4 ms·

> As of 2022 the oracle problem still hasn't been solved What do you mean by that? I thought ChainLink solved that pretty well

hamilyon2 4y ago
Read their white paper. Looks like secure computation with extra steps.
tick_tock_tick 4y ago
ChainLink just distributes the trust around making it harder too corrupt but in no way solves the issue.
- delaaxe 4y ago
  Trying to understand by going through this: https://blog.chain.link/what-is-the-blockchain-oracle-problem/ https://blog.chain.link/what-is-the-blockchain-oracle-proble... Aren't they precisely solving the issue?
  - mypastself 4y ago
    I came away from reading that article with almost no insight into how the problem is purportedly solved. The last section briefly lists a number of security techniques Chainlink uses, adding that these are “just some of [its] many features”, but doesn’t go into much detail. I’m not sure they ever even outright state that this particular network resolves the oracle problem. Are you sure that’s the article you wanted to share?
polygamous_bat 4y ago
Chainlink does not solve the oracle problem. In the best case, they put a "price" on fabricating truth in the blockchain world. If the "source of truth" is distributed, what stops somebody from running a worldwide botnet which pretends to be "anonymous chainlink nodes/contributors" and feed false information to the blockchain? Even if there is a price for each node to participate in the consensus mechanism, then you just set up a price which is required before you are allowed to defraud the oracle. How much are you willing to bet on this possibly defraudable oracle?
- c0mptonFP 4y ago
  Dude, defrauding Chainlink even once would probably cost you trillions of dollars - if not more. Have you read the super-linear staking section in the white paper?
  - mw888 4y ago
    You have the burden of proof, doubly so considering Chainlink has no working product which lives up to its promise. Not saying it’s impossible, but you shouting out a number and some terminology just makes it seem like you are emotionally invested in addition to financially.
    - c0mptonFP 4y ago
      First of all, I'm neither invested in ChainLink nor ETH. I'm interested in working on oracles. Secondly, this is an informal discussion on a mainstream forum - one which is pretty strongly against crypto. So forgive me for not being super keen on writing PhD-level responses The current white paper mentions that for a briber with at most `$d*(n^2 + n)/2`, where n is the number of oracle nodes, there exists a subgame perfect equilibrium for bribers to not issue the bribe and nodes to behave honestly. I can't seem to find a formal proof though. If you bother to check the listed examples in the white paper, my hand-wavey "trillions of dollars" estimate wouldn't be even that far off, assuming wider adoption of Chainlink.

hamilyon2 4y ago

Read their white paper. Looks like secure computation with extra steps.

tick_tock_tick 4y ago

ChainLink just distributes the trust around making it harder too corrupt but in no way solves the issue.

delaaxe 4y ago
Trying to understand by going through this: https://blog.chain.link/what-is-the-blockchain-oracle-problem/ https://blog.chain.link/what-is-the-blockchain-oracle-proble... Aren't they precisely solving the issue?
- mypastself 4y ago
  I came away from reading that article with almost no insight into how the problem is purportedly solved. The last section briefly lists a number of security techniques Chainlink uses, adding that these are “just some of [its] many features”, but doesn’t go into much detail. I’m not sure they ever even outright state that this particular network resolves the oracle problem. Are you sure that’s the article you wanted to share?

polygamous_bat 4y ago

Chainlink does not solve the oracle problem. In the best case, they put a "price" on fabricating truth in the blockchain world. If the "source of truth" is distributed, what stops somebody from running a worldwide botnet which pretends to be "anonymous chainlink nodes/contributors" and feed false information to the blockchain? Even if there is a price for each node to participate in the consensus mechanism, then you just set up a price which is required before you are allowed to defraud the oracle. How much are you willing to bet on this possibly defraudable oracle?

c0mptonFP 4y ago
Dude, defrauding Chainlink even once would probably cost you trillions of dollars - if not more. Have you read the super-linear staking section in the white paper?
- mw888 4y ago
  You have the burden of proof, doubly so considering Chainlink has no working product which lives up to its promise. Not saying it’s impossible, but you shouting out a number and some terminology just makes it seem like you are emotionally invested in addition to financially.
  - c0mptonFP 4y ago
    First of all, I'm neither invested in ChainLink nor ETH. I'm interested in working on oracles. Secondly, this is an informal discussion on a mainstream forum - one which is pretty strongly against crypto. So forgive me for not being super keen on writing PhD-level responses The current white paper mentions that for a briber with at most `$d*(n^2 + n)/2`, where n is the number of oracle nodes, there exists a subgame perfect equilibrium for bribers to not issue the bribe and nodes to behave honestly. I can't seem to find a formal proof though. If you bother to check the listed examples in the white paper, my hand-wavey "trillions of dollars" estimate wouldn't be even that far off, assuming wider adoption of Chainlink.