the most obviously broken thing about the web is that when something goes down you can't just like ask someone else for that thing even though a computer can hold like one million things and sends and receives thousands of them them all day long
hey i'm trying to get a copy of this book, but the library is closed for the day. i get this book like a hundred times a week and so do a bunch of my friends but there is literally no way they could give it to me unless they were a total book hoarder and manually made a scan of it to keep for themselves in this occasion. there is one guy who keeps a copy of all the books and they seem to be doing alright for now but nobody knows how long they can just keep holding all the books for everyone.
if instead we just sorta held onto them for a minute and there was some notion of place where you could be 'near' to a page looking for it and ask around, or had some friends you could ask, that would be cool. someone tells you beforehand 'hey i'm gonna keep a copy of this book so if you ever need one just ask and here i'll prove i am keeping the right ones in advance'
someone is going to come by and tell me some fancy stuff about dns or cryptography but you know i'm right. it should be possible to ask other people for stuff. i know some of you go to the same websites i go to, and i wouldn't ask someone off the street for a favor unless i really needed it but i'd take a website from a friend even if it was a little moldy
@jonny I feel like Hyphanet and IPFS both tried to accomplish this and it just turned out that nobody wanted to be responsible for hosting the things they visited online lol
@jonny @ansuz @jonah I've thought of adding that to @agregore but haven't been prioritizing it because it means you're leaking your browsing history to the network. A bad actor could see all the pages you're seeding with enough effort. 😥 Sharing between friends could make it less risky but even then there's a bunch of risk / hassle. I'm down to build it if someone wants to help crowdfund :P
@jonny @ansuz @jonah Alternately one could use something with IP privacy and zero accounts / names to reduce ability to identify. Requiring the reader to ask a specific URL hash could be good to reduce ability to get *all* hashes. Also have some sort of detection for history brute forcing and block the peer. Lots of options tbh. Not easy and lots of gotchas however
@mauve
@ansuz @jonah @agregore
Yes of course, I would want something like a probabilistic service model that mixes all the pages I visit with all the pages my neighbors in k-space visit or something like that. That is not like "threat modeled" or anything just vibes, but ya straight up serving my browser history is no good but there has gotta be a way