Hacker News Reader

Comment by nightpool | Hacker News Reader

Ethereum has already proven that there's no way to prevent social forks, correct? The DAO incident comes to mind as a very controversial instance, but there are others. Ultimately, it's up to the individual nodes to decide what code they want to run and what rules they want to use to decide what the chain is. I'm not sure I understand how PoS makes this worse (although maybe there's something I'm missing)

View on HN

I found the following article to be helpful in understanding some of these processes

https://haseebq.com/ethereum-is-now-unforkable-thanks-to-def...

This article makes a good point about the fragility of DeFi, but it fails to demonstrate that the broader Ethereum community cares enough about DeFi to let it influence their decisions. It makes a general, hand-wavey gesture to other parts of ETH ("All websites, interfaces, block explorers, and wallets ultimately point to the majority chain. Game operators like CryptoKitties lock their D-ETH contracts so as not to confuse their users."), but it fails to justify this point—it already says "a civil war is brewing" and that individual developers (presumably including CryptoKitties) would have reason to support one or the other.

So while this article is a (good?) argument for why all DeFi operators must follow USDC, it fails to make the case for why anyone else should care about the fate of centralized stablecoins. Maybe these reasons are obvious! But I don't know them.

Right, but then the model is not trustless but is instead a complex implementation of the current system of international banking. Maybe I don't understand the point of cryptocurrencies. Most applications of cryptography are trustless.

"Trustless" is relative. In this case, the protocols are still trustless—the trust lies one meta level up, where you decide which protocols to use. In this situation, there are network effects as well—you can come up with your own perfect protocol in a vacuum, but if nobody uses it, who cares?

In this case, you can think about it as one person saying "I think it's a good idea if we do this" and a few hundred other people saying "I agree with you". Obviously, they're more likely to agree with the first person if they have some sort of institutional legitimacy. But is that "trust"? That's very different from the definition of "trust" that's generally used in protocol design.

Reply on news.ycombinator.com