blossom is definitely a better way to do it, it's simple to add naive propagation mechanisms like forwarding queries to other blossom servers that are storing the same type of data, and then caching them
actually, after i get done with making an actually stable relay i need to write a blossom server
why not just use a p2p platform? clients can then load files via multiple servers at once, can use index and dht lookups, anyone can add servers
DHT does not scale well
scales fine just depends how you use it
'the concurrent number of users of Mainline DHT is from 16 million to 28 million' https://wikipedia.org/wiki/Mainline_DHT
SHA1 haha lol, man this tech is so old... i mean, it works but SHA1 is broken
also still used for git commit hashes... they really should do something about that