Why the surprise?  Blockstream naturally is using every means that they can to kill Classic and retain control of the protocol.  If FUD doesn't suffice, there are bribes, threats, intrigue, etc.

Why would the big holders want to attack one of the chains?  Their holdings are in both chains, and they can keep, move and sell them independently.   Whatever value each branch of the coin has, attacking one branch will kill its value -- which will hurt the holders more than anyone else -- but is unlikely to raise the value of the other branch by the same amount.  

The holders should *pray* that a proposed hard fork will EITHER fail quickly to gather any support, OR quickly achieve majority support and end with a clean non-eventful hard fork.  Any fork attempt that does not resolve cleanly in one of these two ways can only harm the value of their holdings.

If the worst happens, and there is a disputed fork attempt that ends with a coin split (which, in a hard fork with 75% trigger, is VERY unlikely to happen), the best strategy for holders is to keep quiet and wait for the market to define the values of the two branches.  If one of them is quickly dropping to zero, they should dump what they can of it, and then just use the other -- in which case the situation will be the same as after a clean successful fork or clean failed fork, only with bombed buildings all over the place.  If, by some miracle, both branches retain some value, and they seem fairly stable, the holders should keep and use both, or SLOWLY sell the one they don't like to buy more of the other.  

The ONLY bitcoiners who want Core to be the only implementation are the Blockstream people, because their business plans require them having control over the protcol (to add changes that they need, like SegWit, but block changes that their competitors want, like raising the limit).   For all other players, a switch to Classic would be indifferent or better.


Is that right? AFAIK BitFury is based in Ukraine and Georgia, but may have mines in Iceland or other places.

Perhaps the reporter thought that Georgia was the US state?  (Frankly, the US should get rid of the latter.  Their students have enough trouble already with Iran/Iraq, Sweden/Switzerland, Austria/Australia...)

It was no mistake. The concept of a "full but non-mining" relay node is an aberration.  Such nodes are precisely the sort of "trusted intermediaries" that bitcoin was supposed to get rid of.  They do not add anything to the security of the network -- on the contrary, they are a gross violation of the protocol, and a huge security risk.  

What Satoshi called "nodes" were the miners.  Miners were supposed to mine for their own benefit (as users of the network) and for the fees paid by simple clients, who should contact them directly.  Miners have incentives to validate and secure transactions on the blockchain, and serve all clients that pay enough to offset their marginal cost of doing so.  Miners are prevented from defrauding the system by the proof-of-work and most-worked-chain-wins rules.  In constrast, the "full but non-mining" nodes have no financial incentives to be honest, so one shoudl wonder what is their motivation for offering to "help" clients.  There is no way of checking that they actually store or verify anything, so nothing prevents an attacker from cloning thousands of such nodes -- and scamming all clients that happen to talk to only those clones.

"Full but non-mining" relay nodes were invented by bitcoiners who were unable to compete with industrial miners, but still wanted to retain their control over the system, because they felt entitled to it.  Non-mining clients should avoid them, and talk directly to miners (or relays that they cantrust are operated by miners).  The sooner they disappear, the better.

OK.  But note that the Core devs cannot "fire the miners", just as I cannot fire the Core devs or the miners.  They have no power to stop or hamper the CartelCoin, nor the OldCoin. All they can do is what I can do create the CoreCoin (or StolfiCoin) and hope that people will use it.  They cannot even ask people to choose between the two: each client will start with the same coins on each branch, and can move them independently.

And now for something completely different: something tenously related to trading



Note how the volume pattern changed on OKCoin after the last drop. Before the drop you could see the day/night rythm proper of China's time zone. After the drop there is nearly continuous activity, around the clock. Robot traders frantically proping up the price is not a far-fetched possibility...

The volume pattern before the drop itself was different than what I remembered from 2014.  Then it was smother, more sinusoidal.  The one you see on the image above has strong daily peaks that decay abruptly. Maybe some bitcoin ponzi in China updates his webpage once a day?

Depends on what?  If the hashpower is not important, and preserving the PoW is not important, then what is the magic factor that would cause all the value to go to the CoreCoin branch instead of the CartelCoin branch? 

It is amazing how people can get such different readings from the same words.  (Just yesterday I was arguing with someone who, like many libertarian bitcoiners believed that the genesis block quote was proof that Satoshi was libertarian and designed bitcoin to get rid of banks and "fiat" money.

Yes, Satoshi obviously designed bitcoin on purpose to have a finite total issuance.  That was a natural mistake, because he  -- like almost everybody, including myself until 2 years ago -- believed that inflation was a bad thing, and that good money therefore should be inflation-free.

But there is a Grand Canyon between that and "was well aware that Bitcoin not only would be a speculative asset but that it was absolutely expected and even required that it would function that way".  I have read somewhere that it was Hal Finney, in fact, who first thought of bitcoin as investment; and Satoshi's reaction was like "oh, yes, you may want to keep some".  

Note that, in the quote above, he is referring to price increasing as the result of growth of demand, not of speculation itself.  In another quote he tries to estimate the growth of the user base, and says that 20% per year (double every 4 years) would be a "crazy  rate of growth.  By that "crazy"  rate, the price of 0.05 USD/BTC in mid-2010 would be ~0.50 USD/BTC now.  

While he does not say that explicitly, that "crazy" price increase rate would be also consistent with his choice for the parameters of the block reward fomula (halving every 4 years).  If the price increase was less than "crazy", say 10% per year, the value of the block reward too would decrease 10% per year, on average; forcing a gradual transition of the mining revenue from block reward to transaction fees.

Instead, speculative trading cause the price to increase 10x per year over the next 4 years, so that the value of the block reward increase instead of decreasing.  That made mining into an industrial activity, which was inevitably  concentrated (a development that he, and any rational person, would have seen as disastrous for the project).  That level of speculative trading also caused huge volatility, that made bitcoin unsuitable as a currency.  

It is absurdly delusional to claim that Satoshi had foreseen these developments, and that they were part of his goal.  The world did not need another speculative instrument; he would not have spent 3 years designing a crazy one.  There is no sense in having naive investors give 1 million dollars each day to secure a payment system used by less than a million people.  

EDIT: typo "5 years  --> "4 years"

He was proposing Classic to go with CPU mining and leave the 1000 PH/s mining industry to Core.  At this point, I don't know whether he was trolling, making some statement, hoping to trick the Classic devs, calling bluff on their commitment to open governance -- or genuinely believing that they would like and accept the idea.


Was it you that I debated with, 1-2 years ago, about the power of miners to change the rules?

Anyway, I know that this move ("red button"? "poison pill"? "hash seppuku"?) was the official defense against an "abusive" mining majority.  Of course it can be done: anyone can create a modified version of Core and make suitable incompatible changes to it, thus causing a hard fork and a coin split.  And -- funny thing -- it would create a "bitcoin" that I could like, with tens of thousands of independent miners, with not too different hash power.

What is delusional is the belief that the community would follow them, consider their altcoin to be "the" bitcoin, and discard the coins that they have in the miners' branch -- so that their altcoin will retain the pre-fork price of bitcoin, while the miners' coins will become worthless.

It is like the Captain planning to "defeat" a sailors mutiny by rowing out with the First Mate in a small lifeboat, and declaring it to be the real ship.

For one thing, some of the largest mining pools are closely connected to the largest bitcoin exchanges in the world; so, if a majority mining cartel decides to force a fork, these exchanges (and their customers) will use and service the miners' branch.

If that is a serious option, why don't they do it now, and remove the risk of miner concentration?

The delusion is in thinking "bitcoin is where we, the Core devs" are. 

I could do that myself now: make a copy of core, change the PoW formula, and presto: a hard fork happened, and the coin has split.  If the 1000 PH/s is not important, what is the difference between my bitcoin and their bitcoin?

It is not clear whether the USD (or BTC) volume of the settlement traffic will be 100% of the USD (or BTC) volume of LN payments.  IIUC, the LN will not only aggregate payments but will also cancel opposite or circular payments.  Or maybe not, not sure.

Settlements between banks have lots of cancellation, so they are much less than the total volume of interbank bank wires issued by clients.  It would be a shame if the LN cannot not do that

The personalities in "bitcon space" are fascinating, quite unlike those of other communities I have known in the past 30+ years of forum surfing.  Most in a negative way, unfortunately.  People like Mark Karpelès, Danny Brewster, the BFL gang, the Bitcoin Foundation, and many other far too numerous to list.  

I cannot avoid taking sides in the block size limit war.  I am not exactly a fan of Mike and Gavin, but on the other hand I cannot find anything good to say about the Blocstream guys.  Not about character, not about ethics, not about respect for the project and its users, not even about competence.  Bankruptcy is the least they deserve.

A couple days ago, Luke-Jr trolled BitcoinClassic by posting a pull request that would change the PoW algorithm to make it impossible to mine with ASICs.  I then proposed that the Core devs do that on Core.  It was meant to be a just counter-trolling.  But it seems that Greg and Luke-Jr are seriously dreaming the option of doing just that in case the miners switch to Classic.  It is Idiocracy II -- but with bitcoin!

The LN would want to disable bitcoin settlements for the same general reasons that the US government decided to end the convertibility of the US dollar to gold or silver.

Namely, tying LN payments to bitcoin imposes many limitations on hubs and clients.  For example, a big deterrent to LN adoption is the requirement that a new user locks in advance into each new channel an amount of bitcoins sufficient to cover his payments for the next 100 (or next 10'000, according to Adam) payments that she expects to make through that channel.  VISA, on the other hand, gives a bunch of credit to the new client, so she does not have to deposit anything upfront.  The hubs cannot do the same because they would have to lock real bitcoins, not simply let her pay on credit.  if the LN could decouple from bitcoin, it could let hubs create money by credit, like banks do...

Also, one big problem that the LN is still trying to solve is what happens if the bitcoin network suffers a spam attack that delays a fraction of the legit traffic for days.  That could cause many channel settlement transactions to be delayed past the channel timeouts, effectively doing a permanent chargeback on every payment that was made through those channels. And/or it would force the receiving parties of those payments to issue extra transactions in an attempt to push their settlement transactions through with CPFP: and those attempts would eat into their profits, would make the backlog worse, and would be guaranteed to fail in a large fraction of the cases, whatever the fees they used. Again, decoupling the LN from the bictoin system would eliminate that and may other problems...


My reading of history is that Greg Maxwell got pathologically obsessed with the fee market idea years ago, and won't give up on it, no matter what others say.  Some Blockstream employees, especially the guy who is working on LN, even said publicly that the LN would require increasing the limit; but Greg does not seem to be impressed.  That may be the reason why Blockstream has committed to the disgusting SegWit hack, that will provide some relief (especially to the LN, that may require humongous signatures) without touching the sacred 1 MB limit.

(There is an exchange in this forum, some years ago, where Greg says that the 1 MB limit is as sacred as the 21 M BTC limit, and he would not have put any time into bitcoin if Satoshi's design did not include it.  Gavin then points out that the 1 MB was not in the original design, and quotes Satoshi's Oct/2010 post where he explains that the limit could be lifted when needed by a simple 2-line patch and a routine hard-fork release.  I did not see Greg's answer.  Considering that he has never answered Mike's "crash landing" post either, I assume that he just shut that fact too out of his mind, since it did not fit his convictions...)

If a bitcoin channel will cost $20 to set up and close (because of the on-chain transaction fees), while a litecoin channel will cost only $1, I bet that many users would use Litecoin for short-lived connections, in spite of them offering only Litecoin-level securuty.

At this point its totally unclear whether the Lightning Netvapor will be economically and practically viable at all.  There are still huge gaps in the design (think of a sketch of the bitcoin protocol without block rewards) and no one can answer my questions about how simple hypothetical use cases woudl work out.  

Therefore, it is impossible to say what its impact will be on the revenue that miners will get from transaction fees.  Greg is pathologically obsessed with the 1 MB cap and apparently has convinced many (but maybe not all) Blockstreamers that it will provide miners with suitable incentives without making LN fees too expensive.

One thing that we can say is that anyone who wants to open, close, settle, or top up an LN payment channel will have to pay a bitcoin transaction fee.  Thus, if the LN user base grows, the demand for those channel operations will become more desperate, the demand for bitcoin block space idem, the bitcoin transaction fees will become more expensive, and therefore the cost of the LN per user will increase.  

Would this business plan make sense in Austrian economics, perhaps?


Actually Lightning could use any cryptocurrency in place of bitcoin. Or even multiple cryptocurrencies, with some (or all) nodes and hubs doubling as currency exhanges.  That would let users choose the combination of speed, hashpower, and transaction fees that best suits their needs.  

(Maybe that is why BTCDrak of Viacoin, an altcoin that intends to replace Bitcoin, paid Peter Todd to implement OP_CLTV, (aka  OP_DAMN_DAMN_DAMN_I_TYPED_2160_INSTEAD_OF_2016) -- a bitcoin script opcode that is totally useless except for the LN payment channels.  And why Charlie "Litecon" Lee likes Blockstream's (non)scaling plan, in public disagreement with his boss Brian Armstrong.)

Even more: Adam once claimed that the LN would achieve 10'000 transactions for every blockchain transaction.  Leaving aside what that would mean for the coin lock-in periods of the channels, the rate of 1:10'000 is so close to 0:10'000 that no one would notice if the LN just dropped bitcoin altogether, and from then on just shuffled its "cryptochecks" around indefinitely, without ever settling them.  

Of course, LN users would never accept that.  It would be like that time when the US government decided that dollars would no longer be convertible to gold or silver.  We all remember the revolts in the streets, and the huge bonfires where enraged dollar holders burned their then-worthless bills.  Right?

It still works fine as a payment system.  But it is no longer a decentralized trustless payment system -- because 80% of the mining is done by 6 companies, of which 4 are Chinese and 1 is Ukranian-Georgian. 

One cannot notice that detail, but those attributes were what made bitcoin different and worthwhile.  Without them, it is just another centralized payment system -- and a terribly slow, complicated, expensive, unreliable, limited, understaffed, and insecure one.

It is like a Tesla whose electric engine has been replaced by a gasoline one.  It still works fine, but...

May I offer:

• The blockchain that you should trust is the valid blockhain that you can get that has the most proof of work.
• A blockchain is valid if and only if the software you are running says that it is valid.
• "The bitcoin blockchain" is the blockchain that you call "the bitcoin blockchain".

The question "in a persistent fork of the blockchain, which branch will be the real bitcoin?" is basically a religious question.  You should think about it as you would think of "in a schism of my Church, which side will be the True Church?".

What I meant is other people started talking of bitcoin taking a significant slice of VISA's market, either sincerely or to pump the price.  Satoshi himself does not seem to have intended that, and even suggested the opposite:


I read that as "bitcoin is not meant to replace VISA or bank wires, but is meant to be used when a third party is not available or really not appropriate".


Of course.  I was trying to guess what Satoshi was thinking when he defined the halving schedule.

What attracted the scammers and snake oil peddlers was the money.  The big ones came in early 2013, it seems, although Hal Finney is said to have predicted that twist already in 2009.  IIRC, Satoshi reacted has if he had not given it much thought.

Satoshi's basic mistake was to make the currency non-inflationary.  It is forgivable, since he was a computer type and not an economist.  I would have made the same mistake. Inflation is bad, of course -- everybody knows that, right?

Satoshi's disappeared six months after Jed McCaleb repurposed MtGOX to be a bitcoin exchange, and 3-4 months before it was sold to Mark. Maybe Satoshi saw the future when MtGOX started to attract speculators.

Oops, it is actually worse than N², apparently

Trying again: There is a complicated transaction that barely fits in a 1 MB block, that takes 30 seconds to validate.  Someone built a transaction twice as big, filling a 2 MB block, that takes twenty times as long (600 seconds; not just twice as long (60 seconds) as one would expect.

Right now there is no limit to the complexity of one transaction, except that the whole transaction must fit in a 1 MB block.  If the block size limit is increased, then one could issue the more complicated transactions like the one above, probably causing problems for miners.

But there is a known simple solution that (AFAIK) was used in BitcoinXT and/or BIP101.

This is a known protocol design bug: signatures were defined in such a way that the cost of validating a transaction with N signatures is proportional to N² rather than N. 

It has a known simple solution: limit the number of inputs of a transaction to some reasonable value, say 20 or 100, independent of the block size limit.  That will keep the cost of verifying one transaction bounded, and will inconvenience only a few users, by forcing them to break any big transaction into a chain of smaller ones.

IIRC, BiotcoinXT included this simple solution.  Hopefully it will be included in Classic too. 

Again: once, while discussing future user base growth, Satoshi considered 20% per year (doubling every 4 years) a "crazy" rate of growth.  That would be consistent with his chosen block reward schedule (halving every four years).  With a "non-crazy" growth rate of less than 20%/year, the USD value of the block reward would have deceased with time, forcing the system to transition gradually to fees.

Surely he was too smart to risk actual predictions, but on the other hand he cannot have expected the price to rise by a factor of 10 every year for 5 years.  So the USD value of the block reward too increased by almost he same factor, instead of decreasing as he may have expected.

The price rose so fast because of speculative investment and trading, fueled by predictions that it would one day replace VISA.  That surely is a use for bitcoin that he did not expect.  If there is something that the world did not need in 2009, and will never need, is another penny stock.  He would not have bothered creating bitcoin, if he knew that it would turn into that.