No joke there.  Until some 20 years ago, technical conferences used to be organized by reputable scientific societies, and the organizers "played to lose": they looked for the cheapest possible venues and formats, and begged money from sponsors, so that the registration fee could be as low as possible.  Often the conference ended in the red, and sponsors had to be begged again to make ends meet.

But in those 20 years things have changed.  First there were many entrepreneurs ("Nagib Callaos" is a legendary name there) who saw technical conferences as an opportunity to make money.  They would organize humongous conferences, with dozens of parallel sessions, high fees, in popular vacation places; and accept absolutely any garbage that was printed in eight US letter pages and had a title and authors at the top.  They made a profit serving the huge market of mediocre university profs who needed publications in their resumés but cold not get their papers accepted into the serious conferences.

For a few years, serious conferences would still try to keep registration fees down; but then the sponsoring societies realized that they too could use the conferences as a source of revenue, instead of a service that was part of their mission.  IEEE was one of the first to do that (some IEEE top guy even resigned in protest), but others followed.  So, the fees that used to be 150-300 USD are now 1000-2000 USD...

I confess that most of what I know I learned in the last 2 years, watching bitcoin demonstrate all the basic principles as a universal counter-example. 


No disagreement about that.


OK, one can look at bitcoins that way.

(However, the word "right" is meaningless if there is no entity that will assist you in realizing those rights.  No one will come to your help if the miners tomorrow decide to refuse your transactions, for instance.  But OK, I suppose that you intend "right" as "capability", namely you get private keys that, miners willing, will let  you move your bitcoins.)



That does not follow. Even with perpetual inflation, the ledger would still be immutable, and would work in exactly the same way.  Your right to write on it would be preserved, and you could exchange that right with others for money or services, just as you do today.  The only difference between that hypothetical inflationary ledger, and what you today think that the ledger ought to be, would be that the coin rewards do not go to zero, but remain above some value. 


Indeed, but the "slippery slope" argument is notoriously ineffective in preventing changes. 

Promises by some group of people that "we will never change our decision X" are totally meaningless: if one day a majority of those people gets convinced that X was a bad decision, that same majority will conclude that the promise to never change X was a bad decision, too -- and then go ahead and change X.

I have seen that happen many time in the administrative Boards that I had to sit in the university.  Once in a while someone will propose that the Board votes to not debate issue X again -- forever, or before N years.  Sometimes Common Sense will have left the room for a cup of coffee, and such motion will get approved -- only for the issue to be reopened two months later, due to "unforeseen curcumstances"...

By that argument, anything that required lots of resources to build, and cannot be easily replicated, should be worth as much as those resources, and gain value with time.  But there are many obvious counter-examples to this claim.  There are many old abandoned buildings that cost millions to build, may cost even more to build today, but now have negative value -- namely, the land is worth less with them on it than without them.


I know that quite well. (It is one of the two two only things that I managed to learn in my Economics course in grad school.) But trust is not physical, it is a state of mind.  Physically, you could in theory change the mind of all 1 million bitcoin holders, so that they would lose their trust and taste of bitcoin, with a rather modest expenditure of energy -- much less than what you need today to mine one bitcoin.  The difficult part, in practice, is getting that energy applied to the right synapses in their brains...

(The right words and images could perhaps do the trick; but trolling is still not a science, just a crude art...  )

Tough question. There are several "features" of bitcoin that seemed good ideas at the time; but now, in hindsight, they seem to be mistakes, and Satoshi himself would probably agree.

One of them is the expectation of huge increases in value, resulting from the fixed cap and the assumption that it would replace other means of payment.  That turned bitcoin into an allegedly safe and lucrative investment, and a speculation tool.  Most of the big problems that bitcoin is facing today are a consequence of that undue retargeting of the project.

So, one problem that would have to be fixed is making it unappealing to hoarders and speculators.  

Then the value of 1 coin would be determined by its use for e-payments, according to the money velocity equation.

Another problem is ensuring that its value stays in a bounded range, so that ordinary payments can be expressed without too many zeros before or after the point.  Say, between 0.01 and 10 USD.  In the first extreme case (about the value of 1 yen) one would not need decimal fractions, and do all accounting with integers. In the second extreme case, one could truncate after 3 decimals (i.e. use the "milly" as Americans use the penny).

A stable value would seem to attract hoarders and investors, but hopefully the two goals can be achieved with built-in demurrage (negative interest).  Namely, each UTXO loses value at (say) 2% per year, so you can spend only 0.98 BTC of that 1 BTC that you earned 1 year ago.  Those 0.02 BTC would implicitly go to the "Bitcoin Treasury" and would be redistributed as block rewards (so the block reward could remain constant forever while the total amount of coins in circulation would still be bounded.)  

Hopefully that negative interest would be enough to dissuade hoarders and speculators.  

However, in order to keep the value stable, the issuance (e.g. by block rewards) would have to be adjusted to the volume of payments and to the mean time between payments with the same coin. I have no idea how that could be done in a way that would resist manipulation by hackers (computational or financial).  It seems that another Satoshi would have to come out with another ingenious invention...

Another problem is the centralization of mining.  That could be fixed by keeping the block rewards too small to make mining into a profitable industrial activity.  But that may conflict with the need to put more coins in circulation as adoption grows.

Other problems are the centralization of all transactions of the world in one blockchain.  That design limits the scalability of the network and makes nodes expensive to operate.  That could be fixed perhaps by having a large number of separate blockchains, each containing the transactions of a certain subset of (say) 1000 addresses.  Then a 1-in, 1-out transaction would have to be recorded in only two of these small blockchains.

Other technical problems include rewarding all players (not just miners), ensuring that every transaction gets processed, replacing the UTXO-based accounting to the balance-based accounting that banks use, etc.

There are also non-technical problems like anonymity, non-reversibility, lost coins, legal jurisdiction, ... but each would be a long discussion in itself...

That is not "cost".  Even if bitcoins lose all their value, no resources or concrete assets will be destroyed.  The world will not become any poorer. Only, a few hundred thousand people who expected to get hold one day of mansions and Lamborghinis  will see those hopes vaporize, and those assets go to other people instead.

I don't think however that the introduction of extra inflation into bitcoin -- say, 1% per year, beyond the block rewards -- would necessarily induce all holders to dump, and the price to crash.  

In late 2013, when I first learned about bitcon, the Gospel of Antonopoulos was that bitcoins would become extremely valuable because they would replace credit cards and other traditional digital payments, and then demand for that use would drive the price up, according to the money velocity equation.  Moreover, that would happen very soon, because the price had been growing 1000% per year, and no one could look at that straight line on the log plot and doubt that it would continue.

Well, 1% of inflation per year would not destroy that argument; at most, it would delay the growth a little.  After all,  that growth of 1000% per year happened while the miners were creating an inflation of more than 10% per year.

But the Gospel has changed several times since then.  Mircea, for example, wants us to believe that the "Moon" will come just because the supply is fixed.  Even if it were true -- again, 1% of inflation per year, even extended for 20 years, could not possibly prevent the price from rising 1000% in that interval.

Blockstream, on the other hand, wants us to believe that bitcoin will go to the Moon because not only the supply will be capped by 21 million, but the transactions will be capped to 150'000--180'000 per day, and somehw that will induce some companies to pay more for bitcoins. Whatever.  Again, 1% inflation per year cannot possibly spoil such a powerful "whatever".

But, on the other hand,  I agree with one thing: it is very unlikely that those "20 businessmen" who will decide the future of bitcoin will ever agree to putting any sort of inflation or demurrage in it.  That was not a prediction that I was making.  My point was only that the 21 million cap is not "guaranteed by math", but only by people.

I still believe that bitcon's price will go to zero in a few years; but almost certainly not by that route...

Liked that!


Actually inflation *can* be introduced into bitcoin in a number of ways, at nearly zero cost.  The supply cap or 21 million BTC is not "guaranteed by math"; it is decided by humans who choose which software to run.  Paraphrasing the old proverb, "no protocol change is prohibited if the right 20 business men decide to implement it".  

Then there are also altcoins, both original and forks of bitcoin.  Litecoin already has faster confirmation times and better antispam protection than bitcoin (thanks to a measure that Charlie Lee implemented in Litecoin, but the Core devs rejected for Bitcoin).   Viacoin aims to be a replacement of bitcoin; it is noteworthy, among dozens of other altcoins, because at least two Core developers (Peter Todd and BTC Drak) work for it.  

So it is not unconceivable that some other coin will conquer Bitcoin's user base -- and value.

Even if the supply cap of Bitcoin *were* guaranteed to never change, that would not guarantee fabulous prices: scarcity does not imply value.   There is a finite number of tickets for last month's Penssilvania State Lottery, that can decrease but never increase.  With modern micro-analytic techniques, forgery of such tickets would be impossible in practice.  Yet those tickets are now worth the price of dirty paper scraps.  


It is not inelastic at all.   The number of regular bitcoin users is probably less than 100'000.  It is possible that the network will not be able to serve 1000 times as many users in 10 years time.  However, it can certainly serve 10 times as many, even today.

There is a dozen people who think that they own the system, and want to impose an artifical 1 MB lmit on the block size.  Such a limit would prevent the traffic from growing to more than 0.750 MB/block, a level expected to be reached in 2016.  But there is no justification at all for why the block size limit should be 1 MB, rather than 0.100 MB or 10 MB.  Even those guys are now admitting, grudgingly, that the size limit should be raised "now" to 2 MB/block.  The major businesses support BIP101 (with 8 MB blocks now), while the miners support BIP100 (that gives them dynamic control of the block size).  

So, there goes the "inelastic limit in its maximum transaction volume".

That group has claimed that letting the traffic grow beyond the current 0.750 MB/block limit would have a number of harmful effects, such as fewer full nodes, increased miner concentration, increased orphan rates, etc.  But those claims are just FUD.

Miners are too concentrated today and will tend to be more concentrated. The reasons are economies of scale and of location, that are independent of traffic levels or block sizes.

The block size has more than doubled over the past 15 months, and yet the average orphan rate has remained totally flat, at ~1.5 orphaned blocks per day.

The number of full nodes will continue to decrease, no matter what.  Not just because of the increased traffic, but mainly because shutting down a node saves saves money for the node operator, while starting a new one costs increasingly more in bandwidth and time -- even if traffic were to remain constant. (The blockcchain now grows at 100 MB every day, and will soon grow at 150 MB/day even if the 1 MB limit is not raised.)

(And then there are those small-blockians kiling hundred of full nodes because they dared to express support for XT... )

Watching people be duped into putting their savings money into a pyramid scheme, thinking that they are investing in the future global PayPal -- no, that is not fun at all.

Watching gamblers of the day-trading game lose money -- meh, that is part of the game.

Watching scammers tear other scammers apart -- that is priceless.

Report by MtGOX's bankruptcy trustee, 2015-09-09
[ scroll to the end of the document for the English translation ]


That is 2.6 trillion yen. Dividing by 120 to get USD, we get about 21 billion USD, of which 20 billion USD are the top 3 claims.

It would be REALLY interesting to know who are those top 3 claimants who believe that they are entitled to receive twice the entire bitcoin market cap -- each.  If those three claims stand, they will take 95% of the spoils, and the other 24,701 claimants will divide the other 5%.  

It would be fun to watch.

Unfortunately that is not decided yet:


But at least

Could that be a pyramid whose contractors omitted an entire floor, hoping that the Pharaoh would not notice?

Speaking of which: did you, or do you know someone who did, ever try to memorize the Lorem Ipsum text, just as some people memorize digits of pi, or the name of the town of Taumatawhakatangihangakoauauotamateapokaiwhenuakitanatahu?

They already are.

And it will end poverty and corruption. And maybe cure cancer too.


I don't recall saying that BitcoinXT would win; if I did, I apologize profusely.  No matter how much I try, I always overestimate the intelligence of bitcoiners. 

(I just saw a bunch of them exchange hate comments about Gavin on reddit, as if he had always been the arch-enemy of bitcoin.  Amazing what you can do with a smear campaign, if you have 21 M$ of capital to defend and a couple thousand bitcoiner minds to play with.)

I still think that the limit will be raised, and the Blockstream guys will either cede or be left sucking their thumbs.  But I may be overestimating again...


After having admired Dr. Back's words, and even having had the honor of debating directly with him on reddit, I must say that, for his technical and ethical qualities, he is one of the most outstanding bitcoin personalities -- right there besides Roger Ver.  I do hope that his career develops the way he deserves.

It was not a real DoS attack, because it used fixed and relatively low fees. Anyone who paid slightly more than them got trough, as if there was no backlog.

For that reason, it did not convince the small-block camp that 1 MB is too small.  If that was the intent, it was too much expense and effort for a dubious or null result. 

In fact, many in the small-block camp have interpreted the tests as a trick of the big-block camp, and used that to discredit them.  So, if it was indeed a big-blockian trick, it backfired.

I don't really know what to think of them...

In the original schedule, the last day for filing claims was 2015-05-25 or thereabouts, and the trustee was supposed to validate the claims and present a distribution plan to the Court by 2015-09-25 or thereabouts.  But the filing deadline was postponed by 1 month, so I guess that the trustee's deadline will be postponed too...

For example, there are these comments by Satoshi on 2010-10-03, when the block size limit had already been lowered to 1 MB.  Jeff Garzik had proposed a patch that would raise the max block size to 7.16 MB, so that bitcoin would be able to handle PayPal's average traffic at the time:



Up to that date, the traffic had been less than 1800 tx/day on average (today it is 120'000 tx/day), which means ~7 kB per block (today it is ~450 kB/block). 

With 1 MB max size limit, the capacity at the time was the same as today, ~200'000 tx/day or ~750 kB/block.  (It is less than 1 MB/block because certain optimizations by the miners force them to mine some number of empty blocks). 

So, by setting the max block size to 1 MB, the devs at the time set the capacity to about 110 times the demand.  Obviously the intent was not to create "capacity scarcity", any time soon...

But today the capacity is only ~65% more than the demand.


Sorry, the "That" in my quote is not the previous phrase, that you bolded, but the claim that I was replying to: that increasing the block size would lead to increased mining centralization.  That is one of several baseless claims that Adam made (after "increasing the block size limit will cause the number of full nodes to decrease").

My reply in bold is that mining already got centralized over the last 2-3 years; but the causes of the centralization are well understood, and growth of the traffic was not among them.  Those causes will continue for the foreseeable future, and keeping or raising the block size limit will make no difference for mining centralization.

Again, it is possible that the devs do not see how wrong that argument is?

The max block size that was put into the rules in 2010 was not "1 MB" but "a value small enough to prevent big-block attack but still much larger than traffic so that it would not create spurious scarcity, and every transaction would be confirmed as quickly as possible." 

Blockstream now wants to change the max block size into "a value that makes block space scarce so that the fees will rise, even if transactions will be delayed by many blocks."  Is it possible that they don't realize that keeping the 1 MB limit as tre traffic crashes into it is a radical change in the protocol?

That new parameter -- the scarcity control -- happens to be 1 MB now, but they admit that they may set it to higher values in the future.  But always to values that will keep space in the blocks scarce.  Who decides how scarce?  Well, the devs ... or maybe the miners, but by grace and privilege of the devs...


There is NO evidence that the centralization of mining that happened so far was due to the natural growth in the block sizes, or that such growth would affect it in the foreseeable future.  That is just one of several dishonest FUD arguments that Adam Back used at one point.  The economies of scale that resulted in concentration of mining are due to the savings in the costs of equipment, space, electricity, cooling, personnel, management, etc.  that exist  for bulk purchasers and for certain geographic locations. 


[/quote]

The Blockstream plan is to FORCE a particular capacity, by setting the block size limit in the consensus rules, and then let the fees "naturally" adapt to that artificially constrained capacity.

Why is that better than forcing the fee directly, and ensuring that the capacity is well above the demand?  (I ca think of several reasons why it is much worse.)

Is it possible that the devs do not see that their plan is just as "ideologically impure" as setting the fee directly?

The error is in that very first claim.  There need not be a scarcity of capacity for the fees to serve their purpose.  The fees could, and should, be mandatory and set in the 'consensus rules'.

In fact, scarcity of capacity in a network means lousy service, which means loss of users, which means that the fees will not rise as hoped....

Except that the Chinese miners were the first to declare, with signatures and stamps, that they would be OK with a block size LIMIT of 8 MB.

The support for BIP100 seems to be mostly (a) miners who like the idea of deciding things on their own, (b) nerds who like the proposal because it is a nerdish solution, and (c) Blockstream supporters who are pushing BIP100 as a "divide and conquer" strategy against Bitcoin[REDACTED].

By the way, BTC Drak, a co-founder of Viacoin and apparent friend of Blockstream, has just proposed BIP105, another dynamic block size limit adjustment proposal.

And did I mention my own BIP99½?

(I honestly and modestly think that it is the best max block size increase proposal.  I posted it in order to make sure that bitcoiners will not implement it.   That would spoil the fun by ending the size war and giving us another 3-4 years of the same boring agony.  Don't tell that to the bitcoiners.)

Those were not real "spam attacks" because they used fixed and relatively low fees (0.2 mBTC/kB), so the backlog was easily bypassed by clients who knew what was going on.  

A real spam atttack will try to ajust the fees dynamically so as to keep some significant fraction of the legitimate traffic perpetually out of the blockchain, even while the legitimate clients adjust their fees to try to get through.  

At least that interview gave a plausible explanation for the "stress tests":  Coinwallet.eu has a wallet that, they claim, will correctly compute the fee needed to get through the traffic jams... that they create themseves.  Hum.  

It would be diabolically nice if someone else did a REAL spam attack during their stress test, so as to really test their supersmart wallet (and other supersmart wallets out there).  Their "stress test" may even save the spam attacker some coins...