5 ms·

I agree that oracles make many use-cases over-hyped, but IMO you shouldn't overlook that Ethereum is far better at transacting ETH than Bitcoin is for transacti

by markkat 11y ago

I agree that oracles make many use-cases over-hyped, but IMO you shouldn't overlook that Ethereum is far better at transacting ETH than Bitcoin is for transacting BTC. ETH has a powerful use case as programmable money which settles faster and cheaper than any other, and has native support for payment channels and other complexities not only because of it's Turing complete language, but because it does not use bitcoin's UTXO architecture. Not all interested in Ethereum view smart contracts as the best use case.

jerguismi 11y ago
> Ethereum is far better at transacting ETH than Bitcoin is for transacting BTC. Care to elaborate the benefits of using ethereum for value transfers vs. bitcoin? Block size limit is an obvious advantage, but what are the others? > ETH has a powerful use case as programmable money which settles faster and cheaper than any other That is quite a claim, which makes me doubt if you know what you are talking about. In blockchain-based systems the costs are related to market dynamics. Eg. miners can decide what they put to blocks and at what price. Or is ethereum somehow fundamentally different in this aspect? > and has native support for payment channels and other complexities not only because of it's Turing complete language, but because it does not use bitcoin's UTXO architecture. Payment channels are perfectly supported by Bitcoin, what do you mean by "native"? Is it how much easier to use payment channels in ETH in practice, are people actually using them? Why is ethereums account-based architecture better than Bitcoin's UTXO architecture? Isn't really obvious to me.
- woah 11y ago
  Take your head out of the sand. Everything is far easier to program with an actual programming language (even a somewhat mediocre one, like ethereum's solidity). It's always amusing to see the vast number of hoops that one must jump through to shoehorn functionality into bitcoin script (a non Turing complete assembly).
  - DickingAround 11y ago
    He asked a reasonable question; give a reasonable answer. For example 'bitcoin script' or 'non Turing complete assembly' is not a term that makes sense. Describe specifics.
    - woah 11y ago
      It's pretty simple: you can use an actual programming language with Ethereum. If you're not familiar with Bitcoin, "Bitcoin script" is the pseudolang that must be used to program Bitcoin. One of its most prominent features is that it has no loops. Most Bitcoin applications involve a ton of jumping through hoops to even make them work at all. The equivalent programs on Ethereum are short and look like C. If Bitcoin can add real programmability, Ethereum is done for. But so far, the Bitcoin community has not even been able to change a single constant (I'm referring to the block size debate).
    - nix_said 11y ago
      Ethereum relies on a defined opscode language, if anything this is the new thing about Ethereum. The return function in Bitcoin is storage, store the opscode in the return function. In Ethereum you store the opscode in a special data field. The thing the Ethereum client adds is the ability to take the data in the transaction storage (weather it be return or a defined datafield makes no difference) and pass it through an EVM to generate the current state. So there is no reason for that functionality to be limited to just the Ethereum client. You would take that piece of the client and run the opscode in the return data storage across several transactions. I thought this was obvious to anyone who understands the discussion on a reasonably technical level. I suspect many posters in this thread are just Ethereum investors with no programming knowledge.
- DennisP 11y ago
  Here's an article by Ethereum's creator on the reasons they moved to an account-based architecture: https://medium.com/@ConsenSys/thoughts-on-utxo-by-vitalik-buterin-2bb782c67e53#.kwwgetcpl https://medium.com/@ConsenSys/thoughts-on-utxo-by-vitalik-bu...
- markkat 11y ago
  14 second low-variance block times is one that makes an impression from a user perspective. >In blockchain-based systems the costs are related to market dynamics. Eg. miners can decide what they put to blocks and at what price. Or is ethereum somehow fundamentally different in this aspect? Both Bitcoin and Ethereum are fundamentally different from current payment systems, and as Ethereum goes to PoS, it will likely be quite different than Bitcoin's PoW. Both probably represent an improvement on current transaction methods, i.e. credit card with 1-2%, plus $0.2 to $0.3 flat processing fees. >Payment channels are perfectly supported by Bitcoin, what do you mean by "native"? I mean that off-chain channels can be integrated without changes to the current Ethereum code. http://www.arcturnus.com/ethereum-lightning-network-and-beyond/ http://www.arcturnus.com/ethereum-lightning-network-and-beyo... >Why is ethereums account-based architecture better than Bitcoin's UTXO architecture? Isn't really obvious to me. For bitcoin one txn requires n signatures for n unspent outputs. For Ethereum one txn requires just one signature. This has consequences for efficiency of micropayments and payment channels. https://blog.ethereum.org/2014/09/17/scalability-part-1-building-top/ https://blog.ethereum.org/2014/09/17/scalability-part-1-buil...
  - tromp 11y ago
    Adoption of segwit in Bitcoin will pave the way for use of Schnorr signatures, offering comparable efficiency gains.
    - kristianp 11y ago
      Segregated witness and other changes seem like positive news for bitcoin scaling. https://bitcoincore.org/en/2015/12/23/capacity-increases-faq/ https://bitcoincore.org/en/2015/12/23/capacity-increases-faq...

by markkat 11y ago

jerguismi 11y ago

> Ethereum is far better at transacting ETH than Bitcoin is for transacting BTC. Care to elaborate the benefits of using ethereum for value transfers vs. bitcoin? Block size limit is an obvious advantage, but what are the others? > ETH has a powerful use case as programmable money which settles faster and cheaper than any other That is quite a claim, which makes me doubt if you know what you are talking about. In blockchain-based systems the costs are related to market dynamics. Eg. miners can decide what they put to blocks and at what price. Or is ethereum somehow fundamentally different in this aspect? > and has native support for payment channels and other complexities not only because of it's Turing complete language, but because it does not use bitcoin's UTXO architecture. Payment channels are perfectly supported by Bitcoin, what do you mean by "native"? Is it how much easier to use payment channels in ETH in practice, are people actually using them? Why is ethereums account-based architecture better than Bitcoin's UTXO architecture? Isn't really obvious to me.

woah 11y ago
Take your head out of the sand. Everything is far easier to program with an actual programming language (even a somewhat mediocre one, like ethereum's solidity). It's always amusing to see the vast number of hoops that one must jump through to shoehorn functionality into bitcoin script (a non Turing complete assembly).
- DickingAround 11y ago
  He asked a reasonable question; give a reasonable answer. For example 'bitcoin script' or 'non Turing complete assembly' is not a term that makes sense. Describe specifics.
  - woah 11y ago
    It's pretty simple: you can use an actual programming language with Ethereum. If you're not familiar with Bitcoin, "Bitcoin script" is the pseudolang that must be used to program Bitcoin. One of its most prominent features is that it has no loops. Most Bitcoin applications involve a ton of jumping through hoops to even make them work at all. The equivalent programs on Ethereum are short and look like C. If Bitcoin can add real programmability, Ethereum is done for. But so far, the Bitcoin community has not even been able to change a single constant (I'm referring to the block size debate).
  - nix_said 11y ago
    Ethereum relies on a defined opscode language, if anything this is the new thing about Ethereum. The return function in Bitcoin is storage, store the opscode in the return function. In Ethereum you store the opscode in a special data field. The thing the Ethereum client adds is the ability to take the data in the transaction storage (weather it be return or a defined datafield makes no difference) and pass it through an EVM to generate the current state. So there is no reason for that functionality to be limited to just the Ethereum client. You would take that piece of the client and run the opscode in the return data storage across several transactions. I thought this was obvious to anyone who understands the discussion on a reasonably technical level. I suspect many posters in this thread are just Ethereum investors with no programming knowledge.
DennisP 11y ago
Here's an article by Ethereum's creator on the reasons they moved to an account-based architecture: https://medium.com/@ConsenSys/thoughts-on-utxo-by-vitalik-buterin-2bb782c67e53#.kwwgetcpl https://medium.com/@ConsenSys/thoughts-on-utxo-by-vitalik-bu...
markkat 11y ago
14 second low-variance block times is one that makes an impression from a user perspective. >In blockchain-based systems the costs are related to market dynamics. Eg. miners can decide what they put to blocks and at what price. Or is ethereum somehow fundamentally different in this aspect? Both Bitcoin and Ethereum are fundamentally different from current payment systems, and as Ethereum goes to PoS, it will likely be quite different than Bitcoin's PoW. Both probably represent an improvement on current transaction methods, i.e. credit card with 1-2%, plus $0.2 to $0.3 flat processing fees. >Payment channels are perfectly supported by Bitcoin, what do you mean by "native"? I mean that off-chain channels can be integrated without changes to the current Ethereum code. http://www.arcturnus.com/ethereum-lightning-network-and-beyond/ http://www.arcturnus.com/ethereum-lightning-network-and-beyo... >Why is ethereums account-based architecture better than Bitcoin's UTXO architecture? Isn't really obvious to me. For bitcoin one txn requires n signatures for n unspent outputs. For Ethereum one txn requires just one signature. This has consequences for efficiency of micropayments and payment channels. https://blog.ethereum.org/2014/09/17/scalability-part-1-building-top/ https://blog.ethereum.org/2014/09/17/scalability-part-1-buil...
- tromp 11y ago
  Adoption of segwit in Bitcoin will pave the way for use of Schnorr signatures, offering comparable efficiency gains.
  - kristianp 11y ago
    Segregated witness and other changes seem like positive news for bitcoin scaling. https://bitcoincore.org/en/2015/12/23/capacity-increases-faq/ https://bitcoincore.org/en/2015/12/23/capacity-increases-faq...