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.
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.
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.
SQL was great years ago. But after working with things like MongoDB, MNesia, and blockchain tech, SQL based databases just seem inelegant.
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?
As long as someone has the page, it won't go down.
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.
Sqlite is preferred for embedded databases which seems to be what this app is using it for.
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 :)