This post dismisses lightning.network without good reason for doing so.
Instead, it rushes ahead with an implementation that actually employs exponentially increasing block sizes.
There’s no reason for this level of drama, and the solution they chose seems to be a poorly thought out one given the existence of alternatives like Lightning, Sidechains, and Dynamic Blocksize Limit.
More to the point, their actions seem less consistent with an intention to actually solve the problem in an elegant manner, and more consistent with that of an attempted governance coup  .
There’s nothing necessarily wrong with a governance “coup” of Bitcoin (maybe Bitcoin needs a coup, maybe it doesn’t), but doing so under the pretense of a technical issue, when Bitcoins are at stake, is lame, misleading and reckless.
Technical issues are best solved with technical solutions, and preferably elegant ones that work. Of all the options available, an exponentially increasing block size doesn’t even make the list.
That’s not true, he addresses it specifically and links to a more fleshed-out article he wrote on the topic in May. Quote with emphasis added:
The so-called “Lightning network” that is being pushed as an alternative to Satoshi’s design does not exist. The paper describing it was only published earlier this year. If implemented, it would represent a vast departure from the Bitcoin we all know and love. To pick just one difference amongst many, Bitcoin addresses wouldn’t work. What they’d be replaced with has not been worked out (because nobody knows). There are many other surprising gotchas, which I published an article about. It’s deeply unclear that whatever is finally produced would be better than the Bitcoin we have now.
It doesn’t seem like he’s completely opposed to a solution like the Lightning network, it’s more that he doesn’t believe that raising the block sizes is mutually exclusive with the implementation of an off-blockchain payment channel solution. The biggest complaint is that it’s not plausible that such a solution will be ready within a reasonable timeline, compared to raising the block size which is addressing a relatively imminent concern.
It doesn’t seem like he’s completely opposed to a solution like the Lightning network, it’s more that he doesn’t believe that raising the block sizes is mutually exclusive with the implementation of an off-blockchain payment channel solution.
Your interpretation of what he is saying is accurate. It would be nice if that’s what he was doing.
Raising block sizes by some amount is uncontroversial. Everyone is pretty much on board with that, including all of the core devs that I’m aware of.
Mike’s “solution” is not a solution but a massive problem for Bitcoin. At the risk of repeating myself: a hard fork, done in this way, will cause more problems than it fixes (Bitcoin will become way more centralized, the value of BTC may go down, people may lose coins, and the entire community will look even more ridiculous than it does now). And, Mike will have to change his code. That is not a maybe, it is a mathematical guarantee. Exponential growth must be stopped at some point, and an existing BIP proposal does that, but blockchain scaling should be more intelligent. Rushing is foolish.
The biggest complaint is that it’s not plausible that such a solution will be ready within a reasonable timeline, compared to raising the block size which is addressing a relatively imminent concern.
Again: it is not an imminent concern that requires him to create this broken hard fork and split the community.
Fear not, I’m already familiar with all of your cited sources and how the network works.
I get that’s not what you think he’s doing, but he may not be doing the thing you described either. The reality is that the discussion for introducing larger blocks has been on-going for a very long time. It started long before the average block size started creeping up near the limit. Unfortunately, despite numerous proposals, a consensus has not been reached and thus the bitcoin-core repo is stalled on the matter (whether that stall due to lack of consensus or ulterior motives or disagreement in social psychology is another question).
Hearn’s fork is a forcing function. It’s like when Google releases Chrome to create a pressure force on the stagnating browser features of the incumbents. It was never intended to become the top browser. It was intended to illicit a reaction from the competitors, which it did very efficiently.
I do not believe that he’s expecting to win the majority. If I were him, I would be expecting to force a decision by the bitcoin-core maintainers. A decision that I’d expect would be a more conservative but acceptable bump to the block sizes.
Yes, that thought did cross my mind as well.
That would be the most positive spin on what Mike is doing (that I can think of).
The thing is, even if that’s his plan, perfectly reasonable people can disagree whether or not that’s the best course of action.
I for one, don’t think that this method of forcing “progress” is the best one he had available to him, for a few reasons:
FUD and drama is bad enough, but actually creating the potential for people to lose Bitcoins is a whole other ballgame. He’s crossed a very real line at this point.
Starting to sound like FUD here too. :/
I don’t know why you say “starting”, as I said that in my original post. That is something that can happen during a hard fork of Bitcoin (and is more likely, I think, during a contentious one like this).  
It also happens in centralized systems (PayPal, VISA) because of censorship. The whole concern with large block sizes is centralization, and when the limit is increasing exponentially you will get block sizes that only PayPal and VISA can handle.
I’m a total outsider, but he mentions that lightning has yet to be fully worked out, and says that those familiar with capacity planning in this domain are strongly urging rapid action. I don’t know how relatively hacky the alternatives are, but in the short summary you linked to it seems that none of them are without drawbacks.
According to him, SOME action is required in the short-term to stay alive. How much relative support do the approaches you mentioned have, and how soon could they be implemented to prevent short-term network breakdown?
According to him, SOME action is required in the short-term to stay alive.
For any common definition of “short-term”, that is complete and utter nonsense (read the “no reason for this level of drama” link).
How much relative support do the approaches you mentioned have, and how soon could they be implemented to prevent short-term network breakdown?
There is no short-term network breakdown to worry about.
Lightning.network is nearly finished (so I hear). Even if it weren’t, it’s not like it’s some sort of massive undertaking. I don’t know as much about the other proposals.
Anyway, if this was actually an urgent issue with real time pressure the Bitcoin devs would release a hack to increase the block size, and it wouldn’t be Bitcoin XT. Exponentially increasing block sizes are not the answer to this problem, they are a problem in of themselves.
I have a really simple “what” question which I haven’t seen answered by any of the many “why” posts…
Is this copying existing account balances, or starting from zero? If it’s copying them, as of what moment in time - has it already occurred, or is it in the future?
There are a lot of followup questions about currency deflation, financial incentives to do creative things, etc, but if it’s starting from zero, they don’t apply, of course…
By fork, it means that—if accepted by the 75% super-majority—it’s going to “split” from the existing ledger. At that point, balances on both forks would be identical but any transactions on the respective blockchain will diverge henceforth.
If the fork goes through, then anyone still using the old blockchain will be getting data that is no longer accepted by the new blockchain.
Aside: The Bitcoin XT client is structured in a clever way such that if you do upgrade to using Bitcoin XT but the fork does not pass, then it continues to be in-sync with the original blockchain and there is no interruption. The interruption scenario only happens if you stay on the Bitcoin-core client.
Thanks. That’s useful detail.
So theoretically, for example, if both currencies wind up coexisting for some period of time, everyone who held BTC now also holds $new_currency. The value of a currency is in what can be bought with it, and the value of BTC isn’t going to disappear overnight; even in the 75%-acceptance case, there are going to be both individuals and merchants who either don’t understand how to migrate, or have chosen not to.
If BTC eventually dies in favor of $new_currency, that’s going to be “interesting” for everyone involved in BTC transactions post-fork. The mere possibility of it seems likely to trigger a panic, and possibly hyperinflation, doesn’t it? Obviously nobody who stops transacting BTC at the moment of the fork has a huge risk there, since they have $new_currency to fall back on, and they might reasonably assume it’ll be largely unaffected by whatever happens to BTC. I don’t see how one can double the total amount of money everyone has, and not have SOME destabilizing effect, and it’s honestly anyone’s guess what it’ll be.
If hyperinflation does happen, and goes fast enough, maybe nobody will set up a currency exchange, which might indeed largely isolate the two currencies from each other. That seems like a best-case scenario, and honestly not a pleasant one to have even the slightest involvement with.
I guess I don’t have a question; that all seems like it’s entirely intended, and does clarify why this is so acrimonious. People who hold or transact in BTC should probably be making contingency plans now.
Bitcoin XT is not really a new currency. Not in the same way an “altcoin” like Litecoin would be (which starts from a fresh blockchain). Bitcoin XT is actually taking advantage of bitcoin’s built-in upgrade mechanism.
While there is always a possibility that any kind of change (or even non-change) will trigger a panic reaction in bitcoin-land (that’s part of the charm of this lovely endless source of drama, right?) it’s unlikely that Bitcoin XT will have a volatile transition if it ends up “winning”. It’s really up to the miners to decide which blockchain to adopt, and to make the incorrect decision means allocating some amount of time mining a worthless blockchain. If Bitcoin XT starts to get substantial adoption, and a critical mass of miners start switching to Bitcoin XT, then the rest of the miners are very likely to follow suit.
It’s very unlikely that there will be any kind of un-extreme split between miners supporting two different blockchains for any length of time. For this to happen from an economic perspective, the timing and cumulative decisions of everyone involved would have to coincide such that both chains are valued approximately equal to each other with the same number of supporters (note that this can’t happen since 75% is required for this upgrade). Only then will it not make sense to switch to one side or the other, which would be an extremely fascinating social and economic scenario to observe the outcome of. The moment the tides tilt one way or the other, it becomes blatantly rational to quickly migrate to the new winning blockchain (since the other blockchain is going to become obsolete).
This is not unlike the situation if we were to cut the internet connection between two perfect halves of the world, and then later reconcile the connections. For a while, there would be two alternate versions of bitcoin, but quickly one will reach critical mass and win.
If anyone stays on the old blockchain while all miners have migrated to the new blockchain, then the old-blockchain users will not achieve any transactions going through. They’ll never get mined into a new block. Even if some minority miners stay in the old blockchain, the difficulty level would remain being so high that mining a new block will take forever.
I understand what you (and apparently the Bitcoin XT community in general) are hoping for. Reiterating that anything other than the exact scenario you envision is “very unlikely” is not very reassuring.
I wouldn’t say I’m particularly hoping for anything. Either outcome is going to encounter a novel scenario with unknown outcomes.
If Bitcoin XT is not adopted and there’s a sudden rally in usage/price, then the maximum block size will get hit hard and transaction fees will start to climb. Meanwhile, an increasing subset of transactions will sit around in the pool of uncomfirmed transactions. Maybe that pool will keep growing disproportionally, maybe economic forces will play out properly, or maybe there is a bug in the miner transaction prioritization heuristic and the network will get DoS’d.
In order for bitcoin to survive long-term, it needs to be able to handle both kinds of outcomes. Probably better this happens now than someday in an optimistic hypothetical future where even more money is at stake.
That’s pretty fair. Thinking about it that way, you’re right, it’s an important test.