The third Satoshi Roundtable has just concluded in Cancún, Mexico. The Roundtable is a private gathering of 100 movers and shakers within the bitcoin industry, with no media present, and it’s held under Chatham House rules – meaning everybody can use the information shared at the meeting, but never disclose who said what or their affiliation.
First, it can be acknowledged that an event like this can feel like elitism to those who are not invited. Nevertheless, it must be argued it is a very efficient way to make shit happen, and seeing there’s nothing preventing anybody from inviting people to a relaxed meeting, there’s nothing inherently wrong with doing so just because the whole world isn’t invited. This is the third annual Roundtable (and yes, the knight on the picture is the Roundtable mascot).
These are my impressions, and I’m taking the liberty to be blunt. Being blunt saves time that is otherwise needed to guess what people meant.
The good
First, the meeting was superbly arranged by the Bitcoin Foundation and Bruce Fenton. It was set in just the type of location which makes it easy for participants to hang around the meeting venue instead of taking a taxi to whatever favorite other venue and splitting up, which makes all these magic spontaneous small meetings happen.
For me as a Nordic person, it wasn’t too bad to come to 27°C in the air and 23°C in the water in January, either. Not to mention the intense sunlight. (The sun doesn’t rise in large parts of Scandinavia this time of year.)
It was also a breath of very fresh air, after having felt the antagonistic attitude and tensions in online discussions, to see people meet and honestly try to exchange perspectives and gain understanding. In short, people are behaving not just like civilized adults, but like humble thought leaders. You’d never believe this if you just participated in online discussions.
The outstanding
The productivity in a meeting like this is amazing. It’s so intense, you’re having high-impact discussions with everybody you meet, every five to ten minutes at tightest. Almost every single interaction leads to cooperation opportunities. Meeting so many other movers and shakers in such a tight spacetime means there’s an ignition of opportunity pretty much with every person you speak to.
I also had very fruitful and frank – even blunt – discussions with people who represent ideas and objectives seemingly diametrically opposed from mine in the bitcoin sphere, and there was a humility in the air and a willingness to listen and understand that is plain inconceivable if you are judging by the online toxicity. Again, I’m not allowed to disclose whom or their affiliation, but they deserve a metric ton of credit for the attitude. You know who you are.
I’m leaving Cancún with eleven actionable items I wouldn’t have had without meeting all the amazing people here, just one of which would have been sufficient to justify the trip and attendance.
The beautiful
“Blockchain technology is an extinction event to traditional financial institutions. Like the dinosaurs, a few will survive by morphing into agile birds at a small fraction of their previous weight and size, but most will just die.”
I’m not allowed to disclose who said this, but I think it’s a beautiful quote. Roger Ver also posted the first part of it on Twitter, similarly without disclosing who said it, as is appropriate.
There was also a call for leadership among the people present – “we need to act like the leaders we are, and not tolerate bad attitude in the discussions: at the very least, we need to act as good examples and not foster a bad atmosphere”. This was very refreshing to hear and was met with applause.
The bad
The major bitcoin discussions concerned segwit and the present deadlock. As it’s taking most of the energy right now, I’m also going to devote a lot of space to those impressions. There were many hours of discussion with (a non-identified subset of) people present calling for “segwit adoption and activation now”, plain and simple, with frustrated expressions that Chinese miners are “blocking progress” by not signaling, deploying, and activating segfault, basically “because they should be doing so”.
In a speaking slot of mine, I stood up and made the observation that we (people in the room) are acting like a Toyota boardroom who are trying to make a decision that every family should buy the latest Toyota model. “It doesn’t work like that”, I said. “We’re not the Soviet Politburo commanding a planned economy. The reality of the situation is that we’ve made the market an offer, and the market is rejecting our offer.” I made the point that thinking the market should behave differently, no matter how good your reasons, is not going to make the market behave differently in the slightest. The Toyota boardroom doesn’t get to decide what car a family should buy, and the present company does not get to decide what code miners run on their own machines. The world isn’t fair, but instead of complaining about it, play the cards you’ve got on your hand. Give your client what they want and you both benefit.
Some people seemed to take to this argument. Most didn’t, appearing to be stuck in the mindset that miners are there to serve the community, as opposed to the actual objective, serving themselves only as a rational economic actor.
I find it really, really frustrating that you have a room full of otherwise hyperintelligent people, who were told in very clear terms by the Chinese miners what those miners want about a year ago (a hardfork increasing the max blocksize limit for the present type of transactions to at least 2 megabytes), and today, you have the same people asking in frustration why Chinese miners are not adopting segwit when those miners said in bright blinking cleartext a year ago what it is they want, and it is not segwit. The lack of understanding the customer perspective comes across not just as substandard, but appalling to the level of downright confusing.
The conclusion from this meeting appears to be to do mostly nothing and just expect segwit to activate, possibly lowering the activation threshold (which would not go over well at all). I find that disappointing, because it means that 44 weeks from now, when segwit has definitely failed to activate (when the activation window closes), there will be a flurry of confused activity as what to do next. So in 52 weeks, we’ll be at another Satoshi Roundtable with absolutely no progress at all, if this is the only path worked on.
The alternative, of course, is that a hard fork happens in the meantime. There are at least four levels of hardfork that can take place, and the most likely is that enough miners just switch to a non-Core bitcoin distribution with a higher or dynamic blocksize limit – the second easiest level of hardfork.
I am aware of several hardfork initiatives, at three different levels, that are already underway with work progressing. I predict and anticipate an actual fork event to take place six to 18 months from now. And it’s not going to be “firing Core” as some would frame it; more accurately, it’s going to be “firing Blockstream”. The open question would be if such an event happens in time to preserve bitcoin’s first-mover advantage over other, technically superior coins – my crystal ball is very hazy on this point.
Overall, my assessment is that bitcoin is lacking project management. There’s no clear vision of what the community wants bitcoin to be, who the customer is, or what problem it should solve. This is maybe best illustrated by the rather random statement “It’s now clear that microtransactions have been priced out by rising transaction fees”, as one person stated in the meetings.
Yeah, bitcoin just lost two billion users, and it was presented like a freak, unforeseeable, completely unpreventable accident – like a volcano erupting. In reality, it was foreseeable for years and completely preventable. It’s okay to not prevent such a catastrophic loss of userbase as an active choice or as a choice of whom to target – but it’s not okay to merely observe “oh and by the way, this just happened”.
The ugly
The meeting started with the observation that there are now 70,000 transactions in the backlog. Despite this, a subset of the participants insist that there’s absolutely no problem with network capacity. It’s as if that subset of participants and another subset (including me) are living in different realities.
A further subset of people – I am still not going to name or imply an affiliation because Chatham House rules – insist on the importance of not taking risks with the bitcoin network. This comes across as completely counterfactual to me. We’ve seen a capacity ceiling approach for two years, and now we’ve hit it – how can you talk about not taking risks at the same time as you completely ignore the approaching and completely predictable capacity wall for several years? To many, this would come across as some form of arrogance, but calling it that would be to assign bad faith. I don’t believe in bad faith in this company; I prefer to assign the phenomenon to tunnel vision.
I’ve seen segwit rationalizations that doing a hardfork would take 12 months “and we must choose a fast path now”. Well then, the right thing to do would have been to start executing that hardfork lifting the blocksize limit a year ago, when miners declared loudly and openly that this is what they want. But that wasn’t done, and here we are a year later.
The best time to plant a tree was twenty years ago. The second best time is today. –Chinese proverb
Of course, this argument could be trivially rebutted by claiming that something has been done: segwit. But in my blunt world, it’s results that count, and segwit is not happening. Again, this was completely predictable with a modicum of project management.
Conclusion
In conclusion, it was a very productive meeting that was superbly organized and executed. Some of my preconceptions were shattered, particularly with regard to attitudes and professionalism; others were more or less confirmed, particularly with regard to stances and action plans of various actors and actor groups, as well as my assessment of the political situation at hand.
After this meeting, I’m very bullish on bitcoin’s future. There’s one thing I believe in more than ideas, and that’s people. Specifically, people with the ability to execute those ideas — and there’s a ton of them here. They’re not going to let bitcoin die silently with a whimper, but something will happen to resolve this deadlock, and it’s going to happen sooner rather than later with all this energy around. I cannot predict what that event is — as I said, I’m aware of three hardfork initiatives myself — but I am quite sure such an event will come from somewhere.
Microtransactions will only happen as a 2nd layer solution such as LN. There is no blocksize increase in the world that could possibly ever enable on-chain microtransactions of any meaningful magnitude.
And LN and similar 2nd layer solutions require key improvements such as solving tx malleability. And the only solution seriously proposed for this is… Segwit. All focus should be on making LN happen, we urgently need it. Segwit is by far the best chance to get it up and running in 2017.
LN will be to bitcoin what the www was to the pre-1991 web. Nothing, NOTHING, in terms of mainstream user adoption will happen without it. With it however, everything will happen.
> And the only solution seriously proposed for this is… Segwit
Flexible Transactions is a better solution that allows the Lighting Network to work.
Segwit is code-bloat patchwork bow-tied with a fake blocksize increase.
> And the only solution seriously proposed for this is… Segwit
Flexible Transactions is a better solution that allows the Lighting Network to work.
Segwit is code-bloat patchwork bow-tied with a fake blocksize increase.
“I find it really, really frustrating that you have a room full of otherwise hyperintelligent people, who were told in very clear terms by the Chinese miners what those miners want about a year ago (a hardfork increasing the max blocksize limit for the present type of transactions to at least 2 megabytes)”
Bitcoin is being held hostage by two groups. Blockstream and the Chinese miners. And I can’t say I blame the miners. Blockstream is acting strangely and this is affecting Bitcoin gravely.
Although it is true that microtransactions on the blockchain, if carried out at Visa throughput, would cause it to grow very quickly, it is not accurate to exaggerate by claiming that there is no blocksize increase in the world that could support that kind of use. It is correct on the other hand to point that the growth rate would prove inconvenient for what I would presume is the vast majority of nodes on the network, reducing the total node count, which is not good for the network. With increasing popularity of the network all alternatives to allow for accommodating a larger number of transactions with their own significant tradeoffs. This includes every proposal out there: segwit does not allow directly to scale multiple orders of magnitudep; block size increase will reduce the node count; Lightning will most likely lead to centralization of payment channel providers, and significantly-increased complexity and attack surface. Tunnel vision is choosing not to see this big picture.
Although it is true that microtransactions on the blockchain, if carried out at Visa throughput, would cause it to grow very quickly, it is not accurate to exaggerate by claiming that there is no blocksize increase in the world that could support that kind of use. It is correct on the other hand to point that the growth rate would prove inconvenient for what I would presume is the vast majority of nodes on the network, reducing the total node count, which is not good for the network. With increasing popularity of the network all alternatives to allow for accommodating a larger number of transactions with their own significant tradeoffs. This includes every proposal out there: segwit does not allow directly to scale multiple orders of magnitudep; block size increase will reduce the node count; Lightning will most likely lead to centralization of payment channel providers, and significantly-increased complexity and attack surface. Tunnel vision is choosing not to see the big picture.
Great report, Rick!
Great report, Rick!
“Blockchain technology is an extinction event to traditional financial institutions. Like the dinosaurs, a few will survive by morphing into agile birds at a small fraction of their previous weight and size, but most will just die.”
Listen. I have read Hayek, Nash, Smith, and Szabo.
I am the leading authority on Nash’s 20 years of economic philosophy, who’s insight on equilibrium theory is INTRINSIC in bitcoin (thus he was never cited as his work is just standard, known, and implied).
The person that said this quote, has NEVER read a book on economics in their life.
It is wholly irrational.
Well, Real Life has also not read a single book on economy and that is why Bitcoin works at all and why we had (and still are suffering) under the 2007 economic crises.
Thanks Rick, good report.
It doesn’t sound like anyone at Roundtable properly explained the dangers of a hard fork to you. The risk isn’t technical, it’s economic… Currency survives only with trust, and the public will not trust bitcoins A & B after the split. This argument isn’t in any way ‘trivial,’ it is absolutely life & death to bitcoin, especially with all the contention in the community that you have explained in detail.
I liked your analogy of the toyota boardroom. I’d make one change to it though; The public clearly wants a car that can only go forward when it is pointed at a cliff.
If the miners only want software that is sure to destroy their livelihoods (and all of ours with it) I’m 100% behind Core devs on not giving it to them. They have to be shown how dangerous what they want is.
Luke your argument is not economic in nature.
Losing the public trust in money hurts the economy, not the code.
Code was designed to split; it’s doing it’s job perfectly… Currency should never.
Luke you know nothing about economics.
Jal you are a pain in the ass
What’s the correct term then? It sure ain’t technical either.
You also know nothing about economics.
Thanks, Luke!
I deliberately didn’t go into the specific dangers of a hard fork, but I do acknowledge it needs careful planning. Other people (as linked) have claimed it takes 12-18 months to roll out properly.
Accepting that number at face value (although there would be no shortage of people who would challenge it), and going with 12, I observed that it’s now been 12 months since miners told people in Hong Kong what they want to (and will) run, and that another 12 months from today, segwit still won’t have activated.
So even assuming the careful planning needed, a full year has been wasted and we could have had a capacity upgrade, not to mention a much healthier community.
Cheers,
Rick
Thanks for replying.
Since I am not enough of a development wiz to submit code of my own, I don’t feel qualified to judge the 150+ core Devs about being too slow at releasing a scaling solution… Obviously it isn’t an easy task, and I doubt it ever could be one while Bitcoin stays decentralized and open source.
However, the dangers of a Hard Fork are real, and several Devs have submitted their own HF upgrade proposals… But the 12-18 month “safe rollout” is pretty impossible in this highly combative climate.
No matter what they do, HF = getting caught with our pants down. Plenty of powerful banks and governments would be able to destroy bitcoin from even a slight hard fork if they wanted to.
Is anyone really willing to risk giving them that chance just because we’re impatient with the current scaling solutions?
1/ what is a hard fork?
2/if bitcoin is in competition with fiat money wouldn’t the cia’s cyberwarefare unit have already planted programmers within the bitcoin network in order to make the n/w unstable and to then collapse it?
3/ does ethereum have a fixed amount of coins?
thank you,
1 / – A hard fork is where different groups run different code for bitcoin that is just different enough that the blockchain splits in two and a second Coin (Bitcoin B) can be issued, so all bitcoin holders would then have bitcoin a coins and the same amount of bitcoin b coins. – Like a stock split… The price of both could plummet to $0.
2/ Maybe, but the way the bitcoin Core meritocracy works is for hundreds of eyeballs to review all the code and only the best, most agreed upon code rises to the top. It’s unlikely that their influence can make it through that guantlet.
3/ Ethereum has no ceiling, 18 million per year are released and now that means on two different coins. They’re practically worthless, but the Ethereum network is very useful so it won’t be going away any time soon.
Outstanding and thoughtful analysis, Rick. You were right-on about the market deciding where Bitcoin should go and not by a few self-appointed, centralized protocol developers. Yes, the miners have spoken, and they don’t want segwit. I believe the issue will resolve itself by the end of this year, but what it is really going to take is an economic disinsentive to force the miners to decide. As long as there’re making money, the stalemate will continue.
The synergized win-win solution seems obvious – a 2mb hard fork *with*Segwit. Increases transaction capacity, fixes the malleability issue, and allows sidechains like Lightening Networks to come online. Everyone wins, everyone’s happy, or so you would think. And I have no idea why it hasn’t been proposed.
I started a rescan on my backup node last Friday. I had to upgrade, because my 128GiB SSD couldn’t hold the blockchain. It is quite low power, because I want to keep it running on battery for as long as possible, along with network equipment, when my power go out. Due to a fsckup, I managed to delete the rev* files which I had offloaded to an SD card previously while waiting for my new SSD, so I had to rescan . The node is still rescanning after a week! Why don’t big-blockers want to discuss this problem? I really want to know.
Segwit mitigates it in two important ways.
1. Transaction validation cost less.
2. It prepares for moving transactions off the blockchain (Lightning).
The last point is especially important, because everybody knows, and as this experience shows, bitcoin can’t scale on the chain. Unfortunately the gain for the miners isn’t as obvious. They worry about fewer fee paying on-chain transactions. (I don’t think they have to be worried. Lightning commits will pay higher fees due to the high risk of loss if the transaction isn’t mined in time.)
A hard fork is always possible. Don’t blame anyone else for not doing it! That’s just stupid. It is entirely up to you, if you want to fork hard. The advantage of an hard fork is the fact that you don’t need any cooperation from “old” miners at all. As long as someone are willing to mine your fork, e.g. yourself. you are good to go. There will be two coins, of course. That is inevitable with a hard fork. Only soft forks depend on a majority.
I started a rescan on my backup node last Friday. I had to upgrade, because my 128GiB SSD couldn’t hold the blockchain. It is quite low power, because I want to keep it running on battery for as long as possible, along with network equipment, when my power go out. Due to a fsckup, I managed to delete the rev* files which I had offloaded to an SD card previously while waiting for my new SSD, so I had to rescan . The node is still rescanning after a week! Why don’t big-blockers want to discuss this problem? I really want to know.
Segwit mitigates it in two important ways.
1. Transaction validation cost less.
2. It prepares for moving transactions off the blockchain (Lightning).
The last point is especially important, because everybody knows, and as this experience shows, bitcoin can’t scale on the chain. Unfortunately the gain for the miners isn’t as obvious. They worry about fewer fee paying on-chain transactions. (I don’t think they have to be worried. Lightning commits will pay higher fees due to the high risk of loss if the transaction isn’t mined in time.)
A hard fork is always possible. Don’t blame anyone else for not doing it! It is entirely up to you, if you want to fork hard. The advantage of an hard fork is the fact that you don’t need any cooperation from “old” miners at all. As long as someone are willing to mine your fork, e.g. yourself. you are good to go. There will be two coins, of course. That is inevitable with a hard fork. Only soft forks depend on a majority.
I liked your analogy of the toyota boardroom. I’d make one change to it though; The public clearly wants a car that can only go forward when it is pointed at a cliff.
If the miners only want software that is sure to destroy their livelihoods (and all of ours with it) I’m 100% behind Core devs on not giving it to them. They have to be shown how dangerous what they want is.
Great point!
Wish I’d thought of saying that. 😉
All of the stalling, confusion, mixed messages, and purported blindsided-ness makes sense if you just believe 1 simple fact. That there must be some party that actually sees this as the best, quietest, easiest, simplest, way to kill bitcoin’s dream of being a global currency. With total deniability.
nail(head)=hit
Your post reminds me of where I was in this journey of truth and self-discovery in Dec 2015. When miners agreed that 2mb was safe and we can do it, and Blockstream just totally ignored them and went ahead with their geeky and cooler solution instead. This is why we need more than just one implementation team and roadmap.
Bitcoin devs need to diversify. Or die.
Fantastic report! Not that I was at the roundtable, but I agree with every word of it. Especially the Toyota boardroom analogy. And the “mustn’t take risks” people (deploying no improvements for years is a huge risk!). And the predictability of it all – if only Adam Back had stuck to his 2-4-8 proposal…
My only disagreement is I’m not as optimistic, because although I totally agree that the Bitcoin world is full of amazing people, amazing people are no guarantee of an amazing (or even viable) collaboration. But here’s hoping you’re right.
Looks like Joseph have a point! Thank you for this report of the meeting!
Nice writeup Rick.
As others have stated here, SegWit offers benefits and, if a block size increase to 2MB keeps the other camp happy for now, just get on with it and do both. But a dynamic solution for future increases needs to be found as repeating this conversation for years to come is wasteful of precious, limited resources.
Irrespective of that, people need to let go and not be fearful of forking. If they are fearful it is purely because they can not overcome the selfish influence of their decision making process i.e. the fear that it may cost them personally.
What is natures answer to its own “governance” issue?
To fork, and fork again – making many mistakes along the way – but to ultimately survive.
The problem with that proven model – and what this community is having difficulty in accepting – is that the ultimate benefit will only be felt by our descendants and not by ourselves.
What a bunch of selfish bastards we are.
@Scrumeister
Is is so hard to understand that the moment bitcoin B is issued by an exchange the entire planet’s trust in the currency of bitcoin goes to Zero?
It’s not like ethereum’s hard fork because ethereum has a non-monetary purpose. Bitcoin exists to be MONEY, and money does not work if the people holding it fail to trust in it.
Every single thing used as money in the past, from cowrie shells up through gold and arguably paper, have all worked on the following principle: You’d only accept it if you think the value of it will still be there when you go to spend it.
If bitcoin splits, people holding any will all be thinking: Who will be accepting which exactly? Will merchant Point of Sale machines see the new coin, the old coin, both? Which will be the valueable one? Do I have any control over it whatsoever? No? What a shitty money.
Bitcoin will be sub $1 forever, just from the HF split.
Market =! Miners
That’s the whole problem. Miners have too much power and it can be bought.
They wan’t things that are not good for bitcoin, but might bee good for themselves in short term.
Long term: what is good for bitcoin = good for miners.
“We’ve seen a capacity ceiling approach for two years, and now we’ve hit it”
Backlog is a good thing. I can’t understand what is so bad about it. Yes we need on-chain scaling. No, it’s not urgent.
Yes SegWit is nice. No, it’s not necessary. Immutability FTW!
Market =! Miners
That’s the whole problem. Miners have too much power and it can be bought.
They wan’t things that are not good for bitcoin, but might bee good for themselves in short term.
Long term: what is good for bitcoin = good for miners.
“We’ve seen a capacity ceiling approach for two years, and now we’ve hit it”
Backlog is a good thing. I can’t understand what is so bad about it. Yes we need on-chain scaling. No, it’s not urgent.
Yes SegWit is nice. No, it’s not necessary. Immutability FTW!
What’s the correct term then? It sure ain’t technical either. oke nice
Bitman, Its true, same as what is good for Monero is good for Miners. I proved this while mining Monero at https://www.coinimp.com and a Positive change in Monero would always have a great impact to the miners