@jb55 it's a matter of just storing the bits, no matter where or how, just as long as they can be reassembled. All the nodes would have the appropriate means to request the parts it needs

@jb55 content-addressability is neat, i've read some interesting papers on it a few years ago, i guess the difficult thing is how to bootstrap it on top of the current internet architecture, without needing cooperation from every ISP (because that would take forever)

@orionwl @jb55 Isn't that pretty much what IPFS is? Even Cloudflare themselves are experimenting with that: https://www.cloudflare.com/distributed-web-gateway/

@orionwl @jb55 Looking at their website again, it's actually kind of cringey how they tell you that *they* will validate the hashes for you, so you don't have to do anything. Just lean back, don't run a node, don't verify. :)

IPFS does look kind of interesting, i'[ve heard of it a few times but never looked deeply into it

do you know of any FOSS projects, linux distros, etc that (also) host their websites, downloads there?

i guess for most intents and purposes torrents already fill the 'distributed file distribution' niche

@raucao @jb55

@orionwl @raucao @jb55 Iris uses IPFS to store its indexes, if that's of any help.

https://github.com/irislib/iris

@orionwl @raucao @jb55 Nevermind, I misread. I might have something related to the CDN part for y'all soon(tm) though ...

@jb55 i belive urbit's network stack works this way already