Is Bitcoin worth saving?
Posted: Mon Aug 01, 2016 9:00 pm
Anyone who has been using the Bitcoin network recently knows that events are quickly approaching a turning point. The Chinese miners have reneged on their promise to fork the Bitcoin network on August 1, adding yet another lie to a series of their dishonest statements dating back for years. There no longer appears to be any chance that the existing people involved with Bitcoin, most of whom continue to lie and take no action, will do anything to resolve Bitcoin's usability problems. As I suggested a few months ago, there is not enough time for the world to wait for the Lightning Network or for one of the complex solutions the Core wants to implement.
While some are trying to explain the breach of support in bitcoin price as a short-term panic over recent meetings, I suspect that few believe these "discussions" will accomplish anything more than the tens of expensive conferences and "meetups" have over the past year. Instead, the bitcoin price is likely falling due to a more fundamental issue - the network is unusuable.
A few days ago, we rushed to sell all of our Ethereum Classic (ETC). We decided to convert it to 40 ETH and $170, but mistakenly forgot that the GDAX supported ETH - so we thought we had to use BTC to send the cash to be sold as dollars to the GDAX. It took about 24 hours between the time that Chris began to transfer the BTC until we submitted the ACH request. The BTC transfer took 18 hours to confirm, and the confirmations happened while he was asleep, and then he had to come back later and log in, wait for my cell phone because his is associated with another account, and execute the trade. Meanwhile, Chris withdrew the ETH to an offline wallet, where it was received and confirmed in about 30 seconds. The transaction cost 1/100 of the BTC transfer.
Until now, BTC was the only product supported by most exchanges and businesses. When Coinbase added ETH to its wallets, that changed. We will never again use BTC to sell out to dollars, because the opportunity cost in Chris's time in having to log in again later is enormous. With ETH, we can get our profit to the GDAX and converted to dollars within a few minutes, using only one login session. Plus, the gas costs for Ethereum are so much lower that it just doesn't make sense to use BTC when transferring money between customers anymore. We have a task in our bugtracking system to convert the currency we fund customers' Coinbase dollar payouts with into ETH so that we can retain more profit from the reduced fees. This is a small change that we estimate will save $300/yr even if BTC fees never increase any more.
In many of the non-censored Bitcoin forums, it seems that a movement is brewing to hard fork the Bitcoin blockchain without consulting miners. There are multiple suggestions being considered. One idea is to use merge mining like I proposed on July 12. There is already code written to use scrypt, in order to get rid of the Chinese miners while still providing ASIC protection. User singularity87 wants to use a new proof of work algorithm, one which can be mined with GPUs, at least initially.
The major disagreement between all the users seems to be about how the blocksize limit should be increased in the fork, with some people believing that the problem should be resolved permanently and others believing that it will be possible to hard fork again later. While I don't think there is a need for a blocksize limit at all, I fully support all of these efforts and believe that Bitcoin will be better off if someone makes a last-ditch effort before the network begins to collapse.
However, as I've thought about these plans, the enormity of the work required to pull them off becomes very plain. It would take hundreds of hours in a very short period of time to get a fork client ready for release. Advertising would be needed, and someone would have to contact and work with exchanges to support the fork. If the fork includes a new transaction format to prevent replay attacks as some are suggesting, backwards compatibility will be broken and that alone may doom the effort. Testing may reveal a critical flaw that requires significant rework.
Even then, the odds of getting significant support from businesses and others who actually use the network is low. Like Ethereum Classic, it would likely initially be propped up by people pushing a political ideology, but eventually the supply of those type of people will run out and actual users will determine the coin's success. The effort required is huge, and there nobody has stated he will work the 80 hours per week required for the next few months to lead everyone and get the fork released.
Even if the code is able to be tested, and released, and exchanges support it, and support grows, and the old chain becomes worth very little, and it has users beyond the initial political protestors, it will not resolve most of the problems currently present. Even if scrypt successfully bankrupts the Chinese miners, people like Michael Marquadt and Gregory Maxwell aren't going to disappear; they will likely continue to participate in the new currency. Marquadt will still control all of the Bitcoin media, and the DDoS attackers will continue to fight. And in the end, the system will still be technically inferior to many altcoins.
Nobody has yet convinced me why so much time should be dedicated to forking Bitcoin, and that's why I haven't joined any of these efforts. A few months ago, forking Bitcoin seemed like the best solution, but since then Ethereum support has grown dramatically. The most important change is the addition of trading pairs for ETH at exchanges, and within the past few weeks it is now possible to use ETH interchangeably with Bitcoin for transferring value. Ethereum already has blocks 50 times faster than Bitcoin's, it long ago solved the blocksize problem, it has competent leadership, and the level of ethics amongst its Core contributors is higher. It can do everything Bitcoin can do and more.
In conclusion, while I would love to see Bitcoin fork, I think that the time for that has probably passed. If everyone is going to devote thousands of hours to creating a Bitcoin fork just to end up with a marginally better system hindered by some of the same problems that exist today, isn't it better to devote that time to converting existing applications to Ethereum instead? There already exists a system that is tested, widely supported, and outperforms Bitcoin in every way. Those who are considering forking the Bitcoin network should consider that it may be worth letting Bitcoin decline, instead of putting more work into a system to produce something that, even in the best case, will now be an inferior product.
While some are trying to explain the breach of support in bitcoin price as a short-term panic over recent meetings, I suspect that few believe these "discussions" will accomplish anything more than the tens of expensive conferences and "meetups" have over the past year. Instead, the bitcoin price is likely falling due to a more fundamental issue - the network is unusuable.
A few days ago, we rushed to sell all of our Ethereum Classic (ETC). We decided to convert it to 40 ETH and $170, but mistakenly forgot that the GDAX supported ETH - so we thought we had to use BTC to send the cash to be sold as dollars to the GDAX. It took about 24 hours between the time that Chris began to transfer the BTC until we submitted the ACH request. The BTC transfer took 18 hours to confirm, and the confirmations happened while he was asleep, and then he had to come back later and log in, wait for my cell phone because his is associated with another account, and execute the trade. Meanwhile, Chris withdrew the ETH to an offline wallet, where it was received and confirmed in about 30 seconds. The transaction cost 1/100 of the BTC transfer.
Until now, BTC was the only product supported by most exchanges and businesses. When Coinbase added ETH to its wallets, that changed. We will never again use BTC to sell out to dollars, because the opportunity cost in Chris's time in having to log in again later is enormous. With ETH, we can get our profit to the GDAX and converted to dollars within a few minutes, using only one login session. Plus, the gas costs for Ethereum are so much lower that it just doesn't make sense to use BTC when transferring money between customers anymore. We have a task in our bugtracking system to convert the currency we fund customers' Coinbase dollar payouts with into ETH so that we can retain more profit from the reduced fees. This is a small change that we estimate will save $300/yr even if BTC fees never increase any more.
In many of the non-censored Bitcoin forums, it seems that a movement is brewing to hard fork the Bitcoin blockchain without consulting miners. There are multiple suggestions being considered. One idea is to use merge mining like I proposed on July 12. There is already code written to use scrypt, in order to get rid of the Chinese miners while still providing ASIC protection. User singularity87 wants to use a new proof of work algorithm, one which can be mined with GPUs, at least initially.
The major disagreement between all the users seems to be about how the blocksize limit should be increased in the fork, with some people believing that the problem should be resolved permanently and others believing that it will be possible to hard fork again later. While I don't think there is a need for a blocksize limit at all, I fully support all of these efforts and believe that Bitcoin will be better off if someone makes a last-ditch effort before the network begins to collapse.
However, as I've thought about these plans, the enormity of the work required to pull them off becomes very plain. It would take hundreds of hours in a very short period of time to get a fork client ready for release. Advertising would be needed, and someone would have to contact and work with exchanges to support the fork. If the fork includes a new transaction format to prevent replay attacks as some are suggesting, backwards compatibility will be broken and that alone may doom the effort. Testing may reveal a critical flaw that requires significant rework.
Even then, the odds of getting significant support from businesses and others who actually use the network is low. Like Ethereum Classic, it would likely initially be propped up by people pushing a political ideology, but eventually the supply of those type of people will run out and actual users will determine the coin's success. The effort required is huge, and there nobody has stated he will work the 80 hours per week required for the next few months to lead everyone and get the fork released.
Even if the code is able to be tested, and released, and exchanges support it, and support grows, and the old chain becomes worth very little, and it has users beyond the initial political protestors, it will not resolve most of the problems currently present. Even if scrypt successfully bankrupts the Chinese miners, people like Michael Marquadt and Gregory Maxwell aren't going to disappear; they will likely continue to participate in the new currency. Marquadt will still control all of the Bitcoin media, and the DDoS attackers will continue to fight. And in the end, the system will still be technically inferior to many altcoins.
Nobody has yet convinced me why so much time should be dedicated to forking Bitcoin, and that's why I haven't joined any of these efforts. A few months ago, forking Bitcoin seemed like the best solution, but since then Ethereum support has grown dramatically. The most important change is the addition of trading pairs for ETH at exchanges, and within the past few weeks it is now possible to use ETH interchangeably with Bitcoin for transferring value. Ethereum already has blocks 50 times faster than Bitcoin's, it long ago solved the blocksize problem, it has competent leadership, and the level of ethics amongst its Core contributors is higher. It can do everything Bitcoin can do and more.
In conclusion, while I would love to see Bitcoin fork, I think that the time for that has probably passed. If everyone is going to devote thousands of hours to creating a Bitcoin fork just to end up with a marginally better system hindered by some of the same problems that exist today, isn't it better to devote that time to converting existing applications to Ethereum instead? There already exists a system that is tested, widely supported, and outperforms Bitcoin in every way. Those who are considering forking the Bitcoin network should consider that it may be worth letting Bitcoin decline, instead of putting more work into a system to produce something that, even in the best case, will now be an inferior product.