I like Moxie's work and writings, and this article has some great points, but I can't get behind this:
We should accept the premise that people will not run their own servers by designing systems that can distribute trust without having to distribute infrastructure.
I'm not ready to give in. I am happy to leave "normal" (tech illiterate and politically apathetic) people behind to reach my decentralization goals.
I think instead of building centralized infrastructure that does not require trust, we can make it easier to host decentralized infrastructure. Including allowing a "server" to be offline for months at a time, come online for a minute or two, then disappear again. P2P networking is also an area we can improve on, IMO. Too much information is going across the internet instead of point to point. Bluetooth is a terrible protocol, but airdrop (and reverse engineered implementations) seems to be promising.
Email doesn't require DNS. Modern spam solutions for DNS do. You can most definitely use something like `spiped` to create a mutually authenticated, secure channel over IP, and just send mail over that. Or build a VPN overlay network and send mail to raw IPs. If you're going to cloister yourself with your fellow monks^W nerds then this is simple.
Hm, I think my point might not have been clear enough. I would find it hard to function without interacting with central system, like sending an email to someone on gmail. Just today, I emailed a plumber on gmail, but it could easily have been an old friend or a relative or whatever.
> I'm not ready to give in. I am happy to leave "normal" (tech illiterate and politically apathetic) people behind to reach my decentralization goals.
I don't think we have to "leave "normal" [...] people behind". I don't like devices like Alexa, but FFS, look at what millions of people have installed and running 24/7 in their homes. Is someone seriously telling me that a dedicated engineering and marketing effort couldn't build a similar consumer-centric device that functioned as a server (purposes to include but not necessarily limited to http and smtp).
> Is someone seriously telling me that a dedicated engineering and marketing effort couldn't build a similar consumer-centric device that functioned as a server
> I'm not ready to give in. I am happy to leave "normal" (tech illiterate and politically apathetic) people behind to reach my decentralization goals.
Which should be already possible with with the current offerings around selfhosting applications and p2p technologies.
But as the same time you need to accept that the "normal" people would probably be happy to, in turn leave you behind to reach their goal of being able to use all service available without needing to concern themself with running their own server.
NNTP with enforced GPG authentication and PoW like spam prevention could work today (in the narrow technical sense, not in the wide product sense.) It wouldn't even be that large of a lift from current NNTP architecture. Create a moderated Usenet group that only accepts posts that complete a PoW challenge and that sign their messages.
We should accept the premise that people will not run their own servers by designing systems that can distribute trust without having to distribute infrastructure.
I'm not ready to give in. I am happy to leave "normal" (tech illiterate and politically apathetic) people behind to reach my decentralization goals.
I think instead of building centralized infrastructure that does not require trust, we can make it easier to host decentralized infrastructure. Including allowing a "server" to be offline for months at a time, come online for a minute or two, then disappear again. P2P networking is also an area we can improve on, IMO. Too much information is going across the internet instead of point to point. Bluetooth is a terrible protocol, but airdrop (and reverse engineered implementations) seems to be promising.