Centralization is bad for everyone everywhere.
That bring said… I just moved my homeserver to another city… and I plugged in the power, then I plugged in the ethernet, and that was the whole shebang.
Tunnels made it very easy. No port forwarding no dns configuration no firewall fiddling no nothing.
Why do they have to make it so so easy…
The trouble with cloudflare is that there is just one. It’s one of the best registrars out there, the only free/cheap and usable DNS host (have you seen what route53 charges per zone??). That without getting into the whole tunnels and DDoS mitigation end of things, which is nearly unique at any price point.
The problem with cloudflare is that we’re missing three other cloudflares to move to if they decide to pull evil shit.
The bigger trouble is creating a CDN has a stupidly high barrier to entry. You literally need your own data centers across the world, your own server infrastructure, the man power to manage it, etc.
You could try to host it on a cloud provider but you’d go bankrupt even quicker. Unless someone were to try to build a co-op run CDN, it’s just not gonna happen without a profit motive and a large amount of capital.
I mean the optimal cdn is maximally distributed to reduce load and latency right. Unfortunatly the web was not built in a manner that supports this.
Eg if we could have a single url for the same object that could be served by any server that is part of the fediverse then the fediverse itself would be an optimal cdn.
Perhaps we should take some notes from peertube. Plus more legitimate bit torrent content on the internet as a whole is hardly a bad thing make the isp’s jobs harder for places without net neutrality.
Look up Anycast when you get a chance.
I consulted with professor gpt and it seams that it’s basicly just giving the same ip address to multiple servers meaning that any of said servers can serve as that ip.
Also it seems said ips require paying large sums of money to isps. My poiny was more that with the current mainstream internet (http websockets etc) it would require you to run a local service/proxy that can interpret a global id and route to basicly any small server with said resource. Unfortunatly i dont think its possible to build such a thing that would just work across browsers if embedded into a standard webpage.
That’s true. The bizarre paradox of the centralization of edge infrastructure is real.
That said, the other edge-lords (haha) could offer similar functionality, but they chose not to.