XNS: eXile Name System

i just asked the llm cause idk wtf about what but that's always refreshing to see contributions like these
and seems like there are some confusions in the thread regarding XNS and DNS so llm was like:

The confusion

XNS is often described as "DNS on Monero" but it's more accurate to call it a censorship-resistant public-key directory. It doesn't replace DNS resolution — you can't look up an IP from an XNS name. What it replaces is the trusted third party that decides who gets a name. The actual resolver (to convert key → reachable service) is still needed.

What's missing from XNS as "DNS"

1. No record types — can't store A/AAAA/MX/SRV records
2. No hierarchical delegation — every name is flat
3. No resolution protocol — no equivalent of dig or nslookup standards
4. No DNSSEC — integrity comes from Monero blocks, not cryptographic chain of trust
5. No caching/TTL — each indeXer decides freshness
6. No reverse lookup — no PTR mechanism
7. No subdomain support — foo.bar.name may not work (OP didn't specify subname structure)

Where XNS excels

• Tor/I2P — onion addresses are already Ed25519 keys; XNS gives them a memorable name that can't be seized
• PGP/WOT — replace WOT with a Monero-backed claim of eravsar → key
• Self-sovereign identity — your name is provably linked to your key, no registrar involved
• Censorship-resistant publishing — even if a government seizes your DNS domain, your XNS name still points to your key

Summary

XNS doesn't compete with DNS — it competes with registrars and certificate authorities. You'd still need DNS (or Tor) to get packets to a destination. XNS tells you who owns the name; DNS tells you where the server is. They solve different problems.

@SpinningCat it is still a "domain name system" even if not focused on traditional networking, instead it is better to call it that it considers cryptographic identifiers as the network unit rather than IP addresses thats assigned to you by your ISP. Just like how I2P, Reticulum and Tor does it. Thats the cypherpunk way and XNS is made for that. Thank you for checking it out.

@eravsar That's interesting, and I agree that what you are building adds real value and I'd definitely prefer to use it than existing solutions. It seems unnecessary to limit it just to tor and i2p because everyone needs DNS but tor/i2p are not worth it for most websites. If you could find a way to extend it then I'd definitely use it for my http links, even if it requires the user to run a daemon. In the future I see more people relying on technology like this to evade access censorship because digital ID is an active threat.

@eravsar

This is a fantastic idea.

I've been looking at ARweave, storj, Nostr, and filecoin.

One year costs 0.01 XMR

Sadly, $300 per year for a domain name excludes me, and always will.

Who sets this price, and is it variable ?

If variable, based upon which market conditions ?

@CharliePrimero mafs, that's $3 a year, not $300 lol

@CharliePrimero $3 a year lol. it is a protocol constant and always will stay as 0.01XMR, it is a burn transaction, no one makes profit, there is no market. to learn more on how XNS works see xns.rocks/docs

If users register a pubkey when reserving a name, you could enable transfers by requiring a signature from the currently-held pubkey. This will almost certainly enable secondary markets, which may or may not be desired.

@jeffro256 yes i thought of it, but secondary markets is not desired. my focus with XNS is to make an utilitarian name system for those who need it.

for example Namecoin had transfers, and the result was this:

"Our extensive empirical analysis of Namecoin reveals a system in disrepair. Indeed, our methodology for detecting “squatted” and otherwise inactive domains reveals that among Namecoin’s roughly 120,000 registered domain names, a mere 28 are not squatted and have nontrivial content. Further, we develop techniques for detecting transfers of domains in the Namecoin block chain and provide evidence that the market for domains is thin-to-nonexistent."
(source: https://econinfosec.org/archive/weis2015/papers/WEIS_2015_kalodner.pdf)

so not having transfers is intentional. it removes the incentive to squat names and keeps the protocol minimal.

this is XNS in action everyone. read about how XNS can be used as the DNS of Tor here: https://xns.rocks/docs/xns-and-tor/xns-on-the-web

@eravsar can the visitor part be incorporated into a browser extension that can be added to the Tor browser ? Or at least a Tor browser fork with the resolver built in ? Or does it have to be its own separate program ? You can tell I'm not knowledgeable about this type of tech.