#asknostr
How could we get rid of the ips of relays?
At the end, we could have events storing ips to npubs.
If we can get these notes, we can get the ip of what we want.
But we need the ip of the relay that stores this.
We could say, that we connect to any relay, and then each relay would store links to other relays, so that we can resolve our npub to what we want.
But how do we select the first relay to connect to? It seems it has to be a fix ip. But if it is fixed, it can be rugged.
What do you think?
nostr:nevent1qqsyvfmhxy4qe3kavjfsk4d4hkphszw8f3078x0ymuwm7xjwaxe8y8spz4mhxue69uhhyetvv9ujumn0wd68ytnzvuhsygpwl2n3twa5dh2mu6ma4rthqqnx6yt8fwgnhqtc4hd4ctnrmxrnxypsgqqqqqqsqdqe6x
#plebchain #grownostr
Finding seed nodes is a much bigger problem than nostr, even bitcoin has hardcoded domains for finding the first peer.
In practice I don't think this is actually a big issue
Why do you think it is not a big issue?
You can hardcode a large list and only 1 need be honest & operational for your client to find more honest peers.
If 99 of 100 hardcoded peers are down, you're still good. And on the next client update you change the hardcoded list to another working set of 100
I totally agree. In any reasonably useful network there will be plenty of "first peer" candidates.
When you're starting from the assumption that "someone has heard about the network" then it's almost a guarantee that however they heard about it can be a first peer or one can be found from this source.