Hacker News Reader

Comment by dumbfoundded | Hacker News Reader

As someone building in web3, I'll give an example. When I sold my first business, escrow.com wanted >$20K for the domain transfer from godaddy. If you sell a domain on ethereum (ENS), the escrow requires only a smart contract and fees scale with the network so like $50 right now. I'm working on connecting the two.

Intermediaries are a huge reason the web is the way it is. If we can eliminate them, the internet looks very different. My biggest prediction is that fundamental protocols like TLS will altered to integrate with blockchain protocols.

Web3 isn't just a scam or waste of resources. It will enable us to do things we already do but much cheaper than we do them now. Lots of smart people are flooding into the space. It will take at least a decade but it's already happening.

View on HN

> As someone building in web3, I'll give an example. When I sold my first business, escrow.com wanted >$20K for the domain transfer from godaddy. If you sell a domain on ethereum (ENS), the escrow requires only a smart contract and fees scale with the network so like $50 right now. I'm working on connecting the two.

A quick look at https://www.escrow.com/fee-calculator shows that fees only reach $20K if you're selling a $2.2 million domain name. At that level, lawyers should be involved on both sides anyway and your $20K fee is better spent on proper legal services to get it done. Escrow.com is notoriously expensive.

ENS domains may allow for smart contract transfers, but the downside of this is that anyone who gets access to your keys can now permanently own your domain name for that same $50 (likely higher) transfer fee. For companies, this means that a single security slip-up or even a single disgruntled employee with the right access can forever destroy your company's internet presence with no recourse. The current domain name system isn't perfect, but there is recourse for hijacked and stolen domain names.

Nothing comparable exists in the blockchain space, specifically because people don't want it to. Sounds great if you're sitting on the sidelines, criticizing other people for their poor OpSec when things go wrong. Doesn't sound great when you're trying to secure a business and realize that loss or theft of the domain keys means irreversible destruction of your domain name.

> At that level, lawyers should be involved on both sides anyway and your $20K fee is better spent on proper legal services to get it done. Escrow.com is notoriously expensive.

Yea so, that's kinda the point of what he's saying. Why is it that a domain transfer for $2M is so much more expensive than one for $2? Intermediaries.

Smart contracts eliminate the intermediary. Whether you are selling for $2 or $2M, it's gunna cost the same small fee (and the fee is going to be negligible within a year).

> this means that a single security slip-up or even a single disgruntled employee with the right access can forever destroy your company's internet presence with no recourse

This is actually extremely easy to avoid, to the point that it's far more secure than DNS (domain names get compromised via social engineering of registrars all the time, btw). With ENS, for example, the "owner" can be divorced from the "controller", so you can "own" it with a multi-signature smart contract wallet, or even lock it into a smart contract and make it non-transferable with a time-lock and require a quorum of signatures to unlock.

Any mechanism you can imagine basically, but you have full control of how it works. Point is, it's entirely possible, and easy even, to make your crypto assets VERY secure from hacks, it's just the current market isn't that interested in it yet (most easy solutions would be too gas expensive atm).

You seem knowledgeable so maybe you can help me with this. So you said:

> Smart contracts eliminate the intermediary. Whether you are selling for $2 or $2M, it's gunna cost the same small fee (and the fee is going to be negligible within a year).

Okay, fair enough intermediaries suck. Let's replace them with smart contracts.

But what happens if there was a bug in the smart contract? What happens when a situation such as death occurs that wasn't considered when the smart contract was written? Who settles such disputes?

Governance tokens & DAOs are helping to bridge the human divide. Not everything can be coded, humans have to make judgement decisions sometimes. The smart contract is just the agreed upon structure of cooperation.

For cases like escrow over digital goods, it should be perfect in a p2p context without human involvement. It already is for any web3 good like tokens or nfts. Bugs happen but these contracts are often simple and used lots of times. We'll settle on more and more common standards to implement simple transactions.

That's entirely fair, but the perspective of those in web3 is that this property exists on a spectrum.

The present day incumbent side of that spectrum is that the DNS registrar could decide one day to just take your domain from you, and there's not much you could do about it. Sure, you could sue, but that takes time and meanwhile your business could be ruined.

A more concrete example of this is social handles. Earlier this week, there was a story about a girl that had the Instagram handle @metaverse, and Facebook decided to just take it from her with no recourse.

Our contention is that shouldn't be possible.

But here's the thing: if you prefer to have a centralized entity control your assets just like today, you have the choice to do that! Everyone that has their assets on Coinbase today is making that choice, and that's totally fine.

The big innovation is, they can decide at any time to take custody of it themselves if they want, or move it to another centralized holder! Just like you have the choice to leave your money in the bank, or you can withdraw it to cash and keep it in a safe in your house. We're just giving users that same choice for digital assets.

> Our contention is that shouldn't be possible.

Trust me, this will always be possible, "web3" or not. The proper solution here is maybe something federated like Matrix and definitely has to do with encouraging more people to self-host their own services, possible with "web1" tech.

I think it would serve this initiative a lot better if you just completely left the notion of currency, value, and markets _all_ the way out of it.

Focus on IPFS, IPNS, federation, BitTorrent, etc. if you want this to succeed. Leave money out of it. Money is invariably going to be loved by the winners and hated by the losers. It also makes it incredibly challenging to trust incumbents as they have everything-to-gain (and potentially alot-to-lose) by shilling that which increases the value of the coin they hold.

> Intermediaries are a huge reason the web is the way it is. If we can eliminate them, the internet looks very different.

That's what makes me the most skeptical. I don't see a lack of intermediaries in crypto and the crypto intermediaries are far sketchier than their non-crypto counterparts. Just compare Tether to your local bank. Tether could pull the rug on USDT tomorrow and your recourse would be... what? Same with NFTs and the entities hosting the actual art you're paying for. I can have some confidence that Steam will be around 10 years from now; not so sure about boredapeyachtclub.com. The most notable party that's removed from the equation by crypto is the only one I have some small measure of influence on - the government.

From the outside it looks like: "meet the new boss, same as the old boss".

Some things are cheaper in web3 than web2 and web3 can do everything web2 can so it's all going to get rebuilt. The same thing happened with mobile phones. Mobile completely changed the internet and the only "invention" there was the form factor and battery. Everyone all of a sudden needs an app and the whole internet was rewritten for IOS & Android.

The promises for how it changes society are pretty weak. It could change nothing or a lot. Web3 is just a better cheaper tool for humans to use so it'll get used. How it changes society is up to the people using it. Just like how the internet didn't usher in a utopia, web3 won't either, but it will change a lot.

Isn't the whole point of escrow though to trust a 3rd party service? Or maybe I'm confused about what purpose escrow.com served here. Was ecrow.com literally just facilitating sending x dollars from person a to person b and b/c it was a high dollar value it charged a high fee?

Escrow is using middlemen to prevent any potential loss or theft. Every transaction has some counterparty risk. Who goes first, the domain transfer or the wire? To be sure you won't get robbed, you deposit the money with the escrow agent and the seller transfer the domain to the escrow agent. If one party falls through, the escrow agent gives their property back. In web3, the smart contract is the escrow agent and 100% automated. So you deposit your money, the seller puts in the ENS name, and now instead of a % basis for the fee (escrow.com is like 0.9%), you get a flat cost of running the smart contract based on the network fees. It's just like how you can transfer $1 of bitcoin or $1B of bitcoin for the same price.

> If you sell a domain on ethereum

what’s the incentive to sell a domain on ethereum?

> TLS will altered to integrate with blockchain protocols.

you mean the Transport Layer Security, the cryptographic key exchange protocol? what’s the advantage to integrating with a blockchain compared to the way the current protocol works?

> It will take at least a decade

what’s stopping the community from doing it faster? and wouldn’t a decade mean that meta’s metaverse will be long established before this community takes off?

> you mean the Transport Layer Security, the cryptographic key exchange protocol? what’s the advantage to integrating with a blockchain compared to the way the current protocol works?

Sort of lends to be like PGP Web of Trust doesn't it?

If TLS signed the messages with a public key, you put any TLS conversation on-chain immediately. Some web2 apis already implement signed responses that can be used in this way. TLS doesn't though, so you have a provenance problem. You can prove the integrity of the data but not the participants of the conversation. I'm building the piece that works with TLS to add identity verification.

Ah see, but you work in the space and probably hold any amount of cryptocurrency at all, so you're morally compromised and fundamentally a scammer.

Anyways, sarcasm aside, that sounds excellent - have a link to the project?

Reply on news.ycombinator.com