5 ms·

Just going to point out that Tornado cash did invoke chainalysis oracles for all of their interface tools. The problem is that on ethereum it's not possible to

Just going to point out that Tornado cash did invoke chainalysis oracles for all of their interface tools. The problem is that on ethereum it's not possible to censor a deployed contract. There is a possibility that miners could collude not to authenticate blocks with tornado cash transactions in them, but that gets into some interesting game theory in a globally distributed system. Not every miner is subject to US law. It's a very fascinating story. https://www.coindesk.com/tech/2022/04/15/tornado-cash-adds-chainalysis-tool-for-blocking-ofac-sanctioned-wallets-from-dapp/ https://www.coindesk.com/tech/2022/04/15/tornado-cash-adds-c...

hef19898 4y ago
Sanctions against North Korea are not just US law, in fact they apply in basically every jurisdiction that actually worth operating in.
qeternity 4y ago
The onus is for technology to comply with laws, not the other way around.
- dontknowwhyihn 4y ago
  Tell that to Uber.
- everfree 4y ago
  Laws bind people, and people should comply with laws. Technology is not a person. People who make technology should not be expected to add measures to it to make sure that nobody is ever able to use it to commit a crime. This kind of thinking would have resulted in guns being banned until technology exists for the gun itself to detect who was firing it and what it was being fired at, to prevent any crimes from occurring.
  - lern_too_spel 4y ago
    If the people writing the laws wish it, then yes, technologies should be expected to do so. In practice, your gun detector wouldn't solve all gun crimes, just like crypto regulations don't stop all crypto crimes and littering regulations don't stop all littering crimes. They just reduce those crimes to the point that society seems reasonable.
  - waste_monk 4y ago
    >People who make technology should not be expected to add measures to it to make sure that nobody is ever able to use it to commit a crime Why not? It's quite common, e.g. a lot of photocopiers have safeguards to prevent currency counterfeiting and will refuse to copy bank notes containing the EURion constellation and other common bank note markers. Pretty much any decent financial software will have sane defaults for audit trails, separation of duties, and so on to try and prevent fraud. Etc. If you deliberately build something that can facilitate money laundering and ignore finance laws about KYC and mandatory reporting and and so on, you only have yourself to blame if you run into legal trouble. It doesn't matter if you don't like the law or think it shouldn't apply. >This kind of thinking would have resulted in guns being banned until technology exists for the gun itself to detect who was firing it and what it was being fired at, to prevent any crimes from occurring You'd have to ban all blades, any kind of explosives (mining), any kind of vehicle, etc. ... a creative person can misuse just about any technology to commit a crime. At some point it is simply not possible to build safeguards. How would you put access controls on a sharp rock? I don't want to open the whole gun control can of worms, but down here in Australia we have strict firearm control and our firearm-related death rate is very low (0.92/100k population, vs. 10.95/100k for the USA [1]). It still happens of course, but the rate is low enough that strict technological controls wouldn't add much value. And I believe, though can't find the source I remember seeing a while ago, that a pretty big chunk of those are suicides / accidents and there is very low violent crime. [1] https://worldpopulationreview.com/country-rankings/gun-deaths-by-country https://worldpopulationreview.com/country-rankings/gun-death...
    - everfree 4y ago
      > If you deliberately build something that can facilitate money laundering and ignore finance laws about KYC and mandatory reporting and and so on, you only have yourself to blame if you run into legal trouble. It doesn't matter if you don't like the law or think it should apply. Do you believe that code is protected speech? Do you believe that people should be prosecuted for publishing source code that has the potential to be used maliciously if compiled and run?
      - waste_monk 4y ago
bornfreddy 4y ago
Also, with PoS coming in a few days, the game changes. IIUC, stakers will be obligated to validate all (valid) transactions unless they are willing to forfeit their staked ETH... No preferential treatment anymore. (This is still new to me, please correct me if I'm wrong)
- intrasight 4y ago
  If the vast majority of stakers agree to OFAC regulations (which appears to be the case) then the end result is censorship.
  - everfree 4y ago
    What gives the appearance that this is the case? Even if the "vast majority" of stakers agree with the regulation, the regulation is ambiguous as to whether stakers are expected to refrain from including TC transactions in their own blocks, or actively orphan all blocks that include TC transactions. The latter hurts their staking revenue and effectuates a soft fork. (Staking revenue is hurt due to the inactivity correlation factor that the network uses to calculate rewards.) If the regulation only demands the former, then the network will continue status-quo, except that TC transactions may take a couple minutes to be included instead of 12 seconds.
    - intrasight 4y ago
      It'll all become clear soon enough.
joyfylbanana 4y ago
> The problem is that on ethereum it's not possible to censor a deployed contract. It is possible, via a hard work. Similarly as it was possible to transition from PoW to PoS. The question is, if there is political will for that. Clearly changing the protocol rules is possible has it has been done with ethereum in the past, including for censorship reasons (in the earlier hard fork the reason was to cancel a hack).
- toomuchtodo 4y ago
  It is delightfully ironic to see code as law slowly chipped away by Law until law is code and you’re back to finance business logic on a slow distributed computer fabric.
  - everfree 4y ago
    It's too early to say that definitively, imo. As a counterexample to the DAO hack, the parity multi-sig hack of 2017 resulted in over $160M worth of ether being frozen on-chain. There were calls to hard fork Ethereum to return it, but the hard fork was never tenable. That was barely a year after the DAO hack and fork. Personally, I believe there is a zero chance that a hard fork based on Tornado Cash becomes viable. It's not nearly a big enough issue for enough users to care (and exchanges, and wallet software companies, and stakers). Why would the 90% of users who don't use Tornado Cash risk helping the other 10% perform some fork, knowing that the fork would add fuel to the mutability argument and set more precedent for mutability?
    - josh2600 4y ago
      A hard fork based on tornado cash isn't likely, but what if miners refuse to process blocks that have tornado transactions in them? What happens then?
      - everfree 4y ago
        Not much. https://news.ycombinator.com/item?id=32814082 https://news.ycombinator.com/item?id=32814082

hef19898 4y ago

Sanctions against North Korea are not just US law, in fact they apply in basically every jurisdiction that actually worth operating in.

qeternity 4y ago

The onus is for technology to comply with laws, not the other way around.

dontknowwhyihn 4y ago
Tell that to Uber.
everfree 4y ago
Laws bind people, and people should comply with laws. Technology is not a person. People who make technology should not be expected to add measures to it to make sure that nobody is ever able to use it to commit a crime. This kind of thinking would have resulted in guns being banned until technology exists for the gun itself to detect who was firing it and what it was being fired at, to prevent any crimes from occurring.
- lern_too_spel 4y ago
  If the people writing the laws wish it, then yes, technologies should be expected to do so. In practice, your gun detector wouldn't solve all gun crimes, just like crypto regulations don't stop all crypto crimes and littering regulations don't stop all littering crimes. They just reduce those crimes to the point that society seems reasonable.
- waste_monk 4y ago
  >People who make technology should not be expected to add measures to it to make sure that nobody is ever able to use it to commit a crime Why not? It's quite common, e.g. a lot of photocopiers have safeguards to prevent currency counterfeiting and will refuse to copy bank notes containing the EURion constellation and other common bank note markers. Pretty much any decent financial software will have sane defaults for audit trails, separation of duties, and so on to try and prevent fraud. Etc. If you deliberately build something that can facilitate money laundering and ignore finance laws about KYC and mandatory reporting and and so on, you only have yourself to blame if you run into legal trouble. It doesn't matter if you don't like the law or think it shouldn't apply. >This kind of thinking would have resulted in guns being banned until technology exists for the gun itself to detect who was firing it and what it was being fired at, to prevent any crimes from occurring You'd have to ban all blades, any kind of explosives (mining), any kind of vehicle, etc. ... a creative person can misuse just about any technology to commit a crime. At some point it is simply not possible to build safeguards. How would you put access controls on a sharp rock? I don't want to open the whole gun control can of worms, but down here in Australia we have strict firearm control and our firearm-related death rate is very low (0.92/100k population, vs. 10.95/100k for the USA [1]). It still happens of course, but the rate is low enough that strict technological controls wouldn't add much value. And I believe, though can't find the source I remember seeing a while ago, that a pretty big chunk of those are suicides / accidents and there is very low violent crime. [1] https://worldpopulationreview.com/country-rankings/gun-deaths-by-country https://worldpopulationreview.com/country-rankings/gun-death...
  - everfree 4y ago
    > If you deliberately build something that can facilitate money laundering and ignore finance laws about KYC and mandatory reporting and and so on, you only have yourself to blame if you run into legal trouble. It doesn't matter if you don't like the law or think it should apply. Do you believe that code is protected speech? Do you believe that people should be prosecuted for publishing source code that has the potential to be used maliciously if compiled and run?
    - waste_monk 4y ago

bornfreddy 4y ago

Also, with PoS coming in a few days, the game changes. IIUC, stakers will be obligated to validate all (valid) transactions unless they are willing to forfeit their staked ETH... No preferential treatment anymore. (This is still new to me, please correct me if I'm wrong)

intrasight 4y ago
If the vast majority of stakers agree to OFAC regulations (which appears to be the case) then the end result is censorship.
- everfree 4y ago
  What gives the appearance that this is the case? Even if the "vast majority" of stakers agree with the regulation, the regulation is ambiguous as to whether stakers are expected to refrain from including TC transactions in their own blocks, or actively orphan all blocks that include TC transactions. The latter hurts their staking revenue and effectuates a soft fork. (Staking revenue is hurt due to the inactivity correlation factor that the network uses to calculate rewards.) If the regulation only demands the former, then the network will continue status-quo, except that TC transactions may take a couple minutes to be included instead of 12 seconds.
  - intrasight 4y ago
    It'll all become clear soon enough.

joyfylbanana 4y ago

> The problem is that on ethereum it's not possible to censor a deployed contract. It is possible, via a hard work. Similarly as it was possible to transition from PoW to PoS. The question is, if there is political will for that. Clearly changing the protocol rules is possible has it has been done with ethereum in the past, including for censorship reasons (in the earlier hard fork the reason was to cancel a hack).

toomuchtodo 4y ago
It is delightfully ironic to see code as law slowly chipped away by Law until law is code and you’re back to finance business logic on a slow distributed computer fabric.
- everfree 4y ago
  It's too early to say that definitively, imo. As a counterexample to the DAO hack, the parity multi-sig hack of 2017 resulted in over $160M worth of ether being frozen on-chain. There were calls to hard fork Ethereum to return it, but the hard fork was never tenable. That was barely a year after the DAO hack and fork. Personally, I believe there is a zero chance that a hard fork based on Tornado Cash becomes viable. It's not nearly a big enough issue for enough users to care (and exchanges, and wallet software companies, and stakers). Why would the 90% of users who don't use Tornado Cash risk helping the other 10% perform some fork, knowing that the fork would add fuel to the mutability argument and set more precedent for mutability?
  - josh2600 4y ago
    A hard fork based on tornado cash isn't likely, but what if miners refuse to process blocks that have tornado transactions in them? What happens then?
    - everfree 4y ago
      Not much. https://news.ycombinator.com/item?id=32814082 https://news.ycombinator.com/item?id=32814082