Hacker News Reader

Comment by jedberg | Hacker News Reader

The only way an alternate TLD registrar will ever take off is if you get the people who own the current root servers on board. And the only way that will happen is if you figure out how they can make money from owning and operating the root servers.

My guess is that if you create a blockchain for registering DNS, and only the current registrars can issue coins on said blockchain, you might have a solution they would be willing to implement. You'd still have to pay them money to register a new domain or maintain it, but once you had the coin, you could register whatever you wanted and no government could stop you.

View on HN

And more importantly than letting you register, it avoids a .com, .net domain getting one of these on a landing page https://www.justice.gov/sites/default/files/styles/width_12/...

This has been the story for many sites, where domains are seized even if owned by a non-US person/company, with an order to the TLD that operates in the US. With a distributed DNS system like Handshake, 'in theory' this shouldn't be possible, since it would require multiple nodes to alter a domain's DNS records vs. a central TLD authority which has to abide by US law.

It's all interesting, but it still seems we are ways off from the DNS backbone (& my browser) operating with this kind of DNS architecture. It even appears it's becoming easier to take down a domain, not at the TLD/registrar level, but more so at the hosting level (e.g. Parler)

Is it the root servers that you need to get on board, or the publishers of the publications that most people use, or the operating system publishers?

For instance, if the browser makers all started resolving certain TLDs via an alternate network, then people could start publishing websites using domains managed by that network without needing to engage with the root DNS servers at all.

It would be a start, but not all of the internet is accessed through a browser. And I'm not sure you'd get Apple and Google on board unless there was a way for them to make money from it.

Those servers are only necessary for the ICANN tlds and those will continue to act as they do today, even in a Handshake future.

With Handshake, you run your own authoritative server (or use a hosted one) for your own name and that authoritative server's information will be obtainable on chain. No middle-man, no rent seeking.

I understand how handshake works. What I'm saying is that getting my mom to point her DNS servers at an alternate server is a non-starter. You have to get the root servers on board to carry all TLDs or most people will never find them.

But then the root servers can still be used to censor domains by removing them.

They aren't all in the US. Right now they remove a .com by telling Verisign, a US company, to change the registration for the .com. Since Verisign is the only way to change a .com, it works.

Many root servers operate outside of the US. If the blockchain were the authority on IP addresses, then even if they blocked the US roots the others would still have it.

Even today if all the US root servers go off-line every system would find the others and use them.

> If the blockchain were the authority on IP addresses

But it's not -- the root servers are -- by premise here. Premise is we assume zero config results in ICANN root servers having authority. My point is that given this premise, blockchain is ineffective. You are introducing ICANN as informational MITM and censor node in between user and blockchain validation.

True, you are excluding USA, if your only goal is exclude USA then this accomplishes that goal, but if your goal is censorship-proof then this is not it.

Browser extensions and apps can start resolving alternate roots using non-DNS methods for those who opt-in.

Sure, but again, that will never take off. "Install this plugin to access our website" is a non-starter for most people.

It doesn't seem different from "install this app" to access our services though? And clearly, this model has proliferated with fluidity.

Installing apps is easy, built into the phone. And comes from a curated store run by a US company.

Installing plugins is well beyond the capabilities of most people, especially since the plugins for the default browsers come from a curated store run by US companies.

...Only because we (software engineers, developers, etc...) have completely failed in managing to make our field more communicable to users, and have gone down the path of complicating the hell out of things instead of nailing down the primitives and onboarding to the field.

I'm not even talking any wishy washy inclusiveness mumbo jumbo either. I'm talking over complicating the boot chain, over complicating memory management/models/toolchains.

Browsers are a living nightmare of complexity, and I am continuously amazed at the number of programmers who have no clue about the gory details of networking.

Products created for the purpose of lock-in can't invite user participation because the first thing user-participants would do is remediate lock-in.

> … "and have gone down the path of complicating the hell out of things instead of nailing down the primitives and onboarding to the field."

I've thought much this same thing myself for ages now.

Not just "complicating the hell out of things" either. Complicating the hell out of things that should be simple ( or at least should be simplified, or automated, or something ). I see a lot of tedious activities in modern computing that appear to exist purely for the sake of the tedium time-wastery factor.

The flip-side of the same coin is the over-simplifying or "dumbing down" of some things that really should stay "complicated" (and hidden behind an "expert mode" toggle) to the point of allowing willfully uneducated morons to repeatedly "shoot themselves in the foot" and justify blaming everyone and everything but themselves for their error.

That sort of "computers should be an appliance like a toaster" mentality has contributed to a culture which seems convinced they should be able to magically jump from zero to "Hacker-Man" with no effort or learning required, the very same day they decide to "learn" programming.

Reply on news.ycombinator.com