in some fun legal issues soon, depending on country.

You could use it over Tor I suppose - that's what the authors propose, but that seems less than optimal to me and doesn't even approach the sort of anonymity Freenet could.

Looking at it briefly, Tor support is a bit misleading. It only protects content authors and seeders, and not viewers. (Someone could just inject an external image reference to reveal their true IP address as HTML seems to be not sanitized.) This seems to be the key difference from Freenet.

Well, that's an easy fix - view the page in the Tor browser too. Or better yet, run the whole rig inside a VM which is forcibly routed over Tor transproxy. Of course - this is sort of Tor abuse and freenet is the better solution for this purpose, but if you need to view something over this in (relative) anonymity, it is still possible.

While it supports tor, ZeroNet is more about decentralization than it is about anonymity.

Mark my words. This is the ethereum killer.

The problem with ethereum is the novelty factor. Everything under the hood is difficult for the average developer to grasp and it's not clear whats going on as it is quite new and work is on going. Another issue is that there seems to be no clear 'voice' from Ethereum and instead you have many people claiming to be an Ethereum expert. It killed all of the hype and excitement I had in December and January.

Why I think this is an ethereum killer is the tech that powers it is very much mature, proven to work quite well and will be easier for developers and users alike.

But more importantly the end result is what matters. This has been in development since 2014, roughly same time as Ethereum but the end result is far more impressive than what Ethereum offers and ironically it is because ZeroNet offers less is what makes it so attractive. For example, using Tor, bitorrent, bitcoin, already existing and well established technologies provide excellent privacy and anonymity. Sure, ethereum might be a much stronger cyrtographic platform but I think accessibility, speed, scalability, adoptibility gained from ZeroNet's setup is what ultimately blows Ethereum out of the water.

The codebase seems super light and accessible too compared to Ethereum.

Anyways, I could be wrong of course but I think ZeroNet could be it. Look how long Tor, BitTorrent, Bitcoin protocols have survived and grown in adoption rates. Ethereum is trying to pull off all three and as a result it's losing developer passion imo. For instance the forums are less active and ironically attracted lot of politically fueled angsty types who believe in conspiracy theories.

My understanding is that the primary value of Ethereum other then its decentralized nature is the capacity for smart contracts. A cursory examination of this project doesn't look like it supports that.

Right. ZeroNet and Ethereum don't compete with each other. Ethereum will never be a decentralized CDN the way ZeroNet or IPFS are.

I doubt this will hold true in the long term. I build stuff on top of Ethereum and there seems to be huge interest and financial support for the platform.

It is true that Zeronet works out of the box, but Ethereum seems to provide cleaner abstractions. The Web3 vision includes Zeronet functionality (decentralised applications, dapps). I think it all depends on the integration of Whisper and Swarm, which may result from a collaboration with the IPFS project and the Serenity PoS transition. The fact that there already exist node implementations in 7 different programming languages makes me optimistic.

Another interesting project is Hyperledger/OpenBlockchain which is mostly focused on consortium/private chains but provides a pluggable consensus model.

https://github.com/openblockchain/obc-docs/blob/master/white...

If the Ethereum core devs can take the project to that level of standardisation, they win. Unfortunately, its dependency on a currency and the focus on the public chain will also attract some of the worse parts of the Bitcoin community. One of the major risks is also the need for backward compatibility.

In the long term, code does not matter, standardisation does.

What are ZeroNet's main advantages over IPFS?

ZeroNet makes it easy to serve dynamic content with its p2p SQL support.

I don't know.. That just seems inelegant.

SQL was great years ago. But after working with things like MongoDB, MNesia, and blockchain tech, SQL based databases just seem inelegant.

I don't think it is SQL over P2P but rather it just uses SQL internally for syncing. Or at least that is the impression I get from reading: https://docs.google.com/presentation/d/1_2qK1IuOKJ51pgBvllZ9...

Sqlite is preferred for embedded databases which seems to be what this app is using it for.

That's not as bad as I was thinking. It sounded like (and documents are sketchy) that this was some sort of shared SQLite db over their net. It sounded rife with potential problems and interesting injection issues. I've dealt with enough bad SQL statements and poor websites with injections and all. Poor bobby tables.

Now, I thought about how to handle a similar dynamic data issue with IPFS: We had Usenet years ago and it's still creaking along in a few places. What if we bring that tech back, to an IPFS era? I'd be eminently glad to get rid of 'forums', and go back to what we had prior.

The only piece that was missing is how we maintain all the different groups and posts. That's now solved by a blockchain containing the root pointing towards other 'chains, and then to posts themselves. I'm seeing it in my head how to construct this, but it could be relying on faulty understandings too :)

Not sure what you mean by "dynamic" because both systems publish static files and register a link to a root document with a name system (namecoin or ipns).

I had this running as a node on a RPi2 for a few months. Very light duty and did not consume a lot of resources. Every so often zeronet pages would go dark when the owners made the keys private.

Hmm. I'm curious about this - it seems that since this is torrent-based, archivability should be trivial - as long as on person still possess the data, regardless of what the author does, it should be possible for other users to reach that data.

The issues you speak of (with a page going dark) are strictly due to things like DNS changes, right? So other users can still visit webpages by connecting directly to the torrent some other way - right?

Yeah. There's no reason a site would "go dark" unless either there's no more seeders for the site (everyone deleted it off their computer, or the people who have it aren't running ZeroNet anymore) or the Namecoin .bit domain expired and no longer links to the ZeroNet address (at which point you could still access it through the long address that will never expire).

As long as someone has the page, it won't go down.

Does it use bittorrent mutable keys BEP?

The answer is no.

Reply on news.ycombinator.com