Comment by Barrera | Hacker News Reader

Comment by Barrera | Hacker News Reader

> Validators fulfill various functions in blockchains, including the creation of transaction blocks and the updating of data oracles. Ronin uses a so-called “proof of authority” system for signing transactions, concentrating power in the hands of nine trusted actors.

This paragraph perfectly encapsulates everything wrong with the way promoters sell Ethereum. Smart contracts can do little of interest beyond straight monetary transactions without information about the outside world. That information comes from "oracles", or what the article calls "validators".

The security guarantees of this system are far, far weaker than the Ethereum consensus protocol, as the article demonstrates. And yet, the system is hyped to the n-th degree by sheisters who ignore this basic fact with ludicrous claims about security and stability.

Zooming out, basically Ethereum is hyped as a platform for "smart contracts." But the minute a smart contract does anything beyond basic money transfers, it needs an oracle. And with the oracle comes radically reduced security.

Eventually, this will be obvious. For now, shenanigans like this will continue.

Oracles that connect to off-chain data are usually understood as points of centralization, I don’t think Ethereum or it’s developers are selling otherwise.

Most Ethereum developers are advising against relying on bridges across security zones that would be upheld by multisigs and oracles, they are vulnerable to attacks. A better model than a bridge to sidechain would be a rollup - posting proofs on chain without giving the sequencer the ability to steal or control user funds.

A better model than a bridge to sidechain would be a rollup - posting proofs on chain without giving the sequencer the ability to steal or control user funds.

What's a "proof" of an event in the outside world? No such thing exists.

The OP's point that dependence on real world events is a security flaw remains. No alternative programming method can change this because nothing on the Internet can guarantee a real event has happened. This is inherent.

How do you know that a real world event happened if you didn't perceive it with your own senses?

One of the long term goals of blockchains interfacing with the "real world" is to build oracle marketplaces that compete to provide "proof of true events". You can essentially build a reputation by selling valid data. The same way different news outlets and journalists compete for providing the "proof" for events.

https://chain.link/use-cases

In the context of ordinary contracts, the entire court and legal system exists to determine whether a given event happens.

Moreover, the legal system isn't intended as merely a means to determine whether a contract is fulfilled. It's also intended as a institution for preventing "cheating" and it has a concept of cheating. IE, me just satisfying a proxy for my end of a contact while not fulfilling the actual intent will not result in a judgement that the contract isn't fulfilled but also penalties of various severity if I intended to defraud a customer. If I build a bridge out of cardboard and, say, show pictures as proof of my supposed actions, I may well go to prisoner for fraud rather than prospering, even if it takes a while.

Smart contracts, in contrast, inherent can't distinguish between a proxy-value and a real situation - and their inability to do so is seen by their naive proponents as being their appeal. Equivalently, they have no concept of cheating. Thing considered "Hacks" and "fraud" have been prevent in etherium sure but by the opaque actions of governing boards, much less transparently than the ordinary courts.

And you can't make the problem go away with a reputation system since if the benefit lying gets higher than the cost of losing reputation anyone may lie with no other repercussions. And reputation always leaves the possibility of hacks to systems open - as the article shows.

The article in question relates to an incident with a bridge to a proof-of-authority sidechain that was controlled by a multisig. I have not looked deeply at Axie but it is possible that a rollup could replace their bridge and sidechain tech, and their need to "prove" the authority of validators that were upholding the side chain.

You can't prove anything from the outside world.

I still remember in 2017 shills on arkcoin slack trying to sell agricultural insurance that paid based on weather. They kept selling how decentralized it was but at the end of the day they just admitted that a central authority would input that data and we were back at square 0.

> This paragraph perfectly encapsulates everything wrong with the way promoters sell Ethereum... That information comes from "oracles", or what the article calls "validators".

Validators and Oracles are two different things entirely. That factual error aside, this is like saying that the US Federal Reserve is a weak system because Lehman Brothers and the MBS market failed.

Regardless, The Oracle Problem is already incredibly obvious and well-documented[0][1][2][3].

[0] - https://blog.chain.link/what-is-the-blockchain-oracle-proble...

[1] - https://encyclopedia.pub/entry/2959

[2] - https://cointelegraph.com/magazine/2021/12/30/can-blockchain...

[3] - https://www.proquest.com/docview/2524524192

Reply on news.ycombinator.com