That's all it was ever supposed to be.  Bitcoin is a digital payment system that doesn't have a central point of failure and doesn't divert money into the hands of an owner. 

All this crap about bringing down banks and governments is just that -- crap -- and always has been.  We're not revolutionaries here, or if we are we're laughably ineffective ones.  We're just working out a more efficient way to transfer value. Even if we figure out a more efficient way to transfer value, banks and governments have legitimate jobs they are still needed for and can legitimately get paid for doing. 

And transferring value more efficiently does not make anyone an enemy of the state or crap like that; on the contrary it makes the commerce on which the state depends stronger.  If it works this is the sort of thing that earns gratitude and Nobel prizes, not the sort of thing that earns enmity and censure.  Sure, there'll be a few disgruntled money transfer businesses like Western Union.  Boo fuckin' hoo.  Nobody important gives a crap about them, any more than they gave a crap about the coal-oil companies that went out of business when cities electrified their street lighting.

Also?  Quit framing with broad racist stereotypes.  It makes you look like a goddamn hatemongering tool.

As I said, 20MB has been extensively tested.  4GB is okay with the software but subject to practical problems like propagation delays taking too long for blocks crossing the network.  4GB blocks would be fine if everybody had enough bandwidth to transmit and receive them before the next block came around, but not everybody does. Right now I wouldn't expect the peer-to-peer network to be stable with blocks much bigger than about 200MB, given current bandwidth limitations on most nodes. 

Maybe, but I'm not even going to have read far enough to evaluate that part of their protocol. 

I'm going to get as far as "32-bit seed" and "shuffle" and go, "oh crap they can't do  math!"  and go on to someplace else.

If I were habitually looking for opportunities to cheat, I'd see "Oh crap they can't do math!" as possibly good news - and read on to figure out which of us it's empowering to cheat. 

But I'm pathologically honest, and "Oh crap they can't do math" means cheating is possible. Therefore I'm out of there, and usually assume that if somebody is putting up a site where cheating is possible that it's going to be them cheating rather than them allowing the customers to cheat.

Here is the basic scenario as I see it (I'm for the fork FWIW).

If there is traffic to fill up blocks to that size, and the blocks can grow to that size, the blocks grow to that size and the system continues to function. 

If there is traffic to fill up blocks to that size, and the blocks cannot grow to that size, then the system fails.

If there is not traffic to fill up blocks to that size, and the blocks can grow to that size, the blocks grow up to the size of the traffic it takes to fill them, and the system continues to function.   A very rare block mined by a troll might grow to that size, but expect less than five of those a year.

If there is not traffic to fill up blocks to that size, and the blocks cannot grow to that size, the blocks grow to the size of the traffic it takes to fill them, and the system continues to function.

Do we fundamentally disagree on how often we should expect blocks to be mined by trolls?  'Cause I don't see anybody else risking his return on investment in hashing power by growing blocks needlessly without having legitimate transactions to fill them.

I think sidechains as such are doomed to fail because I haven't yet seen a plausible security proposal -- that is, we need to make mining the side chains more profitable than attacking them.  :-/

On the other hand, side chains are just one implementation of a thing that is definitely NOT doomed to fail -- that is to say, blockchain pruning.  We can find a way to reduce the size of the required working set to manageable levels without going to side chains.

"Silk Road" is a name, like "MIckey's Floating Poker Game." 

When "Mickey's Floating Poker Game" becomes well known, as a place where a lot of people want to gamble a lot of money, etc, Mickey starts making tons of cash. 

But then Mickey screws up and goes to gaol.  Now Mikey looks around and starts thinking about how much money he could make if he continued to have "Mickey's Floating Poker Game."  It's not like Mickey's going to stop him; Mickey's in stir.  It's not like there's a legal trademark to worry about; it was an illegal enterprise in the first place. 

So Mikey puts out the word that "Mickey's Floating Poker Game" will meet next week in the basement under the orphanage, and the world goes on. 

Heck, maybe the first Mickey retired, and the one in stir is the fourth or fifth, and the second is running a farm in Leicester now, and our boy Mikey doesn't even know he's the fifth or sixth rather than the second. 

Don't defend it silly.  Just fix it. 

Everybody knows shuffling a pack of cards takes 240 bits.  If you do it with less than 240 bits everybody knows you're cheating.  So they don't play with you, and therefore, even if you're NOT really cheating you lose money.

Everybody knows the randomness used by a game has to be the result of a bit commitment from both sides.  Therefore if your game isn't providing a bit commitment and taking one, and allowing the results to be checked to make sure that the randomness is really the result of combining the numbers committed to, everybody knows you're cheating.  So they won't play with you, and therefore even if you're NOT really cheating you lose money.

Again, it's not worth arguing about.  The only way to "win" the argument is by fixing it so that the site has the desired properties.  It isn't a matter of opinion, it's a matter of math.  Changing anyone's opinion about it won't matter because you can't change the laws of math.  It's either right or it's wrong, and if it's wrong, you can't PROVE that you aren't cheating.  So even if you aren't, everybody will assume you're cheating.  The only way to convince anyone you're not cheating is to use the math to PROVE that you aren't cheating.

I dunno how much Sergio's cards protocol has in common with my dice protocol, but what I use is that each side provides a sequence of inputs where each input is the hash of the subsequent one.  It's a cute trick where each number is also the "bit commitment" for the number that follows.  It requires that each side generate its sequence from the end to the beginning, but if you run out, you can just generate a new sequence and exchange bit commitment numbers again to continue.

At this time the software handles blocks up to 4GB and has been extensively tested with blocks up to 20MB.  So there's not really a problem in terms of API, nor a need to chop discovered batches of transactions up into bits.

The problem is sort-of political; there are a bunch of people who keep yelling over and over that an increase in block size will lead to more government control of the bitcoin ecosystem (as though there was ever going to be an economically significant bitcoin ecosystem that didn't have exactly that degree of government control) and that they don't want just any riffraff in a faraway country to be able to use their sacred blockchain for buying a cuppa coffee (as though allowing anybody anywhere to buy anything at any time somehow isn't the point of a digital cash system).

Neither point makes any damn sense; As I read the opposition they fall into three groups, the Trolls, the Suckers, and the Scammers.

The Trolls know the arguments are nonsense and are yelling anyway because it makes them feel important. 

The Suckers don't know nonsense when they hear it and are yelling because they're part of a social group with people who are yelling.

The Scammers have thought of a way to make a profit ripping people off during or after a hard fork, but it won't work unless there are Suckers who think that the coins on the losing side of the fork aren't worthless, so they keep yelling things to try to keep the Suckers confused as to the facts.

That gain in value is an illusion brought about by measuring value in inflating currencies.  It will be worth larger numbers over time, not larger wealth.

Getting rich on other people's work only happens if you've loaned them money or otherwise ensnared them in a debt-based system, and the beauty of Bitcoin is that it's NOT debt-based.  People who actually produce wealth and don't owe loans they have to pay back, and aren't enmeshed in a gigantic loan made via the whole currency and banking system they use?  They get to keep the wealth they produce.   People who just sit there holding some coins and expect other people's work to make them rich, post adoption, are going to find that it just doesn't work.  At least, I devoutly hope so.  Insofar as I'm idealistic, that's the ideal. 

Bitcoin is second-order productive -- it doesn't directly produce value that can go into someone's pocket.  What it does, ideally, is to cut out middlemen, remove market friction and resistance, and thereby make all the first-order productive stuff that does put money into people's pockets  (people working their assets off) work better.

I want Bitcoin to go mainstream. 

Bitcoin solves a problem - frictionless, conversion-free transfer of value from person to person, world wide.

There are a thousand people standing there wanting to get their hands in your pocket whenever you want to transfer value from one place to another, for whatever reason -- western onion, visa, innumerable currency exchanges, bankers with foreign-exchange fees, money order companies, etc ad nauseam. 

Bitcoin.  Bam.  Go get a job, you goddamn parasites, and try to be productive for a change.

Bitcoin is one stroke that cuts out a whole bunch of commerce-impairing middlemen and toll-takers rent-seekers and obstructionists.  It's going to be regulated, monitored, and taxed.  Because it's money, and that shit happens to money.  But it's not going to be systematically stolen by powerful crooks without leaving any evidence, the way other kinds of money are, because blockchain == evidence.  The petty corrupt can't steal Bitcoin and then pretend that it hasn't been stolen.

That is the primary reason I want it to be adopted.

As an investment?  Short term, if/while it grows with increasing adoption, that's high risk but high potential gain.  Long term?  Not so great.  In the long run, it doesn't represent a group of people using assets to  be productive, and bluntly when nobody's working their assets off, nobody's getting paid.  So returns measured in constant units, in the long run, must tend to zero.  Whether that happens at value zero (meaning complete and utter failure) or at value N>0 (meaning stabilized prices after adoption) it's going to happen.

If it has value in the long run, then it will be useful in portfolios an inflation hedge and a counterbalance against overexposure to the financial-services sector, like gold.  But once it's been adopted, it won't be the primary basis of a portfolio that's actually profitable.

You're so right.  And I don't have to ignore you either.  But I'm feeling generous, so I probably will anyway.


And if I wasn't convinced before that you need to be ignored, I think that will do it.  Seriously?  You don't want people to be able to use Bitcoin to make purchases, but you're still referring to it as money?  You want it to be decentralized, but you want to exclude people who aren't somehow 'local' to each other from making transactions?

You have any idea how much nonsense you're making?

Seriously.  Quoting yourself, quoting Mircea Popescu?  That's supposed to convince me of anything?  

Why am I even bothering to answer?  

Okay.  Visa and Bitcoin, ideally, are both used by people anywhere to pay people anywhere.  As such, they have to deal with many of the same engineering problems.  

The specific problem under discussion - transaction capacity vs. transaction load - is one of the ones they have in common.

I look at real information from Visa because they have a working system.  If you want us to look at real information from some different working system, present the information; it will be useful.  If you don't want us to look at real information, get the fuck out.  We NEED to be looking at real information.  

If somebody in Mali wants to buy Alpaca socks from somebody in Chile, and can't because there's no data-set that Chilean merchant can use to verify the validity of his Malian unspent txouts, that is what we call a failure of the system.  Maybe you don't like the existence of this large data-set.  That's okay; it works whether you like it or not.   Anything that would be as effective at avoiding failures will also work whether you, or anyone else, likes it or not.  In fact working whether anyone likes it or not is ESSENTIAL to avoiding failures.  

Make a proposal that there is less reason to dislike, or reason for fewer people to dislike, but which will still work whether or not anyone likes it, and it may be valuable.  But if you don't have a better idea, then please shut up.

Yes, I was talking about the sigs.  OpenSSL moved to a strict form requirement, which some of the signatures already recorded in blocks did not comply with.

Reading IRC logs makes my brain bleed, but I went and read it because of the extraordinary claim.  As usual for IRC, the  "discussion" amounted to abuse and posturing.  And as usual for extraordinary claims, the evidence does not support it. If you think that Gavin committed to making or not making a hard fork in that log, you have wilfully misinterpreted it.

As one of the "cognitively different" I think far too literally to have deep discourse with normal human beings in realtime.  It takes a long-ish time for me to think through people's written communications to respond appropriately and productively.  And sometimes I still get it wrong.  I VERY MUCH identify with your above request, but long experience tells me it is unlikely to achieve your desired result.

In many cases the problems others are trying to solve are with the conversations themselves rather than the things the conversations are about.  A straightforward plea such as yours to state the problem is generally seen as irrelevant, and a direct answer would possibly be harmful, to actually achieving such goals.  Moreover, a direct answer in that case would require more insight into one's own motives than most normal people have.

As with most things involving normal people, it's far easier to be certain than right.  


That is an excellent statement of a goal I share and some of the values that a good solution must maximize.  I am in agreement with it.  

I think that there is sharp divergence of opinions in the community regarding the relationship between censorship-resistance and scaling to millions or billions of transactors.  Some seem to assume that absolute censorship-resistance and untraceability are necessary to future growth.  Others seem to assume that they would hinder future growth.  Some seem to assume that absolute censorship-resistance and untraceability are more important than future growth, and others that they are unimportant relative to future growth.  

Some seem to think that Bitcoin has ever had any kind of absolute censorship-resistance and untraceability, but this is a misapprehension.  It is 'cash-like' in that if you are very very careful you can acquire and spend it without someone knowing who you are.  That is IMO about the greatest degree of censorship-resistance and untraceability that realistically will not result in aggressive prevention of scaling the network by various law enforcement concerns.  

I would suggest an attempt to clarify the relationship, but any attempt to do so via open discussion and debate would as some say  "generate more heat than light."  It might be worth it, but bring hot dogs, marshmallows, and an asbestos suit if you plan to light that fire.


They most assuredly can kill it, for all practical purposes.  They can make it illegal to be a Bitcoin-accepting merchant, prosecute unlicensed money transmitters dealing in Bitcoin and deny licenses to any money transmitters who plan to start.  That would be an effectively fatal blow.  Network effects and immediate mass dumping by the 'get-rich-quick' speculators would do the rest.  We could be back to that ten-thousand-bitcoin pizza within a few months if the IMF and a few nations signatory to global anti-money-laundering treaties  decide to ban it.  

As you point out, our most effective strategy is to get big enough, fast enough, to disincentivize the delivery of that fatal blow.  Bitcoin needs to be a system that makes significant contribution to the economy - or at least one whose extinction would cost elected creatures substantial numbers of votes - before it can be seen as 'too big to fail.'


Allowing larger block sizes is necessary.

In the future, there will be need for more than 1MB of tx per block.  When that happens, our choices are two: have a larger block size, or fail.  As blocks grow larger, the costs in bandwidth of running a full node become higher so there will be fewer full nodes.  That's bad for security, but unrelated to the block size limit as such.

Blocks will become larger, no matter which limit is in place above them, at the rate transactions grow.  The bandwidth costs of running a full node scale with the number of transactions, not with the limit on the number of transactions.  So full nodes will drop out as actual transactions scale up, not as the limit is raised.  The security question relevant to the higher block size limit then is whether it is better for security to fail or succeed when more than 1MB of tx per block are needed.  I say it is better for security to succeed.

Visa seems to get along fine with a peak tx volume that's about 12 times their average tx volume.  That seems like a reasonable margin to me, both for peak traffic and for sudden growth within a retargeting period, while remaining low enough to limit any DoS attack that could be executed via large block size.  So I would suggest a block size limit set once every 2016 blocks at the same time as difficulty retargeting, at 12 times the average for the previous 2016 blocks.  

My main concern with the proposal this conversation started with (20MB and %40/year) is that adoption of a new technology, if and when it happens, is not that gradual.  This is another reason I advocate a plan capable of growing by up to a factor of 12 in successive 2-week periods; if this happens, the timing will surprise all of us and the speed with which the change happens and the size to which it grows will surprise almost all of us.

That scenario is also why I think that just scaling up the blocks is not adequate.  Bigger blocks would get us through a crisis, but security depends on every user having a manageable set of data that allows them to independently verify transaction validity, and a block chain that grows linearly relative to a userbase that grows exponentially does not make a manageable set of data.  So, IMO, the most important scalability/security job after block size is not side chains - it is blockchain pruning.

FWIW, there is already a financial incentive that makes it unprofitable for miners to generate big blocks.

Bigger blocks have a longer relay time, hence a higher chance of becoming orphaned blocks.  If someone else's block reaches more nodes than yours in a race, it costs you the fees and the block subsidy.

I'm not entirely sure that we still need a block size limit.  Getting a 'spam block' these days would seem to require someone, or some combination, to have invested adequate resources in hashing to get a block, but still not value the fees or block subsidy.  That's a very plausible scenario during bootstrap (or for an altcoin), but the hashing resources required to get Bitcoin blocks are huge, and I don't think trolls will invest that much money just for lulz.

I don't think I ever observed any sign of egotism in SN's stuff. 

He never gave a damn who got credit.  Never bragged on a darn thing.  Had nobody he wanted to prove anything to as far as I know.  Never said "I told you so" about anything.  And was just as hard on his own code and ideas as he was on anybody else's.  He just wanted to get stuff done.  Anytime you wanted to have a conversation that wasn't directly about getting stuff done, he'd just drop it and go get stuff done.

That is all it takes to remain in the shadows.  It's not godlike self-control; it's just a commitment to purpose over ego.

Ooooh yeah. Been there, explained that.  Best case scenario with 5% of the hashing power is dropping 6 out of every 7 transactions on the ground. And it'll take a hell of a lot longer than 2 weeks to get 2016 blocks.

But, you know, I'm waiting to see how they cope with Berkeley Database 4.8, and OpenSSL version 1.01k, and how things are going to fail when google changes the format of its protocol buffers, and how other things are going to fail with the next revision of boost, and ....  well, ALL of that stuff I keep pointing to and saying we need to fix?  All of the stuff that, until it gets fixed, is going to give us occasional maintenance jobs, some of which are emergencies?  

When software depends on a property of a library which is not important to the main purpose of that library and which may be changed at any moment, it is vulnerable to version breaks.  And right now, Bitcoin has several version breaks.  

Berkeley Database is one that happened a while ago but we've never cleaned it out of the code.  OpenSSL broke just last month.  Protocol buffers and Boost haven't broken yet, but the people who maintain them WILL change them eventually.  Protocol buffers, in particular, are going to be a semi-emergency when it happens (same as OpenSSL was) because we have a dependency on the binary form of the data; if it changes, our signatures break, so we can't just convert data.  

But hey, if you're committed to never changing the software?  I guess you never have to build it either.  Just static-link that sucker and wait for the next Heartbleed bug....

True, you'll remain able to get the old version out of the repository, because it has all the past versions.  If you intend never to change anything ever again, I guess that's enough. 

Good luck with that.