Hacker Read top | best | new | newcomments | leaders | about | bookmarklet login

You mention the filter effect: I think what really killed Usenet, at least for me, was spam, much of which was automated, with no great mechanism in place to combat it.


view as:

These days you could fight spam simply by adding a proof-of-work requirement for posting content to the network. Isn't that exactly what this Web3/Blockchain fad is all about anyway?

This is a really interesting use-case. I don’t think posting spam is really a zero-cost thing though. So we wouldn’t be adding a cost, we would be increasing it. So I guess the big question is, how much can we charge per post that spammers couldn’t afford but wouldn’t scare away actual community members?

I think making a non-free Usenet is pretty much anathema to the whole concept of Usenet.

I wouldn't want to pay Real Money for it, but somehow it doesn't seem the same to say "I'll spend some finite amount of computation time/energy to make this comment". That seems like a much lower hurdle.

At the same time, paying ten cents to make a comment on Reddit or HN is very unappealing.


10c of crypto mining is about 30 minutes on an RTX 3090. It would have to be a much, much lower cost!

Wouldn't proof of stake be much more promising?

1) Start off with a trusted group of people (stakeholders)

2) Set up an invite-only system

3) Inviting someone means sharing some stake (i.e. some reputation) with them

4) To post you need some minimum amount of stake

5) With every post that's not downvoted into oblivion you increase your stake until you reach a certain equilibrium point. (So there's a limit and it's not about gaining status points.)

6) Posts that are downvoted into oblivion will cause their authors to lose some stake.


This is the best explanation of proof-of-stake I've seen yet! Thank you! This does sound just like minimum karma limits to me though. Per-community karma though instead of the global kind that Reddit uses.

This was in fact one of the earliest use cases of proof-of-work from the 90s [1].

[1] https://en.wikipedia.org/wiki/Hashcash


>These days

HashCash is older than Bitcoin, and Bitcoin sybil-attack protection is basically HashCash.


That's been attempted for email too, not sure how far it ever went. https://en.wikipedia.org/wiki/Hashcash

Legal | privacy