I’m sorry, but you’re describing an open-source, decentralised, peer-to-peer, permissionless digital payment network. Which is exactly what cryptocurrency is. But I think you know that if you openly advocate in favour of cryptocurrency here, you instantly get downvotes on Lemmy. So you’re doing it in a fairly roundabout way.
I don’t know where you get the idea that blockchains are no longer proof-of-work. Bitcoin is still the largest cryptocurrency and it’s still using proof-of-work. It’s not really what I’m getting at though, when I say that a decentralised system is 1000x the people each doing 100x the work. Even a proof-of-stake system will still have a lot of work that each node has to do, validation transactions, and that amount of data that has to be passed around serves as a ceiling on transaction capacity. Bitcoin is notoriously low at 1 (“virtual”) MB every 10 minutes. But even larger limits or Ethereum’s sharding strategy would be utterly overwhelmed by the transaction volumes of traditional finance.
To clarify for others bitcoins transaction limit is a choice. Bitcoin blocks originally had no limit. One was put in place to limit spam. The restriction could be lifted now but doing so would increase the storage and bandwidth requirements for nodes.
I want to do a sanity check here. The six largest payment cards networks (Visa, UnionPay, Mastercard, American Express, JCB, and Discover) processed about 770 billion transactions in 2024 (source). That’s two years ago, and certainly as more of the world transitions to cashless payments that number has probably increased, but let’s just use the 770 billion number for the sake of calculation.
Bitcoin transactions are actually fairly inefficient in terms of transaction size because of their UXTO-based coin system. An account-based system like Ethereum can result in smaller transaction sizes. Let’s take the minimum transaction data that one would need to store for a payments-only network:
The sending account number, which is usually 128 bits
The amount, let’s say it’s a unit64_t so 64 bits
The receiving account number, another 128 bits
A digital signature generated by the sender. Let’s use Schnorr signatures, which are relatively short. A 512-bit signature will provide 128 bits of security.
We’ll ignore Segwit since witness data still needs to be stored, that’s just Bitcoin’s cheeky way of expanding the block size without explicitly declaring a larger block size.
To allow for indexing, let’s allow some kind of tree structure and say it has overhead of 32 bytes per node (it will probably require more, but let’s just roll with this for now).
Total: 136 bytes per tx
Multiplied by 770 billion gives 104.72 PB. Even if you had a block every 10 seconds like Ethereum does, the block size needed to process that kind of volume would be 33.2 MB.
Storing a blockchain that grows by over a hundred petabytes a year is impossible for all but the most well-funded organisations. That’s two orders of magnitude out of reach for the largest non-crypto open-source projects (the Wikimedia Foundation), four orders of magnitude out of reach for your average open-source project, five orders of magnitude out of reach for ordinary FOSS enthusiasts, and eight orders of magnitude out of reach of everyday users.
Blockchain is a cool technology and I grant that Satoshi Nakamoto was a pretty smart guy and a brilliant computer scientist, but it’s just not suitable for handling the types of volume we deal with in the modern financial system.
I’m sorry, but you’re describing an open-source, decentralised, peer-to-peer, permissionless digital payment network. Which is exactly what cryptocurrency is. But I think you know that if you openly advocate in favour of cryptocurrency here, you instantly get downvotes on Lemmy. So you’re doing it in a fairly roundabout way.
I don’t know where you get the idea that blockchains are no longer proof-of-work. Bitcoin is still the largest cryptocurrency and it’s still using proof-of-work. It’s not really what I’m getting at though, when I say that a decentralised system is 1000x the people each doing 100x the work. Even a proof-of-stake system will still have a lot of work that each node has to do, validation transactions, and that amount of data that has to be passed around serves as a ceiling on transaction capacity. Bitcoin is notoriously low at 1 (“virtual”) MB every 10 minutes. But even larger limits or Ethereum’s sharding strategy would be utterly overwhelmed by the transaction volumes of traditional finance.
To clarify for others bitcoins transaction limit is a choice. Bitcoin blocks originally had no limit. One was put in place to limit spam. The restriction could be lifted now but doing so would increase the storage and bandwidth requirements for nodes.
I want to do a sanity check here. The six largest payment cards networks (Visa, UnionPay, Mastercard, American Express, JCB, and Discover) processed about 770 billion transactions in 2024 (source). That’s two years ago, and certainly as more of the world transitions to cashless payments that number has probably increased, but let’s just use the 770 billion number for the sake of calculation.
Bitcoin transactions are actually fairly inefficient in terms of transaction size because of their UXTO-based coin system. An account-based system like Ethereum can result in smaller transaction sizes. Let’s take the minimum transaction data that one would need to store for a payments-only network:
unit64_tso 64 bitsWe’ll ignore Segwit since witness data still needs to be stored, that’s just Bitcoin’s cheeky way of expanding the block size without explicitly declaring a larger block size.
To allow for indexing, let’s allow some kind of tree structure and say it has overhead of 32 bytes per node (it will probably require more, but let’s just roll with this for now).
Total: 136 bytes per tx
Multiplied by 770 billion gives 104.72 PB. Even if you had a block every 10 seconds like Ethereum does, the block size needed to process that kind of volume would be 33.2 MB.
Storing a blockchain that grows by over a hundred petabytes a year is impossible for all but the most well-funded organisations. That’s two orders of magnitude out of reach for the largest non-crypto open-source projects (the Wikimedia Foundation), four orders of magnitude out of reach for your average open-source project, five orders of magnitude out of reach for ordinary FOSS enthusiasts, and eight orders of magnitude out of reach of everyday users.
Blockchain is a cool technology and I grant that Satoshi Nakamoto was a pretty smart guy and a brilliant computer scientist, but it’s just not suitable for handling the types of volume we deal with in the modern financial system.