4 ms·

Categorically false. Exactly zero energy is required to produce a block. If your statement were true and bitcoin required Switzerland’s worth of energy to produ

Categorically false. Exactly zero energy is required to produce a block. If your statement were true and bitcoin required Switzerland’s worth of energy to produce a block - how on earth did satoshi mine all the early blocks by himself? Your understanding of how it all works is simply flawed. Energy is required to satisfy difficulty requirement, which has to do with chain security, not with transaction processing.

MereInterest 5y ago
Got it, so you're separating out the energy requirements into two buckets, then pretending that those two are unrelated. I agree that in general, processing transactions is not an energy-intensive task. This is why traditional payment processors with trusted nodes are able to do so efficiently. In the case of cryptocurrency, where the assumption is that all nodes are inherently untrustworthy, the resource requirement is stapled onto the transaction processing. When people say "transactions cost X amount to process", they mean "in the current state of bitcoin", not in the past when there were very few transaction validators. The more transaction validators (typically called "miners") there are, the higher the cost of validating a transaction, due to Bitcoin's difficulty scaling. I'm not sure why you separate out chain security from transaction validation and treat them as separate, when the chain security is there to protect against invalid transactions. It's sort of the entire point of it.
- keymone 5y ago
  > When people say "transactions cost X amount to process", they mean "in the current state of bitcoin", not in the past when there were very few transaction validators there are only so many miners (and so much energy spent) because security of bitcoin is what gives it value. so when people say "transactions cost X amount to process" they are saying something non-sensical, because it's not transactions that cost, it's security of those transactions. it's important to separate, because there are many kinds of transactions in bitcoin ecosystem with various security characteristics that land anywhere from very low to on-chain settlement with multiple confirmations.
  - axlee 5y ago
    > because it's not transactions that cost, it's security of those transactions. Logically wrong, but let's follow the reasoning. Let's assume Bitcoin transactions were indeed not secure (and thus, according to you, costing little energy). - Would these unsafe-but-cheap transactions still be considered "Bitcoin transactions" in the way that we know them? - Would these unsafe-but-cheap transactions fulfill the requirements that Bitcoin has in place for itself? - Would an unsafe-but-cheap Bitcoin still be a great option in your eyes to replace the worldwide financial system? If you agree with all these statements and their corollaries, I can follow you in saying "Bitcoin transactions do not cost energy". You are creating a false dichotomy: security is an integral component of bitcoin transactions, and if security costs, transactions cost as well. You cannot just summon it away and pretend it is unrelated to the nature of Bitcoin transactions itself, when in fact it is one of its largest selling points.
    - keymone 5y ago
      > You are creating a false dichotomy: security is an integral component of bitcoin transactions, and if security costs, transactions cost as well that would be true if only on-chain settlements would count as transactions in BTC ecosystem. there are more ways to exchange BTC with security landing anywhere between very low and on-chain settlement. thing is, that it's not even transactions that have security, it's blocks. energy is spent to make sure that it's expensive to override/rollback blocks. and it's true that blocks exist to assert ordering of on-chain settlements, but it's not true to claim that there is 1-1 correspondence between on-chain settlements and transactions. each on-chain settlement can embed trillions of bitcoin transactions (in colloquial sense, exchanges of an asset). what does that do to your "energy per tx" metric?
  - afiori 5y ago
    this is pretty much the same thing as saying that "guns kills people" is nonsensical because there are many non equivalent ways a gun could be involved in killing someone
    - keymone 5y ago

MereInterest 5y ago

Got it, so you're separating out the energy requirements into two buckets, then pretending that those two are unrelated. I agree that in general, processing transactions is not an energy-intensive task. This is why traditional payment processors with trusted nodes are able to do so efficiently. In the case of cryptocurrency, where the assumption is that all nodes are inherently untrustworthy, the resource requirement is stapled onto the transaction processing. When people say "transactions cost X amount to process", they mean "in the current state of bitcoin", not in the past when there were very few transaction validators. The more transaction validators (typically called "miners") there are, the higher the cost of validating a transaction, due to Bitcoin's difficulty scaling. I'm not sure why you separate out chain security from transaction validation and treat them as separate, when the chain security is there to protect against invalid transactions. It's sort of the entire point of it.

keymone 5y ago
> When people say "transactions cost X amount to process", they mean "in the current state of bitcoin", not in the past when there were very few transaction validators there are only so many miners (and so much energy spent) because security of bitcoin is what gives it value. so when people say "transactions cost X amount to process" they are saying something non-sensical, because it's not transactions that cost, it's security of those transactions. it's important to separate, because there are many kinds of transactions in bitcoin ecosystem with various security characteristics that land anywhere from very low to on-chain settlement with multiple confirmations.
- axlee 5y ago
  > because it's not transactions that cost, it's security of those transactions. Logically wrong, but let's follow the reasoning. Let's assume Bitcoin transactions were indeed not secure (and thus, according to you, costing little energy). - Would these unsafe-but-cheap transactions still be considered "Bitcoin transactions" in the way that we know them? - Would these unsafe-but-cheap transactions fulfill the requirements that Bitcoin has in place for itself? - Would an unsafe-but-cheap Bitcoin still be a great option in your eyes to replace the worldwide financial system? If you agree with all these statements and their corollaries, I can follow you in saying "Bitcoin transactions do not cost energy". You are creating a false dichotomy: security is an integral component of bitcoin transactions, and if security costs, transactions cost as well. You cannot just summon it away and pretend it is unrelated to the nature of Bitcoin transactions itself, when in fact it is one of its largest selling points.
  - keymone 5y ago
    > You are creating a false dichotomy: security is an integral component of bitcoin transactions, and if security costs, transactions cost as well that would be true if only on-chain settlements would count as transactions in BTC ecosystem. there are more ways to exchange BTC with security landing anywhere between very low and on-chain settlement. thing is, that it's not even transactions that have security, it's blocks. energy is spent to make sure that it's expensive to override/rollback blocks. and it's true that blocks exist to assert ordering of on-chain settlements, but it's not true to claim that there is 1-1 correspondence between on-chain settlements and transactions. each on-chain settlement can embed trillions of bitcoin transactions (in colloquial sense, exchanges of an asset). what does that do to your "energy per tx" metric?
- afiori 5y ago
  this is pretty much the same thing as saying that "guns kills people" is nonsensical because there are many non equivalent ways a gun could be involved in killing someone
  - keymone 5y ago