Seems some dont realize that nostr relays are powered by http. The first call to a relay is an http request which is then "upgraded" to a websocket (connection). Given that it's http, it just works, can get through a firewall, where most p2p fails. Many dont realize this, in fact, fiatjaf didnt even realize this, until I explained it to him. The http part is what gives nostr a chance of working where p2p inevitablly fails.

Replies (12)

I had the same discovery when I was working on the Http/Guzzle client in PHP previous week 🧐
One of his many lectures, I dont recall which one. I used the term "based" quite loosely, it was more the case that SMTP was there, and people were familiar with it, so he tried to keep it close to what people knew. Same with HTML, it was based on SGML. I could ask him one day perhaps!
Honestly I thought it was the other way around good to know. I whiteness every day how closely related they are (along with ftp kind of) but http is much more simple IMO. > I could ask him one day perhaps! Hopefully you get that opportunity, I would like to hear about it
I normally talk to him every week. We are working on improving http, or tweaking it with some extra properties. He mentioned nostr for the first time last week, when I showed him NIP-98 auth. We are going to try and add it. I also talked to him about his project before the web which was called "tangled". You may enjoy his book, if you are into this stuff.
Pretty much, yes. Firewalls are designed to stop you. You might get somewhere but it's a lot of work. And then fall back to relays. So right tool for the right job. P2P for some small things, relays for bigger things, http for largest things. Simply ask yourself, in 10 years time is amazon going to switch to P2P? Is facebook? Is google. P2P is a good tool, but wont win.