Namecoin is the best example of a real world use case, a proper replacement of a DNS with no central authority. This is clearly one which is very difficult to replicate reliably without a blockchain.
All the blockchain systems have serious scalability issues with, for example, bitcoin transactions times slowing remarkably over time. I don't see how this would be a practical match for a DNS system.
Moreover, at the end of the day, even DNS is a matter of "real world truth" and there's no reason a vote by miners would be particularly reliable source of this truth.
For namecoin, the blockchain doesn’t need to be involved with grabbing the data, just creating a public record of intent.
Right now, accurate propagation of settings throughout DNS kind of assumes a lot. I don’t think Namecoin captures incentives properly either, but I can imagine some blochchain system connecting the dots.
Even if blockchain DNS systems were production quality - which they're far from - why is a 'central authority' inherently bad? The entirety of cloud computing, and the modern internet, is based on the pretense of giving your entire infrastructure to a centralized authority. "Oh, but what if your centralized cloud provider is actually evil and incompetent", the free markets allow you to change providers until you have a provider with absolute highest quality.
> Even if blockchain DNS systems were production quality - which they're far from
That's not an argument against building something...
> which they're far from - why is a 'central authority' inherently bad?
They're not. Sometimes you want a central authority, and sometimes you don't. Both modes can be good. But previously central authorities were necessary due to the physics of DNS. Blockchain enables a new mode. Now the two modes can compete, and we can reach a new equilibrium with both, or one can win out. That's what's cool about it.
>That's not an argument against building something...
At which point did I argue building things is bad? I actually took a lot of time to learn Ethereum smart contract development just to get to the bottom of the whole blockchain ecosystem. Learning and building are almost never bad things.
I'm not bashing the idea of competition between blockchain and traditional centralized authorities either. I'm extremely pro-competition. I'm bashing the fact that all - yes, all - blockchain applications provide negative value compared to their traditional counterparts. The argument that blockchain enthusiasts use to justify their support for such low-quality infrastructure is this philosophy that having no centralized authority is, by itself, valuable. This is what I'm bashing.
If my cloud provider is bad, I can switch to another one or self-host.
You can't switch from the ICANN, no matter which domain you chose, it depends on them, they effectively have a monopoly. I want to self host my domain name.
There's a fair point to be made there. But then I would ask which solution is best: lobbying hard enough to democratize the organizational structure of ICANN, creating a competitive ecosystem, or resorting exclusively to a ledger of entries with no capacity to be modified according to legalities, attacks, or otherwise? What if someone spams the ledger with too many names? What if someone takes domains that should be illegal for them to possess? What if someone takes someone else's private key? What if an organization gets their private key stolen? The vectors for exploitation don't just disappear because you have a decentralized system. The exploitative behaviour just changes.
All these issues currently exist with the existing domain name system as well.
> What if someone spams the ledger with too many names
You need to pay in coins to register a namecoin domain so this attack won't work.
> What if someone takes domains that should be illegal for them to possess?
This should be up to courts to decide if the name is illegal and handle it over, not to some random organisation like the ICANN which has no legal basis.
> What if someone takes someone else's private key?
Same as if some gets access to your registrar password right now, nothing changes with that.
> What if an organization gets their private key stolen?
Again, same as before, nothing is improved or changed by a blockchain here.
Moreover, at the end of the day, even DNS is a matter of "real world truth" and there's no reason a vote by miners would be particularly reliable source of this truth.