Why a single exchange can trigger adoption of BIP101
Posted: Wed Dec 02, 2015 9:45 am
Today, I'll be following up on my previous post about bitcoin's problems being people problems. I was inspired to write by a blog post that showed up yesterday proposing that an altcoin is the only viable solution to force adoption of a blocksize increase. I agree that the current stalemate needs decisive action to force adoption. In the absence of a crisis, people are content to allow bitcoin to flounder while banks and other entities adopt off-chain solutions. It's not an accident that the hype around private blockchains is occurring during this time of inaction.
I'm still amazed at how few people seem to recognize that the blocksize crisis has reduced investment and stalled out development in bitcoin. For good or for bad, we are one of those few; I stopped buying bitcoins a year ago, and will not do so again unless the block size is increased. Chris goes even further and sells off all mining profits he earns as soon as possible. We have no doubt that cryptocurrencies are the future, but see investment in the bitcoin ecosystem as a waste of money when altcoins like NXT and Ethereum, which are led by people trying to solve problems in a positive manner, show much more promise. Holding bitcoins is also a poor idea because the coming DDoS attacks and extortion to prevent a blocksize increase will undoubtedly push prices down to the $100 range, presenting better opportunities to buy if bitcoin survives the upgrade process.
Since an increase is not going to occur slowly over time in the absence of a triggering event, I thought it might be interesting to examine what factors could trigger the adoption of a blocksize increase, and which are unlikely to do so. Then, I'll conclude that the most likely trigger for a blocksize increase is an exchange releasing production code supporting BIP 101.
Events unlikely to trigger an increase
Conferences and physical meetings
The big news this week is the "Scaling Bitcoin 2" conference that will be held in Hong Kong. Some are hopeful that progress will be made, but there are many reasons to be less optimistic. First, conferences are expensive and time-consuming; traveling to Hong Kong can require as many as 36 hours in each direction, with additional recovery time necessary for the timezone changes. The thousands of dollars and two man-weeks of traveling and recovery time could instead be spent actually implementing code. At the end of this multi-million dollar conference, the attendees will return with good feelings but nothing to show for all that effort. The money could have instead been spent on creating solutions and then testing them to discover benefits and drawbacks.
Not only that, but some of the most important people aren't attending the conference. Many discussions will likely come to a conclusion of "this sounds good, but let's ask X, who isn't here." An online videoconference could easily have included all the attendees because the expense and time commitment would not be barriers.
In addition, conferences like the one in Hong Kong have the issue that some of the people attending are not qualified to be discussing all the topics. The organizers, therefore, have two choices: they can either lower the quality of the discussion so these people can participate, in which case all that will happen is that followup meetings will be necessary - or, they can discuss highly technical issues and uninformed attendees will slow down the conversation by needing to be brought up to speed.
Finally, the timing couldn't have been worse. After the conference is over, the holidays will arrive. Everyone will go home and come back in January, having forgotten most of what was discussed. Any goodwill that was earned will be forgotten.
Transaction volume rises
In a previous post, I pointed out that increased fees due to increasing transaction volume causes people to transact through altcoins. It might seem that transaction volume alone would prompt an upgrade, but the people most affected by transaction fees are individual buyers and merchants. The guy who has to pay $1 to send bitcoins has no power to cause anyone to upgrade. To complicate matters, miners have a perverse incentive to keep transaction fees high by retaining smaller blocks so that they can earn more money.
Fee increases alone will simply encourage people to use altcoins, as has been previously shown, which actually decreases volume on the bitcoin network. Thus, transaction fees will never cause people to upgrade and instead make it more likely that a more capable altcoin will eventually surpass bitcoin in popularity and everyday usage.
Time is allowed to pass
After the rise in early November shocked people with how much it cost to send transactions, the status quo resumed. Since it costs a lot for companies to upgrade, no solution will be adopted unless there is a business case for them to upgrade. But since no company can unilaterally upgrade the network, a tragedy of the commons occurs with nobody wanting to go first despite everyone being better off after the upgrade.
It's not a coincidence that the rise in November was prevented from turning into a bubble. If another rise develops, then it too will be stymied by the transaction limit. Volumes will decline after volatility subsides, and bitcoin will continue to remain stagnant. I maintained last year that future bubbles would be aborted by the transaction limit. It happened last month, and will continue to happen (while altcoins continue to see increased usage and rise in value) unless some other factor forces a change.
Furthermore, the idea that "more discussion" will lead to consensus, if that discussion was allowed to occur, has been definitively proven to be false. The Core developers have held at least three consecutive weekly meetings discussing bitcoin's future development and they have ignored the blocksize issue. The transcripts record discussion of minor issues while the one issue preventing bitcoin from advancing is nowhere to be found. Additionally, there has yet to be a public statement made by the Core developers against censorship and for fair debate. While there is some truth to the idea that theymos is dragging down the reputation of everyone on his side of the debate, it is hard to side with those who refuse to condemn his actions.
This isn't an issue of a few people being unable to come to agreement, as it is presented in the media. The lack of any discussion whatsoever demonstrates that the people attending those meetings are completely disinterested in negotiating a solution to the problem.
A large-scale miner upgrades to Bitcoin XT
If a major miner upgrades to Bitcoin XT, then others may take that as a sign that the proposal is gathering momentum. Slush's pool, which will re-enable BIP101 mining soon, could also see a portion of hashrate move over to its BIP101 nodes. But there are several reasons to be cautious about using hashpower as an indicator for adoption.
First, a pool that gives miners a choice of which fork to mine is less likely to initiate an upgrade. While a plurality of support (50.0%) existed for BIP101 several months ago, it takes time for miners to reconfigure their equipment to point to the BIP101 nodes. Time is money, and any time computer software is reconfigured, the opportunity for bugs exists. We only make changes to our mining server when they are absolutely necessary, because a single mistake can cause huge losses before it is discovered. Miners who don't gain any personal benefit from switching will be risk-averse.
If pools aren't able to initiate a change, then perhaps companies that own all their miners would. But even if a company controlling 20% of the hashrate switches to larger blocks, other miners are unlikely to follow suit. The reason is that there are very few nodes installed to accept their blocks, and exchanges wouldn't recognize them either. An orphaned block is a huge deal to bitcoin miners, which causes major losses.
Miners have little reason to be the first to upgrade and will probably only follow suit once other players take the lead. This is why there have been so few BIP101 blocks mined to date. In fact, most miners could upgrade and find that exchanges and merchants are continuing to use the old fork, forcing a downgrade - and knowing that, they wouldn't upgrade in the first place.
Events likely to trigger an increase
theymos is arrested and charged with some crime
In a previous post, I stated that theymos, who controls the majority of the bitcoin media, is singlehandedly blocking consensus on a blocksize solution. By sowing discord with censorship and post manipulation, he is able to waste inordinate amounts of people's time on arguments over his moderation, which could otherwise be focused on the actual problem. He creates arguments between posters and attacks public figures like Mike Hearn so that these people view the other side as demons, making them unwilling to talk with each other.
Since most people support larger blocks, the only way that theymos is able to claim that no consensus exists is to delete discussions that would likely result in consensus. This is the key argument in why simply allowing discussion to proceed naturally would trigger an upgrade. After all, there isn't really any reason why the size of blocks should be any more important than, say, replace-by-fee, which is a much more significant change in the network's behavior but which did not generate controversy because everyone was willing to talk to each other over it. The only reason the blocksize debate is contentious is because people are being made to hate each other.
Interestingly, in addition to blocking adoption of a blocksize increase, theymos has (to his own detriment) forced the discussion for the last 4 months into a "BIP101 vs. nothing at all" debate. With nobody able to present working code as an alternative to BIP101 on his forums, and BIP101 supporters largely having abandoned his sites, he has made adoption of BIP101 far more likely simply because he prevented an alternative from being created.
Since theymos owns so many sites, something that took one of them down would not significantly reduce his influence. However, theymos has been involved in many controversies, including millions of dollars in forum donations that remain unaccounted for and accusations from Coin Fire that he is being paid for his activities. We should be clear that these are not our accusations and we aren't aware of any evidence for or against them. If one of these issues were found to be true and resulted in criminal charges, his inability to run the sites from jail, the likely seizure of assets to pay for maintenance, and the unwillingness of supporters to be associated with him would result in the collapse of his control over the debate and an increased probability that a "compromise" solution is found.
This is one of the black swan events that would probably come with no warning and which traders would be advised to watch out for.
An exchange adopts BIP101 and upgrades its nodes
Another way that a blocksize increase can be adopted is if several major exchanges release code that supports BIP101 blocks, as they have indicated they would do in a letter from earlier in the year. Exchanges, not miners, have the power to decide which fork of bitcoin is the canonical chain, and this principle has been demonstrated in many altcoins. Cryptsy, for example, singlehandedly reverted an attempted fork of Elacoin by refusing to upgrade its nodes to the new proof-of-stake code. In bitcoin's case, even if miners adopt a different solution, their blocks would be ignored by BIP101-adopted exchanges - and coins that are not listed on any exchanges have no value.
Furthermore, because the actual solution to the blocksize problem is not important, as soon as one exchange adopts BIP101, it is likely that others will follow in rapid succession. There are some indications that Coinbase might trigger this event in as little as two weeks. Even if Bitstamp had not already announced support for BIP101, after Coinbase upgraded it would make little sense for them to implement a solution that is incompatible with Coinbase's implementation. The difference between a 4MB blocksize and an 8MB blocksize is simply irrelevant to their business plan when the cost required for a financial company to make a single upgrade runs in the hundreds of thousands of dollars. Shareholders of exchanges would not appreciate the CEO wasting their money on a different solution to make a statement.
Conclusions
In conclusion, there are two triggering events that are most likely to spur adoption of a blocksize increase. First, theymos could be removed from his position, and in his absence, a currently divided community would find that they agree on most things. People would come up with a compromise solution that probably wouldn't be BIP101; however, the removal of theymos is unlikely to happen in the short term. Second, an exchange could take the lead, which would precipitate other exchanges to follow suit, causing miners to see that their large blocks will now be valid on the exchanges. An exchange rolling out production code that supports BIP101 would change the discussion from which proposal should be used to how long it will take for everyone to upgrade.
Of course, any proposal could ultimately adopted by a similar process. However, I focused on BIP101 for this article not only because it is the only working solution, but also because its adoption is the most positive path forward for bitcoin. We support BIP101 not because it is technically superior to any other solution, but because its activation is the easiest way to purge the bitcoin community of negative actors. If BIP101 succeeds, theymos and his censorship will be discredited, criminal activity by hackers and DDoS will have been proven to be ineffective at preventing progress, and the stonewalling Core developers who repeatedly ignore this problem will be marginalized by more effective leadership. Imagine if a different solution backed by the Core developers is implemented, like the much-hyped Lightning Network. With the same people in charge up to the same tricks, would you be willing to increase your investment in bitcoin then?
Right now, the decision of a single major exchange can tip the balance in any direction, and the first mover could end up as decisive. Exchanges involved in bitcoin need to look at the big picture and recognize that the character of the people who end up in charge after all of this is more important than the size of the blockchain in 2020.
I'm still amazed at how few people seem to recognize that the blocksize crisis has reduced investment and stalled out development in bitcoin. For good or for bad, we are one of those few; I stopped buying bitcoins a year ago, and will not do so again unless the block size is increased. Chris goes even further and sells off all mining profits he earns as soon as possible. We have no doubt that cryptocurrencies are the future, but see investment in the bitcoin ecosystem as a waste of money when altcoins like NXT and Ethereum, which are led by people trying to solve problems in a positive manner, show much more promise. Holding bitcoins is also a poor idea because the coming DDoS attacks and extortion to prevent a blocksize increase will undoubtedly push prices down to the $100 range, presenting better opportunities to buy if bitcoin survives the upgrade process.
Since an increase is not going to occur slowly over time in the absence of a triggering event, I thought it might be interesting to examine what factors could trigger the adoption of a blocksize increase, and which are unlikely to do so. Then, I'll conclude that the most likely trigger for a blocksize increase is an exchange releasing production code supporting BIP 101.
Events unlikely to trigger an increase
Conferences and physical meetings
The big news this week is the "Scaling Bitcoin 2" conference that will be held in Hong Kong. Some are hopeful that progress will be made, but there are many reasons to be less optimistic. First, conferences are expensive and time-consuming; traveling to Hong Kong can require as many as 36 hours in each direction, with additional recovery time necessary for the timezone changes. The thousands of dollars and two man-weeks of traveling and recovery time could instead be spent actually implementing code. At the end of this multi-million dollar conference, the attendees will return with good feelings but nothing to show for all that effort. The money could have instead been spent on creating solutions and then testing them to discover benefits and drawbacks.
Not only that, but some of the most important people aren't attending the conference. Many discussions will likely come to a conclusion of "this sounds good, but let's ask X, who isn't here." An online videoconference could easily have included all the attendees because the expense and time commitment would not be barriers.
In addition, conferences like the one in Hong Kong have the issue that some of the people attending are not qualified to be discussing all the topics. The organizers, therefore, have two choices: they can either lower the quality of the discussion so these people can participate, in which case all that will happen is that followup meetings will be necessary - or, they can discuss highly technical issues and uninformed attendees will slow down the conversation by needing to be brought up to speed.
Finally, the timing couldn't have been worse. After the conference is over, the holidays will arrive. Everyone will go home and come back in January, having forgotten most of what was discussed. Any goodwill that was earned will be forgotten.
Transaction volume rises
In a previous post, I pointed out that increased fees due to increasing transaction volume causes people to transact through altcoins. It might seem that transaction volume alone would prompt an upgrade, but the people most affected by transaction fees are individual buyers and merchants. The guy who has to pay $1 to send bitcoins has no power to cause anyone to upgrade. To complicate matters, miners have a perverse incentive to keep transaction fees high by retaining smaller blocks so that they can earn more money.
Fee increases alone will simply encourage people to use altcoins, as has been previously shown, which actually decreases volume on the bitcoin network. Thus, transaction fees will never cause people to upgrade and instead make it more likely that a more capable altcoin will eventually surpass bitcoin in popularity and everyday usage.
Time is allowed to pass
After the rise in early November shocked people with how much it cost to send transactions, the status quo resumed. Since it costs a lot for companies to upgrade, no solution will be adopted unless there is a business case for them to upgrade. But since no company can unilaterally upgrade the network, a tragedy of the commons occurs with nobody wanting to go first despite everyone being better off after the upgrade.
It's not a coincidence that the rise in November was prevented from turning into a bubble. If another rise develops, then it too will be stymied by the transaction limit. Volumes will decline after volatility subsides, and bitcoin will continue to remain stagnant. I maintained last year that future bubbles would be aborted by the transaction limit. It happened last month, and will continue to happen (while altcoins continue to see increased usage and rise in value) unless some other factor forces a change.
Furthermore, the idea that "more discussion" will lead to consensus, if that discussion was allowed to occur, has been definitively proven to be false. The Core developers have held at least three consecutive weekly meetings discussing bitcoin's future development and they have ignored the blocksize issue. The transcripts record discussion of minor issues while the one issue preventing bitcoin from advancing is nowhere to be found. Additionally, there has yet to be a public statement made by the Core developers against censorship and for fair debate. While there is some truth to the idea that theymos is dragging down the reputation of everyone on his side of the debate, it is hard to side with those who refuse to condemn his actions.
This isn't an issue of a few people being unable to come to agreement, as it is presented in the media. The lack of any discussion whatsoever demonstrates that the people attending those meetings are completely disinterested in negotiating a solution to the problem.
A large-scale miner upgrades to Bitcoin XT
If a major miner upgrades to Bitcoin XT, then others may take that as a sign that the proposal is gathering momentum. Slush's pool, which will re-enable BIP101 mining soon, could also see a portion of hashrate move over to its BIP101 nodes. But there are several reasons to be cautious about using hashpower as an indicator for adoption.
First, a pool that gives miners a choice of which fork to mine is less likely to initiate an upgrade. While a plurality of support (50.0%) existed for BIP101 several months ago, it takes time for miners to reconfigure their equipment to point to the BIP101 nodes. Time is money, and any time computer software is reconfigured, the opportunity for bugs exists. We only make changes to our mining server when they are absolutely necessary, because a single mistake can cause huge losses before it is discovered. Miners who don't gain any personal benefit from switching will be risk-averse.
If pools aren't able to initiate a change, then perhaps companies that own all their miners would. But even if a company controlling 20% of the hashrate switches to larger blocks, other miners are unlikely to follow suit. The reason is that there are very few nodes installed to accept their blocks, and exchanges wouldn't recognize them either. An orphaned block is a huge deal to bitcoin miners, which causes major losses.
Miners have little reason to be the first to upgrade and will probably only follow suit once other players take the lead. This is why there have been so few BIP101 blocks mined to date. In fact, most miners could upgrade and find that exchanges and merchants are continuing to use the old fork, forcing a downgrade - and knowing that, they wouldn't upgrade in the first place.
Events likely to trigger an increase
theymos is arrested and charged with some crime
In a previous post, I stated that theymos, who controls the majority of the bitcoin media, is singlehandedly blocking consensus on a blocksize solution. By sowing discord with censorship and post manipulation, he is able to waste inordinate amounts of people's time on arguments over his moderation, which could otherwise be focused on the actual problem. He creates arguments between posters and attacks public figures like Mike Hearn so that these people view the other side as demons, making them unwilling to talk with each other.
Since most people support larger blocks, the only way that theymos is able to claim that no consensus exists is to delete discussions that would likely result in consensus. This is the key argument in why simply allowing discussion to proceed naturally would trigger an upgrade. After all, there isn't really any reason why the size of blocks should be any more important than, say, replace-by-fee, which is a much more significant change in the network's behavior but which did not generate controversy because everyone was willing to talk to each other over it. The only reason the blocksize debate is contentious is because people are being made to hate each other.
Interestingly, in addition to blocking adoption of a blocksize increase, theymos has (to his own detriment) forced the discussion for the last 4 months into a "BIP101 vs. nothing at all" debate. With nobody able to present working code as an alternative to BIP101 on his forums, and BIP101 supporters largely having abandoned his sites, he has made adoption of BIP101 far more likely simply because he prevented an alternative from being created.
Since theymos owns so many sites, something that took one of them down would not significantly reduce his influence. However, theymos has been involved in many controversies, including millions of dollars in forum donations that remain unaccounted for and accusations from Coin Fire that he is being paid for his activities. We should be clear that these are not our accusations and we aren't aware of any evidence for or against them. If one of these issues were found to be true and resulted in criminal charges, his inability to run the sites from jail, the likely seizure of assets to pay for maintenance, and the unwillingness of supporters to be associated with him would result in the collapse of his control over the debate and an increased probability that a "compromise" solution is found.
This is one of the black swan events that would probably come with no warning and which traders would be advised to watch out for.
An exchange adopts BIP101 and upgrades its nodes
Another way that a blocksize increase can be adopted is if several major exchanges release code that supports BIP101 blocks, as they have indicated they would do in a letter from earlier in the year. Exchanges, not miners, have the power to decide which fork of bitcoin is the canonical chain, and this principle has been demonstrated in many altcoins. Cryptsy, for example, singlehandedly reverted an attempted fork of Elacoin by refusing to upgrade its nodes to the new proof-of-stake code. In bitcoin's case, even if miners adopt a different solution, their blocks would be ignored by BIP101-adopted exchanges - and coins that are not listed on any exchanges have no value.
Furthermore, because the actual solution to the blocksize problem is not important, as soon as one exchange adopts BIP101, it is likely that others will follow in rapid succession. There are some indications that Coinbase might trigger this event in as little as two weeks. Even if Bitstamp had not already announced support for BIP101, after Coinbase upgraded it would make little sense for them to implement a solution that is incompatible with Coinbase's implementation. The difference between a 4MB blocksize and an 8MB blocksize is simply irrelevant to their business plan when the cost required for a financial company to make a single upgrade runs in the hundreds of thousands of dollars. Shareholders of exchanges would not appreciate the CEO wasting their money on a different solution to make a statement.
Conclusions
In conclusion, there are two triggering events that are most likely to spur adoption of a blocksize increase. First, theymos could be removed from his position, and in his absence, a currently divided community would find that they agree on most things. People would come up with a compromise solution that probably wouldn't be BIP101; however, the removal of theymos is unlikely to happen in the short term. Second, an exchange could take the lead, which would precipitate other exchanges to follow suit, causing miners to see that their large blocks will now be valid on the exchanges. An exchange rolling out production code that supports BIP101 would change the discussion from which proposal should be used to how long it will take for everyone to upgrade.
Of course, any proposal could ultimately adopted by a similar process. However, I focused on BIP101 for this article not only because it is the only working solution, but also because its adoption is the most positive path forward for bitcoin. We support BIP101 not because it is technically superior to any other solution, but because its activation is the easiest way to purge the bitcoin community of negative actors. If BIP101 succeeds, theymos and his censorship will be discredited, criminal activity by hackers and DDoS will have been proven to be ineffective at preventing progress, and the stonewalling Core developers who repeatedly ignore this problem will be marginalized by more effective leadership. Imagine if a different solution backed by the Core developers is implemented, like the much-hyped Lightning Network. With the same people in charge up to the same tricks, would you be willing to increase your investment in bitcoin then?
Right now, the decision of a single major exchange can tip the balance in any direction, and the first mover could end up as decisive. Exchanges involved in bitcoin need to look at the big picture and recognize that the character of the people who end up in charge after all of this is more important than the size of the blockchain in 2020.