I avoid hackathons π Yes, I'd like to think the things I've built for Nostr bring value to people and solve real-world issues. A decentralized "youtube", a next-gen relay implementation, an anonymous client and more. I'll have to take a deeper look into Pubky core. Is it as easy to build on Pubky as it is Nostr?
Discussion
The problem is that all of the things you are building, regardless of how easy it might be (compared to what, normal web dev?), would be censored at scale because Nostr relies on ICANN/DNS.
The relay model is flawed in that it is just a reconstruction of websites to separate the client from the content. But it behaves no differently than having your data on multiple websites.
Pubky uses public keys as domain names themselves, where the user controls where they point.
This ends up changing everything, even allowing us to choose different data models, and establish truly open and decentralized social graphs and search indexes.
Nostr can't do this stuff without faking half of it.
What about RIPE, Nominet or me.
I run my own DNS with my own naming structure, but even then I don't need to, I use static IP's for some services.
I run my own authoritative name server for me, it subcontracts unknown searches to RIPE or Nominet and these subcontract to other root servers if need be.
IP is the only censorable part, I can run local IP's or I can set my own routing using my AS number to redirect public IPs as a wish, but mostly I wouldn't.
IPs are censorable, not DNS, but even then, I could use my AS number to route around any censorship, I could even advertise higher than the censored AS's to remove the censor.
Ultimately if a state actor like a firewalled country wishes to prevent me accessing content, then they can install a bridge. If I want to stop you accessing my network, I can physically do it. But that's how it's supposed to work, my IP my choice.
There are many 3rd party DNS alternatives that try to solve a problem they don't understand doesn't exist.
Your public key is something you can share with others, yet control. It's portable to any network paradigm in that sense, thus it also transcends IP and DNS.
However, that is only interesting if you have a confident discovery method. We use PKDNS by placing your signed records into the largest, oldest, most decentralized network on the planet.
You can try and reduce this to "any old name server will do" but every other options falls below PKDNS, even when you reduce it to a basic spectrum.
Should people use more than one discovery network? Sure, we all do, right? We use many identities, agents, hosts, networks, etc. But few of those could be interoperable or support user-oriented graphs.
Really? Do tell me more about public keys π
To be clear, I'm not 6 years old.
Just so you know, I've had around 50 companies fail to pitch me on how to "solve" DNS for the past 30+ years.
I'm not here to shit on your parade or to attack you or to invest in you, I'm simply giving you a little bit of background from somebody that actually built the Internet we use today.
I wish you well.
ββ¦ from somebody that actually built the Internet we use today.β
So itβs your fault the WiFi here is shit. Thanks Mike!
Um, okay. I was just replying within the context. I don't care about your personal claims, or your money. Do you have a relevant response or refutation of what I said?
Also, consider this is social media, and I am writing for the benefit of onlookers, not necessarily to persuade you specifically. I already know teaching Nostr cultists is a waste of time.
NOSTR cultists, love it π
I guess I'm also an Internet cultist, a Bitcoin cultist, an electronics cultist, a corporate cultist and a logic cultist, oh, also a cooking cultist and a cruising cultist.
I'm sure there are some other cults I'm in as well.
Good luck going forward.
I'll leave you with this thought.
There is a difference between solving a problem and offering a better solution.
If you focus on the latter , you'll have more chance.
Last piece of advice, you are clearly having difficulty interacting with people if you have to resort to insults. A key to success is not shit on anybody on the way up, you'll be amazed at how quickly all the people you shit on will bring you back down again.
Be nice, it costs nothing and works incredibly well, even if you fake it.
Genuinely, I wish you well and every success.
You acted as if you wanted to have a design conversation, and then resorted to boring style policing when it got real.
It's clear you have nothing useful to say. I doubt you know much about Bitcoin either.
Zzzzzz
Very interesting conversation.
Thank you.
You've helped me make a decision I was unsure about until now.
Interesting! Got a link or something for more detail?
I could write a book I suppose? π
In the mean time:
This is how root servers work in DNS:
https://en.wikipedia.org/wiki/Root_name_server
Here is an example of a typical DNS zone file that would be stored on your or your ISP's DNS server
AS Numbers are described here:
https://en.wikipedia.org/wiki/Autonomous_system_%28Internet%29
BGP routing is how the Internet decides how to route the majority of traffic.
This is a really interesting topic. It also bothers me that this relay model relies so much on dns. But there is the .onion option, which makes it a bit better maybe.
Also considering, just from the point of view, that nostr is such a stupid simple protocol, it probably wouldn't be that difficult to use a different model at some point, to piggy back it on some other tech/transport layer at some point. Especially if the need for it comes.
Its one of those things where things just catch on that are "good enough" and reach some network effect instead of the "best" thing.
I'm torn between nostr (or a nostr-style simple protocol) on top of a transport layer P2P network, and a "full stack" solution like pubky.
If you have more thoughts about it, technically or philosophically, let me know.
I'm also torn, but I guess lets just wait and see. I have been in the crypto space for long enough to realize that there is no perfect end solution, just an endless flow of new stuff and some things just stick even if they are not the best solution. You can use the perfect solution, but you might be alone there, in the end the value is where the people are.
Just like with bitcoin, I could make a case for some xyz coin being a bitcoin killer and having better tech, more private, faster...etc. But its just screaming into the wind. Network effect is a bitch π
But the positive thing is, that even nostr, although not perfect is still 1000x better than traditional social media and maybe total secrecy is not the ideal. Maybe its a good signal for society to see which country blocks nostr relays, then you at least know that that country is shit and you should consider moving out. That is quite a valuable signal in some way. Also it gives countries that are more free an edge, that empowers them with the inflow of citizens, users, revenue from datacenters where people host their relays/other services...etc.
I'm just rambling here, the main point is, that "good enough" usually wins long term, without us knowing or controlling what it will be. right now, even traditional social media is "good enough" since it doesn't censor "that much" yet for it to be a problem that people would talk about. And currently centralized social media is a winner.
I think the Pubky design actually offers people new powers, otherwise I'd agree that "good enough" is acceptable. I sincerely want people to be able to accomplish anything together and the Semantic Social Graph method, combined with PKDNS unlocks that. I'm less concerned with the hobbyist pocket social media community use case, and more interested in changing how everyone uses the internet to establish a better economy and society :)
I totally agree, I just got used to not getting stressed if my favorite thing doesn't gain traction and I am happy with good enough, just an immune system I developed through the years π I'm all in on trying out new tech. Any idea how to get the invite code for pubky?
they can keep you out of it, if you need an invite code, obviously.