We should also be preparing the infrastructure for a global network that's asynchronous, anonymous, and encrypted. It doesn't seem all that unlikely that the whole notion of a real-time internet is going to collapse or become infeasible for anything but entertainment approved by the corporations and states. And of course that has already happened to some extent. We should maintain a relatively slow grassroots samizdat network. Even the remotest regions should be able to receive dispatches through couriers, and in regions where packet-switched internet is somewhat operational we can distribute the samizdat at net speeds. I suppose I'm thinking of an "interplanetary" network. A network fit for the Firefly universe, if you will.

There already are efforts to address this with projects like the IPFS. However, without mass adoption, it seems very unlikely that an alternative internet would emerge before the increasingly-likely collapse of the open internet.

Somewhat agreed. However, there is enough repression to provide incentive for semi-shady people to work on these things. Unfortunately piracy seems to be the biggest incentive for secret communications, and while that is indeed productive, it incentivizes high-bandwidth torrents for large files, which is not exactly in line with the most crucial needs (although of course there are important uses for distribution of videos and photos). An interplanetary low-bandwidth Usenet might be the most crucial necessity for the upcoming dystopian future. So I suggest we start encouraging the spread of subversive poetry and politically uncomfortable discussions, to incentivize such a grassroots network.

You're best off figuring out other avenues to do this - HN is preaching to the choir.

You need a choir to make a chorus.

> However, without mass adoption, it seems very unlikely that an alternative internet would emerge before the increasingly-likely collapse of the open internet.

I am not sure what you mean by mass adoption. Sure, there has to be a critical mass for it to work. But I don't think that everyone and their dog should use it.

Part of what made "the good old days" of the web so awesome was that you needed to put in some hours to actually use it. Thanks to this, it wasn't that interesting for people who didn't want to contribute (somewhat) meaningfully.

It doesn't necessarily apply to IPFS, but the increasing amount of "mobile only" users significantly limits what sort of input people can provide.

You should tell Kimble that message, he's currently determined to launch MegaNet using mobiles to store data when they're idle in a p2p network.

Right, and it's not a bad use of mobile devices, but my point is users can't introduce anywhere near as much input. If everyone were to go in that direction, written word - not to mention programming would take a huge blow.

Am I the only one who prefers the current internet over IPFS? IPFS exposes my IP to anyone who wants to know about the websites I'm visiting. As seen with BitTorrent, that power gets abused significantly. I would prefer a IPFS+Usenet solution, separating the nodes from the clients. What would be even better is if I could choose to connect exclusively to [any country here] nodes, and then I can feel comfortable knowing my traffic is remaining within borders that I can trust with privacy laws.

I believe that the underlying transport mechanism is not part of IPFS; You can use IPFS over Tor to insure privacy.

Also, you can run nodes on other systems. This is effectively what `https://ipfs.io` is.