First it’s not totally accurate, it’s gated by cpu power but the difficulty is very low.
The reason is because it’s not possible to ensure one person per account without a central authority.
Even with a central authority it’s difficult, see Facebook manipulation.
As the space system progresses though you will be able to have verified/authenticated sets of users more similar to a traditional site, but there aren’t any real precautions against alts.
The reason is because it’s not possible to ensure one person per account without a central authority.
Assume you mean one account per person? (Preventing one person from having multiple accounts, rather than preventing one account from being used by multiple people.) Alts can't be prevented on these little reddit clones but they still function, those who care enough will be able to cast a few votes instead of one. Once the reward for manipulation gets too high (reddit) you have no real hope of preventing sophisticated efforts.
I don't see why it would be significantly worse for a distributed system. Unless you're just talking about one person signing in on two nodes (without system checks) getting two votes that way... but that's too minor to mention.
Yes one account per person, shared accounts aren't as big a deal.
I don't see why it would be significantly worse for a distributed system. Unless you're just talking about one person signing in on two nodes (without system checks) getting two votes that way... but that's too minor to mention.
This is the fundamental problem. Traditional reddit clones gate signups by ip address to some extent.
This doesn't work with multiple nodes.
Right now this might not seem like a big deal because the system is more federated and there are only a few super peers.
But once significant numbers of people are connecting indirectly through p2p (how this is going to scale without breaking the bank) that approach becomes worthless.
So my approach is to build a very open system, and tools for building more restrictions on top.
You can create as many accounts as you want on nab, but maybe someone could offer a service for getting accounts and you could have a voting system that only looked at the recommendations of those accounts.
The proof of work voting approach is just something good enough to get things going while I build out the rest of the system. It can easily be replaced or supplanted. One approach would be to replace the current pow voting with a cryptocurrency of some sort and this might be something I explore eventually if nobody else does first.
I'm using tor, many people who care about privacy will be too. Gating signups by IP address seems archaic.
A blockchain ledger seems like the obvious solution for verifying important transactions like account sign-ups over distributed nodes, and it ties in nicely with your present proof-of-work voting system, in that you could use votes to generate hashes for your blockchain.
I assume you know about zerovoat - seems relevant.
Personally, I think the biggest hurdle with notabug is generating critical mass (I know you don't need that now, but you will). Foiling censorship/corruption via distribution is a great selling point, and really I think that's all you need - I'd be concerned about going too crazy with completely unproven systems for website operation, as they might repel users. There are a ton of users out there who are basically happy/familiar with a reddit-like system.
(post is archived)