I feel the article does Pieter Wuille a disservice by mentioning the Blockstream/sidechains conflict of interest without mentioning the fact that he is actually in favour of raising the block size limit (albeit gradually, over time) and has authored a proposal to do so: https://gist.github.com/sipa/c65665fc360ca7a176a6
What's ironic about this whole affair is that four of the five core developers support raising the blocksize limit (Gregory Maxwell is the sole dissenter, preferring to wait until a transaction backlog develops before taking action) but they diverge on exactly how and how quickly to raise it.
I honestly feel that some of the "pro raising" devs aren't actually in favor of raising it anytime in the foreseeable future, and are just dragging their feet in the hopes of that the whole thing will go away.
The whole point of this debate is that we've been litigating this affair for the past three years, each year a several core devs claiming to want to raise it "soon", and then it just gets put down another year.
I'm still not sure about the BitcoinXT gaining consensus, but I'm very glad someone is putting real pressure on the situation, because that's the only way the block size will be raised anytime in the next few years.
My take on it is that the key difference is that Gavin perceives a greater urgency in raising the limit before we reach the point where blocks start getting filled up, than Jeff/Wladimir/Pieter do.
I think he got frustrated that no progress was being made, which sparked his decision to add the relevant code to XT, which is what has brought things to a head. I actually suspect that his preferred outcome is not to fork the blockchain into XT and Core, but to put some pressure on the other core devs to act.
I'm still not sure how I feel about that patch, but to be clear it does not blacklist Tor by default. What it does is that if a particular node detects a DOS attack that originates from the Tor network, each node with then down rate the Tor IPs (since that's where many of the DOSs originate from), giving non-Tor nodes higher precedence. And thus avoiding the DOS attach for non-Tor nodes.
Under normal conditions, Tor nodes are fully integrated into the network.
Like I said, I'm still ambivalent about this patch (for multiple reasons), but saying blacklisting Tor by default might give people the wrong impression of what it does.
In the words of BitcoinXT site [1]:
"This patch set introduces code that runs when a node is full and otherwise could not accept new connections. It labels and prioritises connections according to lists of IP ranges: if a high priority IP address connects and the node is full, it will disconnect a lower priority connection to make room. Currently Tor exits are labelled as being lower priority than regular IP addresses, as jamming attacks via Tor have been observed, and most users/merchants don't use it. In normal operation this new code will never run. If someone performs a DoS attack via Tor, then legitimate Tor users will get the existing behaviour of being unable to connect, but mobile and home users will still be able to use the network without disruption."
I've looked at the patch, and that's pretty much what it does [2].
That would not be correct entirely. Earlier they had 90% as the consensus limit, which they have now brought down to 75% and have suggested driving consensus in regular milestones if that doesn't succeed. Surely they want the fork.
Am I missing something or is that a rather weak argument? If the hashers in that large mining pool don't like the veto, they'll vote with their feet and leave the pool. But if the pool actually has 25% or more of the hashing power it is a clear sign of no consensus. It seems to me that the operator of a pool is not really in control of the computing power.
Combined, the Chinese mining pools have well over 25% of the hashrate using only in-house hashing. I think this was the political way of denying a veto to the Chinese mining operations (who frequently collaborate and will likely vote in block).
>..if the pool actually has 25% or more of the hashing power it is a clear sign of no consensus.
Gavin clearly agrees.
However, I guess he doesn't think that a pool (or a miner like BitFury) that has between 10% and 25% of the hashing power should have the power of veto.
I totally agree with that part of the conclusion that says Bitcoin has failed in regards of it's "no need to trust anyone" goal. It's not a failure for us though. There are so many things we can learn from Bitcoin, like no matter what your technical solution is, you'll always have power bubbles (e.g., for bitcoin the power aligns around the developers and the huge farm owners), you'll always have disagreement in the community no matter how morally high your goals are.
> There are so many things we can learn from Bitcoin, like no matter what your technical solution is, you'll always have power bubbles
If many of Bitcoin's believers had only been less dismissive of the social sciences, they could have known that long ago. In fact, if they had only been less dismissive of politics, they would have known that people who dismiss politics are simply people who don't know what it is, and those are the people who eventually do bad politics.
Politics is a property of any social interaction and is simply influence over others (by any social means -- even technological). As an academic field of study it is the study of the flow of power (simply put -- influence) in a society.
I wouldn't lump all us Bitcoiners together. It was very clear right from the start that Bitcoin is just a tool that lets people rearrange and disperse power, not a magic cure-all (even though some people get rather breathless over The Blockchain™).
If you look at the current dispute, it looks somewhat like what an election campaign for central bankers might look like if western countries were to allow for such things. Or rather what the first one might look like, if the current set of central bankers weren't really on board with the whole idea. You've got different groups with differing visions running something that vaguely resembles an election, but the issues are rather wonkish and technical.
Undoubtably many people will look at this whole thing and say, look how much more professional the dollar or the euro is. You never see such disagreements in the Fed! But I think the social failures that are leading to this crisis are actually also present there too: an assumption that experts never disagree, a preference for opaque unanimity, and a strong mental/emotional link between (perceived) unanimity and credibility.
Put another way, it's difficult to imagine a new central banker announcing that his/her predecessors and peers were all wrong and - for example - the correct inflation rate to target is 0% and not 2%. Even though there are economic papers out there that imply their policies might not be correct (written by economists at central banks, no less), the giant shitstorm that'd unleash would be extremely un-Central-Banker like and be seen as a serious threat to the credibility of that CB or even that country.
Bitcoin, for all its faults, does at least have some kind of mechanism to handle the case where the people making boring technical decisions stop doing it right.
> If you look at the current dispute, it looks somewhat like what an election campaign for central bankers might look like if western countries were to allow for such things.
But they do (indirectly, of course)! Conformity of thought arises not because of the rules of the democratic system, but because of the way power tends to get concentrated in homogenous groups (another thing the social sciences tell us), unless resisted by some civil action.
> Bitcoin, for all its faults, does at least have some kind of mechanism to handle the case where the people making boring technical decisions stop doing it right.
I don't think that the Bitcoin system has any such advantage. The only difference I see is that Bitcoin's leadership are not accountable (even nominally) to any collective social interests that go beyond technical concerns regarding the currency. Whether that's good or bad depends on your ideology.
You cannot fix the bugs in the workings of a system that on paper seems like it should work (i.e. modern democracy) without understanding what the bugs are and how they're created (and even then it's very hard, and entails a lot of politics, meaning influencing people), and you can't learn about those bugs in the mechanism if you dismiss the academic disciplines that have studied it for decades to centuries (depending on the discipline). You certainly can't do all that without an appreciation for politics.
In all western democracies that I'm aware of, the central bankers are meant to be independent of politicians. Of course you can view this as a fiction, but all the players do seem to believe it and it's exceptionally rare for politicians to directly argue with, let alone override, central bankers.
> I don't think that the Bitcoin system has any such advantage. The only difference I see is that Bitcoin's leadership are not accountable (even nominally) to any collective social interests that go beyond technical concerns regarding the currency.
Of course they are. You don't see developers releasing forks of Bitcoin that try to bring about economic outcomes, because such a fork would be almost certainly ignored by the userbase. But that doesn't mean it's impossible, or that it'd never happen even if the users were collectively strongly in favour.
Besides, the two can't be easily divided. The block size debate is a bizarre witches brew of technical concerns, pseudo-technical concerns and social/economic concerns. The line between them is often blurry. So even a fork implementing a very minor technical change is causing massive uproar. Some people just really like the idea of a central planning committee of experts. Regardless, that's not a technical limitation of the protocol. In theory the fork can work.
> In all western democracies that I'm aware of, the central bankers are meant to be independent of politicians.
What do you mean? They are appointed (and reappointed or not) by politicians.
> But that doesn't mean it's impossible, or that it'd never happen even if the users were collectively strongly in favour.
The difference is that in a state these things are done as part of bargains, which help prioritize. For example "I'll support your nomination for the chair of the central bank if you support my law allowing same-sex marriage". Sometimes these bargains can take the form of hostage-taking and be annoying, e.g. "If you don't repeal the ACA, I won't vote to raise the debt ceiling", but that's all part of the same process of keeping people in control. That bargaining, BTW, happens regardless of the political system (i.e. even in autocracies), only in a less explicit way and sometimes with different people involved.
Once you completely separate the power structure for the currency from other concerns, you lose that control, and you lose it in a very radical way. Thing is, it's harder to bargain with people with money because they are relatively immune to many social effects. However, one of the classical checks-and-balances mechanisms in history is that of money and violence (and, at some point, religion). The two are both forms of power (i.e. ability to influence), but unlike money, violence is much harder to concentrate in the hands of the few (technology is changing that, but there are other, subtler sanctions at play that resist total concentration of violence). So the rich could starve the poor and had horses and swords, but they knew that if they went too far, they couldn't withstand an uprising.
The state made an interesting deal: a central authority (though ideally not concentrated -- the two are different concepts) would have a monopoly over violence (up to a point), and, in exchange, it would regulate money. It wouldn't let the poor kill the rich (again, up to a point), but, through regulation it would make sure that the rich don't starve the poor. Once you separate the two, and, in the process, make even extreme violence less disruptive, you break a very important balance that has been reached. Like some sort of a MAD doctrine, you want to keep a mutual, implicit threat going.
Which brings me back to the first point about central bankers. Even after their appointment, and even if they're not interested in another term, they still have a responsibility toward their community -- as I hinted at before, democracy only formalizes certain rules that have always implicitly had an effect, and certain sanctions are always at play even if not made formal by the democratic system. Bitcoin breaks that, too.
> If many of Bitcoin's believers had only been less dismissive [...]
You have really put your finger on something for me. If I had to pick one word to describe the bulk of the Bitcoin advocates I've dealt with, it would be dismissive. Banking, social sciences, politics, history, economics, design, user experience, even a lot of technology.
Years ago I did tech for market makers and dealt a fair bit directly with exchanges, so naturally I read up on the history of places like Wall Street and the Chicago Mercantile Exchange. The technology was mostly uninteresting; things had run fine for decades with shouting and slips of paper. The impressive part was the elaborate social system, the set of subtly interlocking roles that made it all go.
I never met a Bitcoin advocate who seemed to understand any of that. Or even that understood that there was something to understand. So much of what I heard and read was in the true-believer vein, as if someone had crossbred gold bugs with partisan gadget fanboys. That's not to say that more responsible thinkers don't also like Bitcoin, just that I haven't managed to find any of them talking about the stuff I knew from experience was important.
I still hope people find something useful to do with blockchains, as they're a really cool technology. But it seems to me that if one has a solution looking for a problem, then actually understanding problems would be a valuable step.
Yes, this comment is exactly my experience with Bitcoin advocates, too.
I often post the exact same critique of Bitcoin in these stories, that there's no political accountability in the Bitcoin technology to its users.
Bitcoin, of course, has economic assumptions baked in that influence how people use it, but those people have no real recourse to effect change in those economic assumptions. So its users, unlike with any national currency, have to take what they are given, without ever being able to have a say in how the currency is run.
This is antidemocratic at its core, of course, but the near-ubiquitous response I get is, "sure it's democratic, you just need to fork it and then persuade 51% of the network to use your fork," and occasionally someone comes out and says "people aren't fit to make economic decisions, it's right that Bitcoin makes those decisions for them."
Never has a single Bitcoin advocate ever said, "you're right, there should be some method in place for Bitcoin users to have a say in how it operates." Not once.
I haven't seen any other community so willing to either take power (for those who think of themselves as the techno-elite) or give it away (for those who recognize they aren't).
It reminds me of Terry Pratchett's running gag about how the people of Ankh-Morpork want kings back.
Absolutely. The whole public story is "decentralized techno-magic freedom money with no nasty governments!" But as was pointed out long ago in "The Tyranny of Structurelessness", a lack of visible structure just means the true power structure ends up implicit. Right now it's fascinating to watch nominally technical decisions being made based on what mysterious Chinese mining organizations will think.
I can appreciate the notion of having machines implement the transparently-published rules humans decide on as a way of increasing market trust. I'm not sure if it will work, as expressing rules sufficient to operate economies is very hard. But regardless, I'd like to see some rules about the humans who make the rules. Rule by law has been working out pretty well the last couple centuries.
This really is the takeaway. Bitcoiners like to point to the remittance industry or the wider money transfer industry and say how a bitcoin future would be so much better, but they're slowly learning the hard way a lot of problems actually need solving, and the cost with bitcoin is just as high as non-bitcoin.
"Solutions" start to appear which look remarkably similar to the existing industries they were trying to disrupt. For example "off chain" transactions which seem w.r.t bitcoin remarkably similar to fiat banking with respect to gold.
Yeah and in theory we all know that, right? But especially with Bitcoin I totally forgot about it, although I'm probably far less enthusiastic than most Bitcoiners (don't read much about it, only own $2 worth of Satoshis, etc). For some time I really thought Bitcoin would solve money. And that's where I think the real learning comes in, when you transfer your theoretical knowledge and see how it applies to reality.
I'm sorry, I don't understand. Who, in this situation, has exercised power over me, the Bitcoin user? As far as I'm aware my Bitcoins at safe and I am still able to transact them with anyone who has an Internet connection and a little time. Without interference from any governments or banks or what have you. Isn't that the Bitcoin contract? What exactly has failed here, due to a power bubble?
Bitcoin's problem is that it pretends to be distributed when it's really decentralized. Or vise versa, depending on your definition of those terms. What I mean is that Bitcoin is a distributed way of building a global variable (the blockchain). As long as we go on thinking a global blockchain is any different from centralized banking... I'm more interested in what happens when we throw out the idea of commodity money altogether: http://thenewstack.io/why-art-could-become-currency-in-a-cry...
This is very good news (that there is a fight). Open sourced currency that could be forked at any moment is a feature not a bug.
The fact that government currencies can't be forked, and thus never guarantee they stay competitive, is the main reason a bitcoin like currency is nearly certainly the future.
The downside is some medium/long term uncertainty in that exact currency. However this is a NECESSARY component.
I've tended to agree. Bitcoin is really just an early iteration of the cryptocurrency concept, and many of its "failures" are, in the long run, successes. This is why the "altcoin" ecosystem is so vital; they're doing a lot of the fundamental experimentation that's necessary for cryptocurrency in general to approach perfection.
Unfortunately, a lot of people fail to realize this, and champion Bitcoin specifically as the be-all-end-all of electronic money despite the sheer number of problems with it. This attitude, while understandable, is regrettable.
Bitcoin XT also has other changes decided solely by their two devs without fanfare or discussion; for instance, it disfavors Tor nodes and hampers anonymity.
This is not widely reported but is indicative of what we can expect from such a BDFL model. A BDFL is the end of Bitcoin.
The reason to oppose XT has nothing to do with blocksize and everything to do with control.
Ah, sorry. Even the article I linked is potentially stale. Things seem to be moving quickly. The "near majority" statement was based on unconfirmed reports that another large pool has hopped onboard: https://twitter.com/mkomaransky/status/636534235508178945
The Bitcoin core has definitely left the stage of rapid innovation. Instead, we can expect slow, consensus-based progress at best for the future. This is normal for a successful protocol and happened to html, smpt, tcp, etc. as well. Committees have taken over.
The Bitcoin core has definitely left the stage of rapid innovation. Instead, we can expect slow, consensus-based progress at best for the future. This is normal for a successful protocol
But that alone isn't indicative of the protocol's success, is it, can't that also be attributed to infighting?
Committees may propose things, but they've not really taken over as it's still the community(those controlling the networks computational power) that decides what changes to adopt right? And the only people who can actually implement changes are those with access to the core right? Or former core people who've splintered off(XT.) So what do committees have to do with this?
Sidechains are a kind of complex mess. After reading about them I'd definitely chalk them down to terrible, committee based software development vs. just bumping the size like the XT people want to do.
Suppose a client and server were connected with a connection running with speed of light. Even then, there is some finite limit as to how many bits can be transferred. How do we handle scalability beyond it then? By abstraction. Rather than sending raw data we only send what is relevant & decode it at either ends.
Now transactions are raw data that flows in the bitcoin network, but there will always remain a practical limit. Throughput of the network is number of transactions per unit of time it can handle, which translates to number of blocks per unit of time.
To increase the value of this equation, 3 things can be done: 1.increase block size 2.Decrease block time 3.Abstraction
The problem is that XT side considers increasing blocksize is the only way. The core side has presented sidechains as one of the methods of abstraction, but since they invented it they formed a company out of it and now it appears as conflict of interest. Anyways, XT side is painting the Core side as anti scaling, which is totally untrue, but since sidechains are not even being heard, they are inviting other proposals to be presented at the workshop in Montreal on 12th Sept.
What I don't get is how quickly the debate started and then dissolved with a potential hard fork as the outcome.
What was the urgency to make a change? Yes, Bitcoin couldn't scale as it was, but there wasn't even close to enough demand for Bitcoin transaction for the topic to have escalated so quickly.
Are Bitcoin people so delusional about how popular Bitcoin is or will be, that they thought the amount of transactions that could be put into a block was of such dire importance?
"Suppose a client and server were connected with a connection running with speed of light. Even then, there is some finite limit as to how many bits can be transferred. How do we handle scalability beyond it then? By abstraction."
Also by additional (independent) connections and parallel transfer/processing of those bits. Personally I think that abstraction (and especially its levels) should be governed by the law of diminishing returns, not by hard limits. The hard limits must be dealt with through other means.
By stating that, I meant there is some limit to hard limits always. And bitcoin is unlike traditional networking; better network does not directly affect scaling.
What's ironic about this whole affair is that four of the five core developers support raising the blocksize limit (Gregory Maxwell is the sole dissenter, preferring to wait until a transaction backlog develops before taking action) but they diverge on exactly how and how quickly to raise it.