I agree with this, at least in general. Linking content location to content ID is a serious mistake.
I'm not convinced it's worth investing time or effort into something that might make it harder to address the problem of URLs in the future. I get "solve one problem at a time", but I feel like URLs are so fundamental of a problem, you have to get them right first before you move on to other problems. And for the most part, we have gotten them right -- build whatever web replacement you're working on on top of DAT/IPFS, unless you have a really good reason not to :)
A decentralized web is a web where content can be re-hosted and verified by anyone.
I'm not convinced it's worth investing time or effort into something that might make it harder to address the problem of URLs in the future. I get "solve one problem at a time", but I feel like URLs are so fundamental of a problem, you have to get them right first before you move on to other problems. And for the most part, we have gotten them right -- build whatever web replacement you're working on on top of DAT/IPFS, unless you have a really good reason not to :)
A decentralized web is a web where content can be re-hosted and verified by anyone.