Replies (27)
DHT works well with enough redundancy. It's weak if there aren't enough instances of each data object stored on enough clients. Think when you're downloading torrents and one of your pending torrents only has a couple seeds and they both drop off.
Looks like trash. Took a conceptually good idea and totally fumbled the implementation.
it's well publicized, quietly centralized garbage
nostr isn’t p2p, therefore it works
By synonym
(A tether company)
Do what you want with that information
Punky might be shit, but the youtube video above is talking about the underlying tech, which may or may not be shit.
The core idea is retarded: look up your home IP by a public key, the data distributed over DHT (bittorrent).
Do you want people to find your home IP with your npub? If you do, you have never heard of opsec.
It’s p2R2p which is superior
Punky yes, but Kademlia might have value
If there is no way around this, then yes it's retarded, but torrents can be used safely through a vpn
Can the systems be integrated together? I don't see why not.
I love being told the network I use reliably on a daily basis has failed. I hope it continues failing like this indefinitely.
I'm running a publicly reachable nostr relay on my home IP, but my IP geolocates to Denver, CO, about 1000 miles from where I live.
Same thought
Let's continue to fail together bro 🫂
*beginning search in a 2500 mile diameter donut around Denver*
No, you dont understand. That is the purpose of the "pubky core". It's essentially looking up IPs with an npub. If you want to run a "homeserver," which supposedly is the entire purpose of pubky, you post your IP alongside your npub and this is published so people can connect to your homeserver. It's literally the dumbest idea I've ever heard of. Why not just enter the person's IP address directly? Well, I suppose you get the user's public key as a result that can be used to open an ssl connection, but that is not how it's being pushed. I doubt he has a solution to multiple pubkeys being published for the same IP, but if he does, kudos - he solved a problem for a solution nobody needs. It's claimed to be "better than nostr" but it provides zero anonymity, and it solves none of the problems that nostr does.
The guy is just an attention whore trying to get people to dox themselves with his retarded software.
That is the exception, not the rule, and that information can be subpoenaed from the ISP and yielded trivially.
Guessing the pub key serves both the encryption and DNS, basically. Yeah this sounds dumb. There's some obvious ways to both make this private and add some distribution of encrypted data for redundancy. I like nostr's ad-hoc client-relay architecture. I write my events to my own home relay and also forward them to a few public and invite-only relays. But everybody publishes relay lists also so clients can request their content from other relays not on your normal relay list. It's all interoperable and you can customize it however suits you. How is his idea better? 🤣
"Technically more elegant" he claims.
There is no DNS, that's my whole point. It doesn't "improve" DNS, it replaces the entire concept of a domain name with a public key random string of characters. There is no "domain" or "name" as part of it. It's really nothing more than a wrapper for the IP address, which, is equally difficult to remember. That's the whole idea behind a domain name, not that domain names really are anything today other than a vanity-plate for your IP address, since nobody can remember your domain name either.
We’re the most failingest.
The fact that Pubky is trying to do SMS verification proves that they're not really serious about good design.
It's all released under MIT licence.
Devs need to feed themselves.
Tether have lots of coin.
Although - agreed - not a good look.
