Hacker News Reader

Comment by scrollaway | Hacker News Reader

Distributed discovery is possible, but you quickly run into spam/noise problems. Centralized communities generally have features such as curation and deduplication (duplication is bad for torrents, as it causes peers to be spread across multiple slightly-different packages of files, reducing the available peers for both).

It's not a technical problem, it's a social one.

View on HN

It's a social problem that is caused, though, by the technical constraints of relying only on BitTorrent, rather than adding other technologies to the mix.

I started to write out an example of what it would look like to fix the problem by adding some more layers around the BitTorrent layer, but it got a bit long; sort of turned into a feature spec:

https://gist.github.com/tsutsu/0441fffa6b989121018b5f970d1f4...

The new steemit.com alternative to reddit is entirely powered by Ethereum on the backend. It works remarkably well and I think a similar system could power a decentralized torrent site.

It might even allow for benefits like people who post torrents get tipped every time someone upvotes.

It wouldn't stop the domain name wack-a-mole, but it would allow a server hosting the domain to just host static files that pull in data from a distributed ethereum based API.

It would also allow for local applications to operate that work off the same chain so people could still use the service even if the servers powering the HTTP based interfaces are shut down.

I wonder if mediachain [1] could be repurposed as a decentralised tracker.

[1] http://www.mediachain.io/

In a blockchain-based system, proof of work could give users points so they could spend them to show their trust on a file (or maybe transfer them to whoever they trust so they can release files with a number of points from the start). For the duplication problem, solve it like IPFS does, the file's hash is it's Uri so a duplicate would just point to the same resource. Actually, people could build KAT-like sites entirely in IPFS and users loading (and thus, serving them to other users) said sites would be the trust mechanism.

Proof of work burns money and I guess users won't pay for some content's reputation score, there are no incentives to do it other than altruism. POW-based currencies work well because you can exchange the points for stuff or services, not because you can spend them within the system (proof of stack does exactly this indeed and its long-term reliability has been questioned many times).

Reply on news.ycombinator.com