9 ms·

Ethereum could quickly outdate itself by defining a fixed language. How will the language adapt? Very slowly by consensus if Ethereum hopes to become a large ma

Ethereum could quickly outdate itself by defining a fixed language. How will the language adapt? Very slowly by consensus if Ethereum hopes to become a large market player. Bitcoin could be easily made Turing complete. If each transaction can store 80 bytes of information, then one could use 48 bytes to store code and the remaining 32 bytes to refer to another transaction storing the next block of code and thus chain together code. Unless I am missing something, a new currency is simply unnecessary, and will always be at a second-mover disadvantage. Soon, I am sure a prototype of what I described will be released on bitcoin thus obscuring Ethereum. A solid use-case for distributed computing would be distributed HTTP delivery. If ethereum could offer a product for decentralized web-hosting, then I would buy into their product, but until then the whole situation is just vaporware to me.

adrianmacneil 11y ago
> Ethereum offers nothing more than Bitcion It offers a more powerful scripting language to define contracts. > Ethereum could quickly outdate itself by defining a fixed language. If offers a "bytecode" type language which can be created by compiling contracts in >=2 languages already. Bitcoin currently lets you define contracts in the equivalent of assembly code, with no higher level languages available. So this statment seems to apply more to Bitcoin than Ethereum. > Bitcoin could be easily made Turing complete. Bitcoin developers can't even agree to a simple block size increase. What makes you think they would be able to do something as complex as adding turing-complete scripting capabilities? > Unless I am missing something, a new currency is simply unnecessary. Since bitcoin does not support 2-way pegged sidechains yet, it's not possible to launch a new blockchain without launching a new currency to power that blockchain. Besides, you need a way to reward the miners. > Soon, I am sure a prototype of what I described will be released on bitcion thus obscuring Ethereum. That's basically what Rootstock[0] is supposed to be. Although at present, it seems to be vaporware. A year ago I used to have a similar viewpoint to you, that Bitcoin had such a massive first mover advantage and network effects that it would remain the dominant public blockchain, and that even as new inventions appeared, they would be backported to bitcoin. But over the past 6 months I've realized how useless Bitcoin is without a genuine respected leader/BDFL, and now think that without a leader it's doomed to design by committee and stalemate on key decisions, which leaves the playing field wide open for competitors like Ethereum. [0] http://www.rootstock.io/ http://www.rootstock.io/
- Anonobread 11y ago
  > and now think that without a leader it's doomed to design by committee and stalemate on key decisions As a regular in the cryptocurrency world, it's evident that no matter what you do, someone is going to find a reason to criticize you. One day these people are criticising Bitcoin for being run by a cabal of closely knit developers, the next you see this guy saying Bitcoin doesn't have a leader and is instead being designed by an apparently leaderless committee. They seem to alter their concerns and their approach to criticising Bitcoin depending on which way the wind is blowing. BTW, I'm going to assume something and you should too: adrianmacneil here is almost assuredly long ETH. He probably albeit less assuredly so doesn't own any BTC. In any case you can assume I have the opposite portfolio. First and foremost, Bitcoin and Ethereum couldn't possibly be more different. Bitcoin is a "Peer-to-Peer Electronic Cash System". Bitcoin places a higher value on securing the cash. It places tamper resistance above all else, and generally takes no chances on that. That's a feature, not a bug. Because if that core unit fails, then the entire cash system collapses. You can't exactly have a global currency that fails mid voyage. Modern financial systems don't do that very often and neither should Bitcoin. Conversely, Ethereum is a smart contracts platform through and through. It has 15 second block times. It has its own VM. This VM has had critical vulnerabilities in the recent past, and is less than 1 year into production. Notice how Bitcoin could never be about smart contracts like Ethereum does, at its core, since Bitcoin could only do that by taking on existential risk? However, that doesn't mean Bitcoin can't better manage its risk by making what is essentially a "unikernel" for Ethereum functionality on top of the Bitcoin network. Sure it won't be totally perfect, but it will do everything Ethereum does, without exception, and will do so without compromising the core of Bitcoin's network, when doing so would jeopardize the cash system. And this is what the poster I'm responding to completely misses when he says: > [Ethereum] offers a "bytecode" type language which can be created by compiling contracts in >=2 languages already. Bitcoin currently lets you define contracts in the equivalent of assembly code, with no higher level languages available. So this statment seems to apply more to Bitcoin than Ethereum. The point is, once you have the unikernel model down, who cares if it's Ethereum? Ethereum's VM is very first generational, highly unproven, highly unstable, has God knows how many vulnerabilities not to mention competitors who develop competing languages and competing VMs. We don't want to be in that business. We want to be in the business of running the best of it as a unikernel without even remotely jeopardizing the entire system all at once, though I understand Ethereum kind of had to do that in order to raise money. Turns out if you don't raise a bunch of money from a bunch of investors, you don't get people to voluntarily promote you on HN and Reddit. > That's basically what Rootstock[0] is supposed to be. Although at present, it seems to be vaporware. Speaking of Ethereum promoters, they do have a knack for reminding you how the unikernels of cryptocurrency will surely never exist. Nothing to see here.
natrius 11y ago
Arguing about which blockchain will be used for computation is like arguing about PowerPC outcompeting Intel chips: it only really matters to people with a financial stake in the outcome. Don't get so distracted by speculation that you overlook that as a programmer, this technology allows you to build powerful new kinds of software.
- nix_said 11y ago
  I have used it and it has many many limitations. A lot of the people coming up with random ideas + a blockchain don't realize that a lot of those ideas would not be improved by adding a blockchain.
eyci 11y ago
Hmm your post led me to discover that it's not permitted to have more than one OP_RETURN output per transaction. Any idea why this is the case?
bronz 11y ago
Can someone explain why chaining code together on the block chain is useful? I have heard this idea before but never seen an explanation of its utility.
sgt101 11y ago
Turing complete means "has loops and branches" it is a property of languages, not transaction protocols and data stores like Bitcoin.
riprowan 11y ago
> Bitcoin could be easily made Turing complete. A simple increase from 1MB blocks as proposed in 2010 by the software's creator is treated by obstructionists like Greg Maxwell as "theft" and a violation of social contract and blocked for six years. https://bitcointalk.org/index.php?topic=1347.0 https://bitcointalk.org/index.php?topic=1347.0 Some of the original obstructionists are still at it, and so far, Mike Hearn is right: it's working. How gullible would one have to be to think that reengineering Turing-completeness into the system will be remotely possible?

adrianmacneil 11y ago

> Ethereum offers nothing more than Bitcion It offers a more powerful scripting language to define contracts. > Ethereum could quickly outdate itself by defining a fixed language. If offers a "bytecode" type language which can be created by compiling contracts in >=2 languages already. Bitcoin currently lets you define contracts in the equivalent of assembly code, with no higher level languages available. So this statment seems to apply more to Bitcoin than Ethereum. > Bitcoin could be easily made Turing complete. Bitcoin developers can't even agree to a simple block size increase. What makes you think they would be able to do something as complex as adding turing-complete scripting capabilities? > Unless I am missing something, a new currency is simply unnecessary. Since bitcoin does not support 2-way pegged sidechains yet, it's not possible to launch a new blockchain without launching a new currency to power that blockchain. Besides, you need a way to reward the miners. > Soon, I am sure a prototype of what I described will be released on bitcion thus obscuring Ethereum. That's basically what Rootstock[0] is supposed to be. Although at present, it seems to be vaporware. A year ago I used to have a similar viewpoint to you, that Bitcoin had such a massive first mover advantage and network effects that it would remain the dominant public blockchain, and that even as new inventions appeared, they would be backported to bitcoin. But over the past 6 months I've realized how useless Bitcoin is without a genuine respected leader/BDFL, and now think that without a leader it's doomed to design by committee and stalemate on key decisions, which leaves the playing field wide open for competitors like Ethereum. [0] http://www.rootstock.io/ http://www.rootstock.io/

Anonobread 11y ago
> and now think that without a leader it's doomed to design by committee and stalemate on key decisions As a regular in the cryptocurrency world, it's evident that no matter what you do, someone is going to find a reason to criticize you. One day these people are criticising Bitcoin for being run by a cabal of closely knit developers, the next you see this guy saying Bitcoin doesn't have a leader and is instead being designed by an apparently leaderless committee. They seem to alter their concerns and their approach to criticising Bitcoin depending on which way the wind is blowing. BTW, I'm going to assume something and you should too: adrianmacneil here is almost assuredly long ETH. He probably albeit less assuredly so doesn't own any BTC. In any case you can assume I have the opposite portfolio. First and foremost, Bitcoin and Ethereum couldn't possibly be more different. Bitcoin is a "Peer-to-Peer Electronic Cash System". Bitcoin places a higher value on securing the cash. It places tamper resistance above all else, and generally takes no chances on that. That's a feature, not a bug. Because if that core unit fails, then the entire cash system collapses. You can't exactly have a global currency that fails mid voyage. Modern financial systems don't do that very often and neither should Bitcoin. Conversely, Ethereum is a smart contracts platform through and through. It has 15 second block times. It has its own VM. This VM has had critical vulnerabilities in the recent past, and is less than 1 year into production. Notice how Bitcoin could never be about smart contracts like Ethereum does, at its core, since Bitcoin could only do that by taking on existential risk? However, that doesn't mean Bitcoin can't better manage its risk by making what is essentially a "unikernel" for Ethereum functionality on top of the Bitcoin network. Sure it won't be totally perfect, but it will do everything Ethereum does, without exception, and will do so without compromising the core of Bitcoin's network, when doing so would jeopardize the cash system. And this is what the poster I'm responding to completely misses when he says: > [Ethereum] offers a "bytecode" type language which can be created by compiling contracts in >=2 languages already. Bitcoin currently lets you define contracts in the equivalent of assembly code, with no higher level languages available. So this statment seems to apply more to Bitcoin than Ethereum. The point is, once you have the unikernel model down, who cares if it's Ethereum? Ethereum's VM is very first generational, highly unproven, highly unstable, has God knows how many vulnerabilities not to mention competitors who develop competing languages and competing VMs. We don't want to be in that business. We want to be in the business of running the best of it as a unikernel without even remotely jeopardizing the entire system all at once, though I understand Ethereum kind of had to do that in order to raise money. Turns out if you don't raise a bunch of money from a bunch of investors, you don't get people to voluntarily promote you on HN and Reddit. > That's basically what Rootstock[0] is supposed to be. Although at present, it seems to be vaporware. Speaking of Ethereum promoters, they do have a knack for reminding you how the unikernels of cryptocurrency will surely never exist. Nothing to see here.

natrius 11y ago

Arguing about which blockchain will be used for computation is like arguing about PowerPC outcompeting Intel chips: it only really matters to people with a financial stake in the outcome. Don't get so distracted by speculation that you overlook that as a programmer, this technology allows you to build powerful new kinds of software.

nix_said 11y ago
I have used it and it has many many limitations. A lot of the people coming up with random ideas + a blockchain don't realize that a lot of those ideas would not be improved by adding a blockchain.

eyci 11y ago

Hmm your post led me to discover that it's not permitted to have more than one OP_RETURN output per transaction. Any idea why this is the case?

bronz 11y ago

Can someone explain why chaining code together on the block chain is useful? I have heard this idea before but never seen an explanation of its utility.

sgt101 11y ago

Turing complete means "has loops and branches" it is a property of languages, not transaction protocols and data stores like Bitcoin.

riprowan 11y ago

> Bitcoin could be easily made Turing complete. A simple increase from 1MB blocks as proposed in 2010 by the software's creator is treated by obstructionists like Greg Maxwell as "theft" and a violation of social contract and blocked for six years. https://bitcointalk.org/index.php?topic=1347.0 https://bitcointalk.org/index.php?topic=1347.0 Some of the original obstructionists are still at it, and so far, Mike Hearn is right: it's working. How gullible would one have to be to think that reengineering Turing-completeness into the system will be remotely possible?