Kind numbers are inherently centralizing.
Took me way to long to see that.
Discussion
It's a limited namepace that doesn't even tell you anything about what the thing is.
At least with MIME tags (m/M) you know what the thing is
I don't think that is correct. I think the acceptance mechanism for Kinds and NIPS is . . . really irksome.
It's an open protocol. Do what you want with it.
Hashes of feature descriptions with a nonce.
Ow, I don't grasp what you mean by that sir.
🤔
Instead of letting people pick nip identifiers, which requires a central authority to organize, identify the nip by a hash of it's content. If I want to create a nip for secure blackmail I could create the following definition
{
title = "nostr blackmail",
description = "End-to-end encrypted message from a burner npub that will be automatically followed up by another event revealing the embarrassing secrets if no zap is detected."
nonce = "16 Byte base 32 encoded random integer."
}
This way even if someone else has the exact same idea the nip hash would be different allowing competing implementations.
Ahaaaaa!
Thanks, interesting 🤔