pull down to refresh
604 sats \ 7 replies \ @justin_shocknet 12 Mar \ on: Why no decentralized namespace yet? bitcoin
It's been tried a countless times, by myself included... here's some of what I've learned:
ICANN DNS isn't really broken enough and censorship of it is an imaginary problem. If Alex Jones and the State of Iran have functional websites, censorship isn't a valid concern.
Simply because of how the internet is architected, whatever system we try to displace it with still rides on legacy DNS. All we can really do is abstract it which doesn't solve for much.
Usability is an almost impossible problem too because of the above, everything browser related depends on SSL. So now you're not a naming system but inventing a whole new application stack.
Name squatting is a positive market function, and new systems make bad trade-offs trying to prevent it. Squatting must be possible since the point of authoritative namespace is because scarcity, so it must be market liquid.
My next stab at it will be non-authoritative naming from a social graph, where there is no one true "MyCompany dot com" for example, but rather, MyCompany is keyword derived from you and your social graphs address book- that resolves to a public key.
That's the other thing, we don't really need a new name->ip address system, we need human readable names for Public Keys for things like Nostr.
An arbitrary suffix can scope these keywords into contexts (.nostr) for example
thanks for sharing. these are great lessons.
for the record, I'm less interested in DNS-replacement specifically and I agree that there's a lot of legacy rails on which we would still dependent.
I'm more interested in the "human readable names for public keys for nostr" idea. it's a mess to share my npub and it's hard to explain to non-technical people why we need to do nip-05 verification
reply
Glad we're zeroed in on something... To help me think about solutions and my products mind answering some Q's?
What in particular prompted the OP?
In what situations would you share your npub directly vs. a URL or QR code?
What have you seen non-techies encounter with nip05? This in context of a custom domain or Nostr apps not doing it automatically?
reply
OP was prompted by a reply I got to a different post I made. (explained in the details of this post actually!)
I would like to have an npub on my twitter bio, in my email footer, etc. I have seen a lot of people use Primal links to share their profile like this: https://primal.net/dk
I am a supporter of Primal and I think they do a great job with what they're building, but primal.net/dk is yet another pointer to where my npub can be found which is my actual unique identifier.
I just see non-techies stumble when they can't figure out why they can't just choose a username like every other social network they sign up to. It feels awkward and strange when they already have a mental model/pattern for how this "should work".
reply
Fair to say then a unique identifier thats usable by multiple nostr apps not just Primal?
If you're sharing the npub already that's ideal, but only works where copy/paste is available... I think you mean a search term that users can use in whatever app they're already into? (as opposed to a URL)
can't just choose a username
Bingo, I see this too.
I think this is at the crux of the username problem, there's no consistent application of them across the Nostr apps because every search would have to respond with strictly an opinion on who
dk should belong to.My ruminations on this are as follows, we should only need a very loose framework for arriving at those opinions. It can't get anymore fractured than it already is.
By self declaring
dk, having followers that label you as dk, perhaps some other metrics from their relay that add to a rank, Primal could easily give you weight over a lesser and rogue claimant of dk.Reputational conflict resolution.
This is non-authoritative and not truly unique ofc. It gets interesting when another platform sells influence, and takes money to assign the name
dk to someone else in their social graph that users may bootstrap from. Since their social graph differs, users weighting that platform resolve a different dk. This is basically what we have today.Which begs the next question,
Is the incentive stronger to sell reputation, or is consistency with social consensus rewarded more?
My theory is the latter, going against social conensus will eventually render what you're selling worthless. If correct, its relatively simple from a technical standpoint because it's still mostly open to interpretation (worst protocol theory) how to weight a social graph.
Only one way to find out, back to work...
reply
I think it's a mistake to think of namespace as just DNS. Assume you had a namespace that is decentralized, it does resolve to a single name and is universally accessible. It would create an alternative to trademark laws and other institutions created on the foundation of the nation-state for managing intellectual property, reputation, etc. It's a missing piece for decentralized identity. When you combine a decentralized network that enables you to control assets (bitcoin) and another one that does the same for names you can build an alternative to the State.
reply
I agree on DNS not being a target for replacement, my focus is on human readable names for keys at this point as I mentioned.
I disagree though with the step-change of utility in tokenizing titles. We can already tokenize any digital thing, but namespace is an abstract social thing.. This is why decentralized namespace hasn't worked yet, because approaches have been technical-centered instead of socially-centered.
Physical things, even if traded digitally, are ultimately enforced in the meatspace where they exist. The internet is made of physical servers, datacenters, powerplants... This inevitably results in an apex force that is the most efficient at killing people and breaking things in an area... that's not going away no matter what you want to call it.