Namecoin Forum

I ran across this again.

magnet:?xt=urn:btih:cb7caf0b4c0ee266cce5fbd8f8ba3903f5efa82e&dn=bootstrap.dat+-+Bitcoin+Blockchain+%40+295000&tr=udp%3A%2F%2Ftracker.openbittorrent.com%3A80&tr=udp%3A%2F%2Ftracker.publicbt.com%3A80&tr=udp%3A%2F%2Ftracker.istole.it%3A6969&tr=udp%3A%2F%2Fopen.demonii.com%3A1337

Turns out the link can be quite short: magnet:?xt=urn:btih:cb7caf0b4c0ee266cce5fbd8f8ba3903f5efa82e
We could shrink it to:
"btih" : "cb7caf0b4c0ee266cce5fbd8f8ba3903f5efa82e"
or
"btih" : "y3yvC0wO4mbM5fvY+Lo5A/XvqC4=\n" (base64)

magnets it's the way to go if you want to store files on the blockchain, but seems redundant, as they still cannot shutdown even torrentz.eu , since it's a search engine.

A more usefull use case would be to store bootstrap information for p2p networks. Any new coin, or any new bittorrent-like program need some peers to start, the same way bitcoin first used to connect to some irc channel to get peers. Namecoin could be used for that.

So we may see lite namecoing-quering components embedded into p2p programs in the future.

tosh0 wrote:magnets it's the way to go if you want to store files on the blockchain, but seems redundant, as they still cannot shutdown even torrentz.eu , since it's a search engine.

Trusting a centralized torrent index not to self-censor seems suboptimal. Some torrent indexes seem to be doing a pretty good job at the moment on this front, but decentralized would be better IMHO.

tosh0 wrote:A more usefull use case would be to store bootstrap information for p2p networks. Any new coin, or any new bittorrent-like program need some peers to start, the same way bitcoin first used to connect to some irc channel to get peers. Namecoin could be used for that.

So we may see lite namecoing-quering components embedded into p2p programs in the future.

This would be an interesting use case. Can you post this in a separate thread?

biolizard89 wrote:

tosh0 wrote:

tosh0 wrote:A more usefull use case would be to store bootstrap information for p2p networks. Any new coin, or any new bittorrent-like program need some peers to start, the same way bitcoin first used to connect to some irc channel to get peers. Namecoin could be used for that.

So we may see lite namecoing-quering components embedded into p2p programs in the future.

This would be an interesting use case. Can you post this in a separate thread?

It does indeed sound interesting, but don't you then have exactly the same problem - to access the Namecoin index, you need to find bootstrapping nodes for the Namecoin network. Or do I miss something here? (Of course, if we assume that someone already runs Namecoin, then it is indeed an advantage.)

It does indeed sound interesting, but don't you then have exactly the same problem - to access the Namecoin index, you need to find bootstrapping nodes for the Namecoin network. Or do I miss something here? (Of course, if we assume that someone already runs Namecoin, then it is indeed an advantage.)

Yes, you need bootstrapping servers but it's kosher as long as there are significant free-speech use cases and the core project remains nonprofit.

I've been waiting for a low-latency WebRTC DHT to come along but the standards are in flux and there is still lots of research in this area. Then speech.is/frame resolution could utilize nameservers as well.

Some sort of ephemeral data store comes up regularly and there is a strong need for publishers need a way to transmit content signed using the credentials stored in the blockchain that's decentralized and resistant to censorship. I would love too see our P2P network upgraded to provide a decentralized nameserver/keyserver/whatever. However, I do not want to be responsible for guaranteeing anonymity for others

Ideally we would co-op an existing network, however, few of these are tuned to low latency. It is possible to get 200 ms response time, which is a lot in terms of perception but it's pretty good for a decentralized system.

Telehash is the new hotness but it isn't a traditional DHT, you can't query for data you have to subscribe to a channel and then wait. BitTorrent is currently the best solution for generic file sharing but it isn't amenable to changing content/customized responses and it has very high latency.

There is plenty of great research on new kinds of overlay networks, linear network coding is particularly exciting and lends itself well to anonymity. The author of Kademlia came up with a new overlay networking protocol based on electric routing, but then he got involved in Go and has put the project on the back burner.

Perhaps it would be best to writeup a proposal for the characteristics we are looking for on the wiki and allow interested parties to take a crack at it.

biolizard89 wrote:

tosh0 wrote:magnets it's the way to go if you want to store files on the blockchain, but seems redundant, as they still cannot shutdown even torrentz.eu , since it's a search engine.

Trusting a centralized torrent index not to self-censor seems suboptimal. Some torrent indexes seem to be doing a pretty good job at the moment on this front, but decentralized would be better IMHO.

My hunch is that the real "service" torrentz.eu and similar sites provide is a community to manage the index. Gnutella and others had built in search capacity but they can't keep the signal/noise ratio high enough compared to the BitTorrent indexes. There are efforts to provide that kind of community management in a decentralized fashion, but the UX is generally poor and I'm guessing that the income from ads on the website or subscriptions employ core members.