It's good validation that we've been on the right track, keeping in mind Nostr itself is an iteration of what we attempted with GUN for those who've been around long enough to remember Lightning.page (as is LNC for that matter), so it's all working out as expected

NWC is actually a direct response to Pub because we wouldn't implement WebLN on Lightning.Video, which is fine, I'm glad the imitators have proven our architectural concepts ;)

NWC isn't bad perse, its just a nerfed higher time preference version of kind 21000 that we could easily accommodate with an adapter if it becomes sticky.

The reason we're not just expanding on NWC is the lived experience of having implemented LNURL in the wild. Everyone implements specs in their services just a little wrong, any proposed changes to specs are bikeshed to hell, and actually writing specs before seeing how things work in a live implementation is just backwards imo.

Nostr only works to the degree it does because everyone uses nostr-tools in their projects, if every service implemented NIP's from spec, it'd be a disaster. That's a problem because more logic is required in Lightning node middleware than is in-scope for nostr-tools.

For this reason, Lightning.Pub and kind 21000 are a proto, from which you can autogenerate a client for any language. New working methods can then be PR'd to the proto rather than the bureaucracy of higher level NIP, and work consistently across clients.

That all said, whether its our stuff or more collaborative protocols, the fusion of web relays and Lightning is the future of the internet.

My vision since 2018 first playing with lightning has been that of an open-source decentralized censorship-resistant WeChat-like stack be built around it, and that is now most definitely happening.