Are there any alternatives to the Internet Archive that are built around P2P, so that everyone can contribute to hosting/sharing web archives? Seems like having all these important archives hosted by a single organization isn’t the best idea for longevity/redundancy
There was a ActivityPub wiki clone, no idea where it got to.
The major upside of IA being built and owned by one central company is trust. We can (so far at least, if I’m wrong please correct me) trust IA to not censor/rewrite history. As soon as every man and his dog can contribute, that gets a lot harder to guarantee.
Edit: https://github.com/Nutomic/ibis
Don’t take me linking it as endorsement, I think federated wiki’s for anything other than fandom stuff to be madness.
Sounds like a potential application for a blockchain techs, that allows to do verifications, voting and consensus.
Ive had similar ideas on verified unaltered media to combat fake news. Such an archive but for journalism would be a great start.
Yeah, quite possibly. Could still be very hard to get right. Region blocking might make consensus difficult.
Edit: just occurred to me, any method of consensus could be used to ddos sites as well. Might be best left for people smarter than me
I think that proof-of-work approach to blockchain can make ddos attacks much harder, but I’m not an expert too :)
I figured that every node would need to scrap the site, in order to validate the content. If there are thousands of nodes, that would ddos the site.
I don’t really understand how PoW would solve that, can you explain?
with cryptocurrencies the blockchain is distributed, its not stored centrally. your idea could work similarly
I think it can be done like a NFTs on top of Bitcoin. In this case evey archived page is NFT and all the blockchain is available, so there is no centralized cite. If each action will require some computations (PoW) then ddos attack or spam attack will be very hard to implement.
Thats for proving its untampered with right? I’m more thinking of validating the archive copy is a “true” copy when adding it initially, which requires each node to check against the live site?
Its definitely an intriguing idea though, but I don’t know enough to know how feasable it can be