Re:Coin Carbon Cap – PoW cryptocurrencies ranked by energy e...
by nivexous on Sunday May 16, 2021 @07:57PM
(#27178479)
Almost all IP traffic goes through an Internet backbone router which can process 100Gbps+. It's no problem. I suggest people think of Bitcoin miners like this in the future. These would be big operations, yes, but they are economically incentivized to honesty, and there will be plenty of competition. Once you use your imagination, or try to build a node yourself, you'll see the only real bottleneck in Bitcoin's architecture is the updating of the UTXO set, and this can be made extremely fast - eventually in hardware. All other transaction and block validation and propagation can be parallelized or removed as a bottleneck. Users share transactions p2p and settle them onchain, and apps and services create sub-networks of transactions they care about. It all works. But to do that people need to drop the assumption that home users need to run a node - it's a broken idea that holds Bitcoin back. The original design of Bitcoin was remarkable in its ability to scale, which was written about by Satoshi and ignored, and this is being demonstrated again in Bitcoin SV. On the other hand, almost every other cryptocurrency has deviated from that simple design. Ethereum for example requires a global state database and cannot parallelize this way.
Re:Coin Carbon Cap – PoW cryptocurrencies ranked by energy e...
by ntr-- on Sunday May 16, 2021 @08:14PM
(#27178609)
"Trusted backbone transaction processors" that don't use megawatts of power to do this work have existed for some time as "banks".
Re:Coin Carbon Cap – PoW cryptocurrencies ranked by energy e...
by nivexous on Sunday May 16, 2021 @08:21PM
(#27178661)
No. In this design users don't trust miners. Users trust Bitcoin. This is important. Banks can censor. Banks can go down. Banks can be forced to confiscate your money by a single court. But if one miner were to act up, other miners around the globe could keep them in check, and apps and users could switch around as desired. It's an incentivized complete system that you can't reduce this way.
Re:Coin Carbon Cap – PoW cryptocurrencies ranked by energy e...
by rawtxapp on Sunday May 16, 2021 @07:43PM
(#27178369)
A single transaction on the blockchain most likely corresponds to thousands if not millions of transactions in another layer.
Unless you buy straight from a miner, the Bitcoin you buy changed hands many times before reaching you and those transfers didn't all occur on the blockchain (could have happened on an exchange, could have happened on lightning network, could happen on another blockchain like eth).
If you look at actual transactions on the blockchain, you'll see that they have an average value of ~300-400K [1] per transaction which means it's acting as final settlement.
1: https://bitinfocharts.com/comparison/bitcoin-transactionvalu...
Blockchains all have a bottleneck. All transactions in the world have to go through the miner. And the situation is actually worse than that because we don't know who will solve the useless Proof of Work hashcash function ahead of time, so EVERY potential transaction is gossipped to EVERY potential miner, and then the miners ALL go to town wasting electricity. How great!
Anyway, because of this architecture (preventing double-spending by literally aggregating all ambiguities into N potential histories of EVERYTHING, and then having them duke it out with PoW or PoS or any other way), we literally have a cap on how many "TPS" (Transactions Per Second) can go through.
Before Bitcoin, no one ever asked how many transactions per second an Internet protocol could support. It would be ridiculous to ask that of HTTP, or FTP or SMTP because there is no central bottleneck: the number of websites, files and emails that can be sent increases with the number of computers on the network.
With Bitcoin, Ethereum and other blockchain-based networks, this is not the case. The max number of transactions that fit inside a block is the most throughput you're going to get. The best thing to come out of these blockchains is Microsoft's Sidetree protocol championed by @csuwildcat but that's for identities, not currencies. So basically, you have a payment system which is super-secure and also super-wasteful.
It's as if BitTorrent required every computer to seed every movie, and BitTorrent maximalists called everything else that used a Kademlia DHT a ShitTorrent.
Except here, it's actually worse because you're not doing Proof of SpaceTime (like FileCoin) but you're requiring a growing number of computers to participate in a lottery proving they burn electricity and get rewarded sometimes. And this amount of computers is growing, just so we can do one operation (transfer money) 10 times a second. At least Ethereum made it into virtual machine capable of executingmore than one operation, thus enabling Turing-complete programming languages like Solidity, and smart contracts. Still, the way it's designed as a "world computer" it's basically a glorified mainframe, and gas is basically paying for time on the mainframe. It is the reason "flash loans" are even possible, they are artifacts of the idea that literally ONE transaction in the entire world can happen at any given time. Nice for atomicity guarantees, I guess, but overall not so great for almost every real world application.
Blockchains with global consensus about every transaction in the world are NOT what I would think of when I think "decentralized" in the sense of having no central bottlenecks. Actually @vbuterin wrote a great article years ago about the different senses of the word "decentralized".