Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The math prodigy whose hack upended DeFi won’t return funds

bloomberg.com

Post reply on HN
Print view View original

401–409 of 409 posts

Re: The math prodigy whose hack upended DeFi won’t return funds

#401

Post by Dylan16807 »

criddell wrote: post #398
Dylan16807 wrote: post #397

Earlier quoted context omitted.

No no you cut me off there. NFTs do it, and certain smart contracts do it. If you get picky about the latter part, then I retract it and just say NFTs for simplicity.

Ownership can’t be defined outside of the law. Ultimately these are boring old contracts in a digital form and they may or may not be valid. The validity can’t be determined within the NFT, smart contract, or blockchain. Some code may initiate a transfer of something and that gives you possession of it, but ownership isn’t guaranteed. For example, you can write a smart contract that would watch for divorce papers bei…

If it goes from a wallet you control to a different wallet you control, neither system would say ownership changed, so I'm not sure how that example helps.

Another point of comparison I'll bring up is cars, where if you let someone have your keys then you have limited or zero recourse in many situations where they do something bad.

Re: The math prodigy whose hack upended DeFi won’t return funds

#402

Post by beaugunderson »

xur17 wrote: post #400
beaugunderson wrote: post #396

Earlier quoted context omitted.

> It's not a guarantee though, the miner could decide to still frontrun The transaction bundle will fail if the success criteria is not reached (often a certain level of profit), so the worst that happens is that the profit margin falls to that level or the transaction is not included with zero cost to the sender

Is that something the mev apis implement? The Ethereum network would still include the transaction and just have it fail.

Exactly, flashbots and similar implement a gateway for submitting transactions directly to the miners. The transactions are only included if they succeed, because part of the transaction transfers payment to the miner for including it and another part of the transaction guarantees success only if it is profitable enough for the submitter. The transaction bundle is not mined unless it succeeds and thus avoids paying gas for failed attempts.

Re: The math prodigy whose hack upended DeFi won’t return funds

#403

Post by criddell »

Dylan16807 wrote: post #401
criddell wrote: post #398

Earlier quoted context omitted.

Ownership can’t be defined outside of the law. Ultimately these are boring old contracts in a digital form and they may or may not be valid. The validity can’t be determined within the NFT, smart contract, or blockchain. Some code may initiate a transfer of something and that gives you possession of it, but ownership isn’t guaranteed. For example, you can write a smart contract that would watch for divorce papers bei…

If it goes from a wallet you control to a different wallet you control, neither system would say ownership changed, so I'm not sure how that example helps. Another point of comparison I'll bring up is cars, where if you let someone have your keys then you have limited or zero recourse in many situations where they do something bad.

I think most of our disagreement is about the word ownership. I think many of the times you've used that word, it would be more accurate to say possession.

Your car example is a good one. If somebody has your keys (either with permission or stolen) and they take your car, they have possession of it but at no point do they own it.

Another example - lots of companies have crypto wallets and own Bitcoin and other currencies. Corporate wallets are controlled by agents of that company. The people who know the keys don't own the crypto, the company does. If an employee transfers from a company owned wallet that they control to a wallet they own, that's theft the same as if they had wired money to an overseas account of theirs.

Possession and ownership can align, but they don't have to. Smart contracts can be legally binding, but that isn't guaranteed. None of this exists outside of our legal system.

Re: The math prodigy whose hack upended DeFi won’t return funds

#404

Post by grumple »

postalrat wrote: post #384
grumple wrote: post #376

Earlier quoted context omitted.

I know how to pick locks. The locks are designed to open when pins are raised by a certain amount. Doesn't mean I have a right to anything I can unlock. The intention of the system is to work to as security. Just because you find a way around the security doesn't mean that you are now immune to the law regarding ownership. When SHAX gets cracked, that doesn't give the person figuring it out a right to transfer all th…

If private keys are proof of ownership then what's the difference between keys generated before vs keys generated after. If both keys are valid then both have then both are proof.

Private keys are not proof of ownership, they are a technical implementation detail. That crypto has been poorly designed so that keys are all people have to prove ownership is a failure of design but doesn't fundamentally change the divisibility of ownership and access. This is easily observed with a shared wallet, for example coinbase's. If an employee stole their keys, I assure you that employee would enjoy a jail yard view for the next lifetime.

Re: The math prodigy whose hack upended DeFi won’t return funds

#405

Post by Dylan16807 »

criddell wrote: post #403
Dylan16807 wrote: post #401

Earlier quoted context omitted.

If it goes from a wallet you control to a different wallet you control, neither system would say ownership changed, so I'm not sure how that example helps. Another point of comparison I'll bring up is cars, where if you let someone have your keys then you have limited or zero recourse in many situations where they do something bad.

I think most of our disagreement is about the word ownership . I think many of the times you've used that word, it would be more accurate to say possession . Your car example is a good one. If somebody has your keys (either with permission or stolen) and they take your car, they have possession of it but at no point do they own it. Another example - lots of companies have crypto wallets and own Bitcoin and other curr…

> I think most of our disagreement is about the word ownership. I think many of the times you've used that word, it would be more accurate to say possession.

But also I'm saying that a foundational pillar of NFTs is that possession on the blockchain is the only thing that matters.

It's possible to disagree with that pillar, but removing it would destroy a huge chunk of what NFTs are.

If you want me not to say 'ownership' for that concept, that's fine.

Also I'm not making this claim about crytocurrency in general. It's only specific subsets.

Re: The math prodigy whose hack upended DeFi won’t return funds

#406

Post by aidenn0 »

motohagiography wrote: post #292
aidenn0 wrote: post #276

Earlier quoted context omitted.

> If he has the money, fleeing before charges were laid was probably even rational, as there is no reason to expect the legal system is equipped to deliver justice in something so new. TFA claims he was originally offered to keep 10% (over a million dollars) from this hack, free and clear. Not agreeing to that deal meant willingly putting himself at the mercy of said legal system. Talking about a single decision as r…

This is like a casino offering to let you leave with 10% of your winnings because their dealer didn't shuffle the deck. Not an reasonable expectation. I will go back to this idea of an endogenous logic calculation vs. exogenous parsing errors description, where so long as he did not misrepresent the identities or sabotage the functions of any of the sources or destinations of the funds he used in his system, he shoul…

Probably nobody is reading this thread anymore, but I remembered an analogous situation in a casino with a slot machine (it credited $1 inserted as $10) and the people who kept playing it for that reason were potentially liable for fraud.

https://www.techdirt.com/2007/07/24/criminal-charges-for-usi...

Re: The math prodigy whose hack upended DeFi won’t return funds

#407

Post by xur17 »

beaugunderson wrote: post #402
xur17 wrote: post #400

Earlier quoted context omitted.

Is that something the mev apis implement? The Ethereum network would still include the transaction and just have it fail.

Exactly, flashbots and similar implement a gateway for submitting transactions directly to the miners. The transactions are only included if they succeed, because part of the transaction transfers payment to the miner for including it and another part of the transaction guarantees success only if it is profitable enough for the submitter. The transaction bundle is not mined unless it succeeds and thus avoids paying g…

My point is that this is something the flashbot service does before submitting the transaction. There is nothing technical stopping them from submitting it or frontrunning it (besides destroying their reputation). In addition, when uncles and reorgs occur, the rest of the network sees the transaction, and hence could frontrun it / submit it and have it fail.

Re: The math prodigy whose hack upended DeFi won’t return funds

#408

Post by Handytinge »

zecken wrote: post #123

I feel like the fact this person, per the article, is a white supremacist who used the n-word in his code repeatedly is under-discussed here. Folks here jumping through hoops to rationalize why what he did is actually demonstrative of either flaws in crypto or the purity of arbitrage come off seeming very tone deaf.

Someone can be a white supremecist and you can be impressed with their (non bigotry) actions.

People aren't that black and white (sorry).

Re: The math prodigy whose hack upended DeFi won’t return funds

#409

Post by Handytinge »

DangitBobby wrote: post #36
trasz wrote: post #21

I find it disturbing that Medjedovic was prosecuted in the first place. If anyone is guilty of this situation, it's Kellar and Day.

Absolutely. They are guilty of writing vulnerable code, being hacked, and stolen from.

If _they_ are not treating it as seriously as actual regulated finance, why should the courts be treating the attacker in that way?
Post reply on HN