Is it possible that you and I may be operating under different definitions of what it means for a system to be decentralized?
To my mind, email is a 100% decentralized system that we have today. This, to my mind, suggests that we fully capable of designing, implementing, and widely deploying distributed systems. Can you help me understand where I have erred in this logic?
We are fully capable of designing them. Good luck getting consumers to use them. Particularly using decentralized at every step of the process, in a way that people use Google or Facebook as a one-stop-shop for getting everything done.
If a decentralized solution is not easy to pick up, and does not offer clear everyday benefits to users over existing centralized solutions, then it will not win with users. (Just "better privacy" is not really a good everyday benefit for the average person; compare DDG usage to Google.)
email can use IP addresses using the "direct send" protocol. DNS servers are provided for convenience, and they're widely used, because they are much more convenient than the direct send protocol.
Now you're just adding on features hoping to win a battle you've already lost. But while we're on that topic: spam can use on-server text filters, AI processing, and IP blocks...
If you choose to keep going down the rabbit hole, please learn first about how email works.
Email is centralized in the exact same way Facebook, Twitter, or Netflix are centralized--probably more so. There is less CDN, etc. infrastructure built for email unless you are one of the really big players.
Emails don't need CDNs. They don't need images. They can be sent directly from one computer to the other.
There isn't a large infrastructure set up for email because it's so damn efficient that a 1990s era computer can serve as an email server for thousands of people.
Every organization can host its own email programs: servers, clients, spam filters, etc. Most businesses and companies still host their own email.
To my mind, email is a 100% decentralized system that we have today. This, to my mind, suggests that we fully capable of designing, implementing, and widely deploying distributed systems. Can you help me understand where I have erred in this logic?