4 ms·

I'm not sure exactly why the US government thinks Bitcoin is not a security, but I think it's reasonable to speculate that the resistance to making incompatible

by Vecr 3y ago

I'm not sure exactly why the US government thinks Bitcoin is not a security, but I think it's reasonable to speculate that the resistance to making incompatible changes, and hence the probable agreement between the government (the prosecutor) and the defendant to run compatible full nodes may have something to do with it. Etherium is different, more willing to change, has a central personality with more control, and is frankly harder to set up full nodes and independently verify. Imagine the prosecution explaining to the judge what a "beacon chain" is, and the defense showing command line output on the 4K TVs they brought in, trying to explain why it's been two weeks and their nodes is not synced yet.

hakfoo 3y ago
I wonder if the "central personality" aspect is the key. Even the promise of massive wealth and clout has failed to unmask Satoshi, so it's largely seen as a self-managing system. There's no real figurehead you can say is responsible for it or able to steer it for personal gain at this point.
- latchkey 3y ago
  Matt Levine talks about this in one of his pieces recently. https://www.bloomberg.com/opinion/articles/2023-06-07/when-is-a-token-not-a-security https://www.bloomberg.com/opinion/articles/2023-06-07/when-i... "Some of them did securities offerings, but by the time the SEC noticed they were too entrenched and decentralized and it would have been a pain for the SEC to go after them. Ethereum, most notably, very very clearly did an ICO in 2014, raising about $18.3 million by selling ETH tokens. If they did that today, or in late 2017, the SEC would have some serious questions. But by the time the SEC got around to cracking down on ICOs in 2017, Ethereum was big and decentralized and the SEC would have had a hard time, practically and legally, challenging its 2014 ICO. And so everyone sort of grudgingly concedes that ETH is not a security."
jlokier 3y ago
> trying to explain why it's been two weeks and their nodes is not synced yet. Fwiw, it's possible to sync to Ethereum in a few hours now. The features to enable that were added in 2021, before the merge. If you want to sync an archive node instead, which contains all historical transactions verified from the beginning, that takes about a day and a half, but current work will bring that down to a few hours as well. It used to take nearly 2 months, and the state was smaller then, so this is an interesting technical achievement because it's just the same data represented more efficiently.
- analognoise 3y ago
  >It used to take nearly 2 months >so this is an interesting technical achievement When you consider how much work a $35 Raspberry Pi 4 could do in 2 months, is it still an interesting technical achievement, or is it an embarrassment?
  - jlokier 3y ago
    It used to take nearly 2 months on computers much faster than a Raspberry Pi 4, so yes it's an achievement. It's not insanely complex or anything, but there have been clever technical innovations in the the storage structures (how the database works) and in the network protocols designed to be sympathetic to the storage layout, that have sped it up a lot. Solving these was not easy, though seeing how they work is obvious in retrospect if you understand how storage and networking performance behave generally, which most people don't. The syncing bottlenecks were mostly storage performance bound, not CPU bound, so the Raspberry Pi thing is not as relevant as it may look to someone who doesn't know the difference.
    - analognoise 3y ago
      Ah, right, storage performance bound. The Etherium blockchain is roughly 1122.97 GB. Cool, how long would it take me to write all that data with an RPI4? Let's pick the ~20th fastest SSD listed, just so we ensure we don't have a ringer in the data: https://pibenchmarks.com/benchmark/53760/ https://pibenchmarks.com/benchmark/53760/ Awesome, let's do some division. Let's pick the DD speed: 1122.97GB/174 MB/s = 1.79273627 hours. Hm, that can't be it. The RPI4 could write the entire Eth blockchain in under 2 hours. Hmm. Well, maybe there's lots of data being added in a torrent at the end? That could be a real problem? https://ycharts.com/indicators/ethereum_chain_full_sync_data_size https://ycharts.com/indicators/ethereum_chain_full_sync_data... Looks like it's adding about 3GB a day. A rip roaring rate of 34.7222222 kBps. Ok with the data in hand, this looks like a pathetic system and it doesn't look like a technical achievement. It looks like an embarrassment to me. Could you check my numbers?

by Vecr 3y ago

hakfoo 3y ago

I wonder if the "central personality" aspect is the key. Even the promise of massive wealth and clout has failed to unmask Satoshi, so it's largely seen as a self-managing system. There's no real figurehead you can say is responsible for it or able to steer it for personal gain at this point.

latchkey 3y ago
Matt Levine talks about this in one of his pieces recently. https://www.bloomberg.com/opinion/articles/2023-06-07/when-is-a-token-not-a-security https://www.bloomberg.com/opinion/articles/2023-06-07/when-i... "Some of them did securities offerings, but by the time the SEC noticed they were too entrenched and decentralized and it would have been a pain for the SEC to go after them. Ethereum, most notably, very very clearly did an ICO in 2014, raising about $18.3 million by selling ETH tokens. If they did that today, or in late 2017, the SEC would have some serious questions. But by the time the SEC got around to cracking down on ICOs in 2017, Ethereum was big and decentralized and the SEC would have had a hard time, practically and legally, challenging its 2014 ICO. And so everyone sort of grudgingly concedes that ETH is not a security."

jlokier 3y ago

> trying to explain why it's been two weeks and their nodes is not synced yet. Fwiw, it's possible to sync to Ethereum in a few hours now. The features to enable that were added in 2021, before the merge. If you want to sync an archive node instead, which contains all historical transactions verified from the beginning, that takes about a day and a half, but current work will bring that down to a few hours as well. It used to take nearly 2 months, and the state was smaller then, so this is an interesting technical achievement because it's just the same data represented more efficiently.

analognoise 3y ago
>It used to take nearly 2 months >so this is an interesting technical achievement When you consider how much work a $35 Raspberry Pi 4 could do in 2 months, is it still an interesting technical achievement, or is it an embarrassment?
- jlokier 3y ago
  It used to take nearly 2 months on computers much faster than a Raspberry Pi 4, so yes it's an achievement. It's not insanely complex or anything, but there have been clever technical innovations in the the storage structures (how the database works) and in the network protocols designed to be sympathetic to the storage layout, that have sped it up a lot. Solving these was not easy, though seeing how they work is obvious in retrospect if you understand how storage and networking performance behave generally, which most people don't. The syncing bottlenecks were mostly storage performance bound, not CPU bound, so the Raspberry Pi thing is not as relevant as it may look to someone who doesn't know the difference.
  - analognoise 3y ago
    Ah, right, storage performance bound. The Etherium blockchain is roughly 1122.97 GB. Cool, how long would it take me to write all that data with an RPI4? Let's pick the ~20th fastest SSD listed, just so we ensure we don't have a ringer in the data: https://pibenchmarks.com/benchmark/53760/ https://pibenchmarks.com/benchmark/53760/ Awesome, let's do some division. Let's pick the DD speed: 1122.97GB/174 MB/s = 1.79273627 hours. Hm, that can't be it. The RPI4 could write the entire Eth blockchain in under 2 hours. Hmm. Well, maybe there's lots of data being added in a torrent at the end? That could be a real problem? https://ycharts.com/indicators/ethereum_chain_full_sync_data_size https://ycharts.com/indicators/ethereum_chain_full_sync_data... Looks like it's adding about 3GB a day. A rip roaring rate of 34.7222222 kBps. Ok with the data in hand, this looks like a pathetic system and it doesn't look like a technical achievement. It looks like an embarrassment to me. Could you check my numbers?