Live data from Hacker News

Is Web3 anything?

chris-granger.com

111–120 of 442 posts

Re: Is Web3 anything?

#111

Is Web3 even a thing? Outside HackerNews, I only see it mentioned in crypto-circles, but on HN there’s another article at the top every day.

Interesting. For me it's all around pretty much everywhere, from TV ads to billboards.

To be fair: Not necessarily web3 but generally crypto, but still.

Re: Is Web3 anything?

#112
post #9

All of the examples of having to trust the government in this article neglect the fact that trusting the government is not just trusting the government, and it’s not just the government storing a deed or giving you a Social Security Number that makes it real. It’s the whole complex web of laws and regulations that created these policies that make them real, and the people (your fellow citizens) who have collectively…

Also, it's important to realize the core reason why people trust the government with these things. It's because the government has the ability to actually _enforce_ the contracts if required.

I mean, having a deed of owning some property on the blockchain is all cool and dandy. Provided that the polity where the property is located recognizes this. Otherwise you're still trespassing.

Re: Is Web3 anything?

#114
post #101

Earlier quoted context omitted.

> Democracy solves the Byzantine agreement problem just fine already. That’s nice when you and your counterparty are part of the same democracy, or potentially allied ones. But how do you enforce contracts when your government and their government are at war? (The governments say “you shouldn’t; they’re the enemy.” Well, screw the governments, I want to make positive-sum trades in the hopes of increasing global GDP a…

If that’s what you want web3 for, I’m actually quite happy for you to record your treachery on a public ledger. There is a reason the government doesn’t want you to do it, and that’s that sanctions are one of the only reasonable deterrents to war that we have. I can’t believe you would advocate for interfering with that on a public forum, in 2021 when western democracies are already struggling to make economic sancti…

You’re assuming I’m talking about selling things to North Korea, or arms dealing, or something; but in the multilateral case, I’m actually talking about enabling trade in situations where arbitration of otherwise-legal contracts is impractical due to enforcement of contracts requiring ongoing positive relations (e.g. for mutual extradition) which have lapsed, or requiring the crossing of a DMZ — leaving parties on either side in a legal “state of nature”, free to screw one-another over with no consequences.

There were no economic sanctions imposed between East and West Germany; but making a contract with someone on the other side of the wall was still impractical, due to lack of diplomatic relations. Smart contracts solve that problem.

Other angles on this: the regular financial system in the Western world doesn’t let anyone transfer fiat currency to countries like Nigeria “for your protection”, even if you’re literally trying to run payroll for the registered employees of your Nigerian subsidiary company. (In other words, a de-facto trade embargo, but not a legal one, and not one to do with a any war.)

The only good way to do that payroll at this point is to send them crypto.

Re: Is Web3 anything?

#115
There is certainly some hope: for example Decentraland joining Blender as a sponsor equal to Aws, Facebook, or Nvidia. It's all about who controls the money

Re: Is Web3 anything?

#116
post #8

Earlier quoted context omitted.

The irony of course being that you would have to fully trust the programmer who reviews your contract.

Hmm, we already trust programmers enough to design our cryptography (hash functions and such)

With millions of dollars being stolen from various smart contracts every other week... No, I don't trust those programmers.

Designing hash/crypto functions is a very different story than implementing contracts. You would need to know basically 0 of the math that would be needed to implement a hash function.

Re: Is Web3 anything?

#117
post #47

Earlier quoted context omitted.

Ethereum has twice hard forked the blockchain to revert the results of "code is law", clearly saying that code is only law when a quorum of powerful individuals accept the outcome. Code isn't law, "code is law" is a slogan with as much veracity as slogans usually have.

Miners choose which chain to follow, not the quorum of powerful individuals. If you believed that ETC was the correct option in that decision, you were free to mine that chain all you want.

Miners choose where it's profitable to mine, which will always be the more active blockchain. It's called voting with your feet, and clearly Ethereum miners followed the quorum of powerful individuals. What matters your massive holdings of ETC when there's no buyers?

Re: Is Web3 anything?

#118
post #114

Earlier quoted context omitted.

If that’s what you want web3 for, I’m actually quite happy for you to record your treachery on a public ledger. There is a reason the government doesn’t want you to do it, and that’s that sanctions are one of the only reasonable deterrents to war that we have. I can’t believe you would advocate for interfering with that on a public forum, in 2021 when western democracies are already struggling to make economic sancti…

You’re assuming I’m talking about selling things to North Korea, or arms dealing, or something; but in the multilateral case, I’m actually talking about enabling trade in situations where arbitration of otherwise- legal contracts is impractical due to enforcement of contracts requiring ongoing positive relations (e.g. for mutual extradition) which have lapsed, or requiring the crossing of a DMZ — leaving parties on e…

I updated my comment to clarify. You did say enemy.

Re: Is Web3 anything?

#119
post #101

Earlier quoted context omitted.

Democracy solves the Byzantine agreement problem just fine already. I like the aspects of decentralisation that help you preserve democracy, but beyond that it’s not necessary to replace every single aspect of a functioning democracy with a cryptographic commitment scheme. It would be worse if you did. That vision of society sounds terrible. (I can’t comment on this without noting that of all the interpretations of t…

> Democracy solves the Byzantine agreement problem just fine already. That’s nice when you and your counterparty are part of the same democracy, or potentially allied ones. But how do you enforce contracts when your government and their government are at war? (The governments say “you shouldn’t; they’re the enemy.” Well, screw the governments, I want to make positive-sum trades in the hopes of increasing global GDP a…

In the thought experiment of a true anarchy, how do you ultimately enforce the off-chain part of a transaction? If BTC is traded for goods or services, how do you enforce that the goods and services have been sufficiently provided?

Re: Is Web3 anything?

#120
post #84

Earlier quoted context omitted.

If I understood Web3 correctly, it would be exactly like traditional web, except that: * The source code of all websites would be open-source, and accessible to anyone. It could still be changed, but it could not be done secretly. * All the actions of the users of the websites would be public. If we take social forums, for example Hacker News, it would mean that: * The algorithm behind the front page of social forums…

> The message would still be accessible on the blockchain. What about the right to be forgotten and GDPR laws?

That is a tough issue.

A blockchain is essentially an Event Sourced system. And Right To Be Forgotton is a common issue in all event sourced systems.

Several technical and one non technical solution are common:

Encrypt events with per actor keys. Delete the key when user wants to be forgotton. Or encrypt just certain attributes in the same way.

Keep actual data outside of events but only include a link to the data and checksum of the data.

Avoid storing PII in the first place. Does your domain model really require an IPaddress or First Name?

All have tradeoffs, none are silver bullets. But it is a solved problem in Event Sourcing, so it is a solved problem with blockchains (if only new programmers new this and reused patters instead of continually reinventing them)

Post reply on HN