Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Control-F: distributed (0 results)

Control-F: peer (0 results)

Control-F: p2p (1 result)

Can we seriously not come up with a way to make a distributed reddit?

I imagine running a client on sandstorm.io which is pulling content from subreddits that are each hosted in a sandstorm.io instance.

Is the problem with that that it becomes too easy to doxx someone? That we like having a centralized authority that we can "trust" for authentication / authorization / pseudo-anonymity?



I had worked on one a few years ago, which used a DHT to exchange messages which were signed GPG JSON. At one point I had it using a Dining Cryptographers-style algorithm to post messages fairly anonymously. I don't recall if that stayed in.

One problem with P2P these days is that you need people to run servers, which they don't typically want to do. I made a small RaspPi box they could run, and planned to run a public instance for people who didn't want to go that far.

One problem I ran into is how to stop terrible posts. Freenet/etc have a problem with people sharing illegal and harmful material, and I didn't particularly want to help with that.

Additionally, when you create an alt-platform, it gets used be people who are kicked off of mainstream platforms. I didn't really want to end up creating a service primarily used by pedophiles and nazis.

I think there are ways to solve the technical challenges, but it's still difficult to solve the community-management in a distributed/sane way.

(The solution I came up with btw was opt-in mods. People could publish a ban list, and you could subscribe to whomevers you want. But I don't think that's a great long-term solution)

I still think there is some good work to be done in this space, and I may circle back to it someday, but the community elements are vexing.

You can see my version at https://github.com/e1ven/Tavern


>Is the problem with that that it becomes too easy to doxx someone? That we like having a centralized authority that we can "trust" for authentication / authorization / pseudo-anonymity?

Dox, send child pornography through, use to sell narcotics, plan hitman assassinations with, etc, etc. Not to mention all the fun stuff organized bands of miscreants can do to other people. Just look at Wil Wheaton's adventure with Mastodon for an example.


So, I happen to believe in government regulation. Is your assertion that in the predator-prey relationship of free speech vs speech regulation, that the government cannot possibly keep up? That any (new?) forms of communication will enable people to engage in illegal speech, and so therefore are bad?

I'm not trying to egg you on, I'm genuinely curious what you think our attitude should be.


My argument is more that people would prefer to not be on a platform where they are at the mercy of organized miscreants or one that is known for lots of illegal activity. So centralized platforms provide a layer of safety that people want. It's really hard to make a platform that allows broad speech, is actually decentralized/p2p (ie: rather than just federated), and is safe from organized abuse.

As for the government, I believe the government is fallible, corruptible and that privacy has value. The government can keep up but in doing so you tend to lose more freedom than you had before. For example, Australia's new laws on encryption which would, combined with mass traffic surveillance, lets them know everything you say. So I'd prefer the government not care enough to actually solve the speed regulation problem because doing so puts dangerously much power in it's hands.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: