But you start from the idea that people are using a centrally designed code.  The idea of a decentralized system is that there are hundreds or thousands of different client codes in principle, in other words, that ideally, everybody writes his own code.

Putting a hard coded check point is a bit like thinking that one can ban certain web sites from being visited by putting a hard coded ban in the official, centralized, unique web browser software, no ?  A priori everybody can recompile his version of web browser, with is preferred check points.

It seems that people attach a lot of importance to the rules in the code, but that is because most crypto currencies are totally centralized in their coding, and only one entity is writing the code.  Core had that centralized power until recently in bitcoin, and in most other coins the "dev team" is the centralized decision force ; usually about the *protocol*.  However, if they start putting block chain check points in the code, they are also doing transaction consensus centralization.

If that were true, we wouldn't need proof of work.  We could use proof-of-non-sybil-node.  Now *that*'s a boy of a consensus algorithm 

In fact, we should have a guy with a web site that publishes the IP of all the good nodes (according to him) that are entitled to vote, what about such a consensus system ? 

It didn't occur to me that that was a block chain protocol hard fork, in the sense that the blocks made afterwards were incompatible with the valid protocol before (definition of a hard fork) ?

(and no, orphaned blocks are not a "hard fork" )

That said, what happened just before WAS strictly speaking a hard fork, when the block size limit of 500 KB was raised !  Funny how that was possible back then when it was a purely technical parameter, and causes so much troubles right now !

Indeed, franky1 is AGAIN confusing both.

Of course, users use nodes.  But users market weight is not proportional to the NUMBER OF NODES.

A big exchange has one single node, but has a bloody heavy market weight.  I'm running a node too, and my market weight is essentially zero.  The 5 times a year I buy something with bitcoin is entirely negligible.  Do you really think that my node should count as much as a big exchange's node ?  Do you think that my node is as heavy as the Winkelvoss brothers' node they may use ?

There are about 4000 bitcoin nodes.  My weight is 1/4000 of "nodes".  I don't control $5 million dollars of bitcoin (at all - I'm only using it as a currency, and very rarely so).  I'm maybe even going to set up a second node, for the fun of it.  Do you think that my vote should count twice ?

THAT is the reason why proof of work is used as a consensus "voting" system, and not "proof of node".

Users only need ONE single node to connect their wallets to. 

Here you go again.  If pools want A, and they have a serious $$$ stake in A, and the currently active nodes want B, what stops pools to fire up 3 times more nodes voting A for a tiny fraction of $$$ ?

Did you still not understand that proof of work was invented to avoid vote by node, because the above Sybil attack is far too easy ?

And don't confuse users (people exchanging coins for value in the market)  with nodes again...

I told you already: because ANY hard fork that is not 95% consensus will make 2 coins, of which the one "leaving" is the alt coin, and because the large majority of miners wants to keep 1MB blocks and transaction scarcity, because the fees are needed to pay for the PoW in the future.

Miners have to follow the market in a hard fork, at least, if the PoW algorithm is the same so that they have the right hardware.  But for that to happen, a hard fork is needed first.  Now, the thing going for bitcoin is its brand name, so nobody is going to fork off without being able to get the brand name ; and miners are not going to help him.  For the current miners, status quo is best, so they will keep with status quo and keep the brand name as long as they can.  Any forking, with alternative mining, with PoS or whatever, will be an alt coin without the reassuring security and continuity of what the current miners are continuing to make.  Most users will keep using the old system, and the "real" bitcoin.  

So it is essentially impossible to hardfork and get the bitcoin brand name without the current miners.  Yes, you can create your alt coin and see what it does in the market.  For a PoS fork, you ONLY have to write new code.  For an alternative PoW fork, you have to have sufficient PoW in the new algorithm available to garantee the security of your altcoin if it pretends to be a bitcoin fork, which is near to impossible.

This is why it won't happen, until bitcoin's brand name is essentially eroded.  However, when that is the case, why forking, and not using another crypto that has the features you want all together ?  

Wise words.  Agree with everything in this post.
(apart from the hard fork in 2014 ?  Hard fork in the block chain protocol ??)

Bitcoin has decentralized aspects, but I don't know how long this will still last.  Of course, there was a strong centralization in the protocol, by Core.  But for a long time Core respected essentially the immutability apart from technical issues that don't matter much: bitcoin was *economically* working like it was lined out in Satoshi's paper, and Core hardly changed anything to that.

You have to distinguish the *software* written by Core and the protocol adhered to by that software.  The protocol has only undergone small changes that were of very technical nature, and didn't actually impact anybody at the moment of the modifications ; in other words, those aspects of the protocol were essentially unrealized and hence open to modification.

So, in as much as there was "leadership", that leadership didn't modify the protocol in essential ways, and was hence equivalent to its absence.

In functioning, bitcoin went from highly centralized (Satoshi and a few guys), to very decentralized, to back rather centralized with 5 or 10 or 14 mining pools.  On the other hand, Core, trying to impose its leadership, actually lost it and now on the protocol side, bitcoin is in fact more decentralized (over 5, 10 or 14 pools) than when Core had de facto leadership.

Is a group of 5, 10 or 14 antagonists decentralized or not ?  In as much as they don't collude, in fact, yes.

You can have a decentralized system from the moment that you have 3 antagonists that can't come to an agreement.  But it is fragile, and you never know if there is collusion or not.

So, bitcoin is still somewhat decentralized in as much as these 5 - 10 - 14 entities do not collude, which nobody really knows.  On the protocol side, bitcoin got just more decentralized with the fight between classic, Core and BU, because it wasn't and there was leadership, but there isn't much any more, which is good.

I am exactly of your opinion.

Ok, let's not waste your time (mine can be wasted ).  But I just want to say that I am closer to your position than you think, and that I don't think that crypto is ONLY for dark markets, but for those places where centralized systems cannot go, or became too corrupt to be efficient.  Your belief in the corruption of these systems is maybe different from mine.  I think they are corrupt, but not to the point where they have lost so much efficiency that decentralized systems can compete, except in niche applications.  Of course, the day that they break down, decentralized systems will have a large ecological niche to take, but today, that is not the case yet.

The idea that bitcoin, or whatever decentralized system AS OF TODAY will compete to death fiat currency without major crisis, is ridiculous.  Of course, you can count on the big crash coming.  Sure.  But your experiment will not give any results of significance before.

Yes, you're right.  I meant to say that if you fundamentally change the protocol, you have a different coin.  But of course, if you start a new chain, you ALSO have a new coin, as you correctly point out.


My idea is: where does immutability even come from ?  Why should people join a system of which the rules can arbitrarily change ?  Would you buy a coin of which chances are that tomorrow, "a majority" votes that you don't own that coin any more ?  Would you trade value for transactions that can be undone the very next day "by majority" ?

Even though in principle, any consensus mechanism can change just anything, any time, you would EXPECT that the basic rules you subscribed to when entering the system, are "enforced" by some mechanism, no ?

But as that "mechanism" has no centralized leadership, and is to be trustless, it cannot have any "moral decision power".  It cannot distinguish which changes are in fact "good", and which changes are "bad".   So there *must be a mechanism that enforces stability of the rules*.  That mechanism is what guarantees immutability.  If you don't think that such a mechanism is at work, you must presume that there are only 2 possibilities:

1) there is a central form of authority, that will distinguish good from bad ; for instance, a genuine majority vote inspired by charismatic leadership IN WHICH YOU HAVE TO HAVE TRUST

2) anything can happen any moment, and your holdings, transactions, whatever, can and will change in an ARBITRARY way any moment.

In 1) you don't have a decentralized, trustless system.  You have something very similar to normal banking.
In 2), you have a totally arbitrary system that is unstable.  Impossible to generate any reasonable monetary belief.

My idea is that 1) is possible, but that there IS a mechanism at work: decentralized antagonism.  No antagonist is strong enough, and can find enough collusion, to get HIS changes imposed by majority.  No majority can be reached over anything else but status quo on IMPORTANT (economical) aspects of the system, simply because there are different ways to change the system and too many different antagonists which are not colluding and have no common leadership (= my definition of "decentralized").

As such, only status quo can have a majority "vote" by so many non-colluding antagonists.  This is the cause of immutability, and the fundamental property that can make decentralized trustless systems develop monetary belief.

If there is a majority that imposes change, it means that the system was centralized under leadership.

However, you are right that hard forks (new coin creation) can be performed by just any antagonist.  In that case, people vote in the market: but this is not "leadership".  Any antagonist, no matter how small, can do this, and the success of his modification emerges in the market, leaderless.  The old system, however, remains immutable.  (but can die)

Hey, I'm not saying that no crypto system can be developed if it has leadership.  But if it has leadership, it is centralized, period.  Now, for FAR MOST applications, centralization is beneficial, because decentralized systems cannot beat centralized systems in efficiency.  However, in the long run, all centralized systems become corrupt and make a balance between getting out of the market what the market can bear as advantages for the "leadership".  It is only when this corruption is so strong that it kills its own efficiency, and that decentralized systems start becoming competitive.  But new systems with fresh, competent leadership make centralized systems that outcompete all decentralized peers.  Because decentralisation is, in most markets, a huge burden that centralized systems don't have to face.

Facebook and google are great.  No decentralized system can compete with that.  They are not yet corrupt enough for any decentralized system to be able to take their place.  Fiat systems, although heavily corrupt, are STILL way way more efficient for legal and normal payments than bitcoin or whatever clunky crypto payment system.  Yes, there are niches.


1,2,3,3: facebook, google, Apple....

4: because Satoshi had the stupid idea of proof of work. (not stupid for burning seigniorage, but utterly stupid as a decision mechanism).

So, yes.  You can invent another facebook.  You can sell it as fake decentralization.  Like DASH and ethereum, and up to a certain point, bitcoin.  Ebay or OpenBazaar ?  VISA/paypal or bitcoin ?   Skype or jitsi ?   DNS or onion ?  

Decentralized is useful only in those cases where centralized systems cannot go, for legal reasons, because they became too corrupt or other.  So if you want to be successful, yes, make a centralized system.  Much more customers.  If you can sell it and hype it as (fake) decentralized, get in the gullible's money.  That's what gullible money is for.

This.

Every single PoW algorithm has a threshold over which special hardware becomes beneficial.  Economies of scale.  SHA-256 was a particularly poor choice in that respect.  Scrypt didn't hold out much longer.   Cryptonote is holding out, but also has a threshold.  Hell, if you can do it with a computer, made of silicon, of course you can do it with special-purpose silicon !

However, the economies of scale to work out the chip, make prototypes, and put it on a board are such that only from a certain level of performance-over-CPU this becomes interesting.  So on chains that will not go to the moon, CPU/GPU mining may hold out with a sufficiently sophisticated hash function.  From the moment you hope to go to very high prices, and hence, very high block rewards, this investment will always become useful.

Point is, however, why try to get consensus over that with bitcoin ?  If you want that, there are several alt coins *doing that already*. 

You would have to modify bitcoin so profoundly that it is easier just to switch to another crypto that is already up and running than to find any form of consensus over such a profound modification.   In fact, you can do it on your own, and publish a hard fork for bitcoin using such a function.  But your alt coin will not take off.  Better buy an existing one in the market.

Ah.  Me too.  Because it permitted to transgress laws.  But if it is not transgressing laws, we already have good legal payment systems: fiat.  I don't see any, but absolutely any, advantage for people to get involved in other payment systems that are risky, clunky and will, in the end, not allow anything else than what you can do with the legal payment system.  I don't see the use of adding the burden of trustlessness and decentralization (it is a HUGE burden as compared to a swift centralized system) if there's nothing more that you can do with it than what you can do with the existing fiat system.  Bitcoin has been costing between $12 and $8 dollar a transaction the last few years.  It is not scalable, except if you go to a banking system, worse than the fiat banking system.  You take on top of that the high volatility, the risk of being stolen and the burden of having to be extremely careful, because everything is irreversible.  You broadcast your financial transactions open in the blue to the whole world, cryptographically graved in stone ; a fiat bank is way, way more private.

No, honestly, bitcoin doesn't serve any purpose, doesn't bring in the slightest bit of value over fiat payments if you can do it with fiat payments.  The whole of Amazon is a credit card click away.  If there is a problem, I can discuss with my bank.  If someone steals my stuff I can complain, and this is reversed or in any case, guaranteed.   With bitcoin, you are depending on 5 Chinese guys to make your block chain, when they don't get enough fees, they block the network with spam, your payment doesn't get confirmed. 

Why on earth would I want people to adopt such a wasteful, clunky, dangerous and volatile payment system to buy coffee or books on Amazon when everything works much, much better with fiat ?

Only when I cannot do it with fiat.  And banks will try to make you do everything which is legal with fiat.  So, essentially, crypto only has economic value when it is unregulated, illegal or criminal.

There is no reason for Joe to adopt crypto if he doesn't want to go there.  Fiat is much, much better and unbeatable: it is centralized.  You can never build a system with the burden of decentralization and trustlessness that can compete with a centralized system.  So the only ways where such a system makes sense, is if there's no centralized system that can compete.


You don't understand the basic principles of these systems.  Primecoin tried to do something of the kind.  PoW is a very rough concept, that only makes sense to kill seigniorage, but is a horrible idea to secure a cryptographic system.  In fact, it is the worst possible form of cryptography, because the "good guys" and the "bad guys" are on equal footing.  Normally, a cryptosystem gives an advantage to the "good guys" (for instance, those that know a secret key).  PoW is the worst form of cryptographic protection. 


No, it is totally meaningless.  If you do something valuable with PoW, then it is not punishing you if you double-proof.  You get then the "nothing at stake" consensus divergence.

Suppose that I do PoW to sell heat, in such a way that I earn (money wise) exactly the cost of my "mining PoW".  I wouldn't have the slightest problem upscaling a thousand-fold, selling thousand times more heat (in competition with non-PoW heat providers which is a huge market), and redo the block chain.  It is only when the PoW that I have to deliver is actually WASTEFUL and COSTLY that I will not do so.  If I can sell the by-product of my PoW for the same value as what it costs me, I do not care about wasting PoW and it loses entirely its security aspect.


Because these projects are naive and misguided (and yes, me too, I bought some prime coin because I liked the idea before I understood much better these systems).


I hope not for monero's sake !


I don't see the use of it.  If they do so, it is to "mine the speculators".  But bitcoin has no use for the public, if the public can do it with fiat.


Getting rid of state and law is freeing up humanity, not "destabilising" it, but this is difficult to understand for most endoctrinated people.

No.  I want these things abolished.  That's why I like crypto.

But Spoetnik, the real question to you is: why do you want to favor adoption of crypto ?

Suppose that I was saying: "I want pink noise to be integrated with the music industry, because I want greater adoption of pink noise listening".

Would that make sense ?  Why would I want people to listen to pink noise ?  Maybe I'm selling pink noise.  That would be a motive.  Maybe politically, I want pink noise to rule the world.  That also makes sense.

But why would you want people to adopt crypto, and bitcoin in particular ?  What good does that bring ?  To you ?  To people ?

I don't want crypto adoption for crypto adoption.  I think that crypto is great for those things were legal money cannot go, and as I think it is important that a lot of illegal things can happen, I think crypto is great, for that purpose.  

But you ?  Why would you want crypto adoption ?  What good does it bring ? If it goes where legal payments also go, what good is it ?  LIke what good is it to make people listen to pink noise if they listen already to music ?

The point is that if transactions are not scarce, fees will be low, and without tail emission (of which bitcoin is proud), these fees are the only thing that will pay for ledger security.  The ledger is ONLY be secure because it is too expensive to do it over, that is, in order to do it over, one has to spend more than what is potentially at stake (double spends).
As such, the block benefit (reward + fees) must not be significantly smaller (orders of magnitude) than the largest individual transactions it contains ; otherwise, it is beneficial for the emitter of that transaction to redo the chain, orphan the block in which the transaction took place, and double spend it to another of his addresses.

For instance, suppose that you do a 1000 btc transaction in block X.  Suppose that the total reward per bloc is about 5 btc.  That means that mining will not cost more than about 5 btc per block in proof of work, or miners would lose money mining.  If you wait for 10 blocks, your counter party thinks that the transaction is safe.  You obtain what you wanted from your counter party. 
Next, you pay miners 200 btc to redo the last 10 blocks, and insert another transaction of your 1000 btc (to one of your addresses).  For miners, this is profitable: they discard 50 btc of bloc reward and the hash rate they spend on it by redoing the chain, but you pay them 200 btc.


There are ways, but not with PoW. 

Note that the price of bitcoin doesn't really matter in this story.

Ha, coindesk has censored again one of my contributions.  You'll see why:

There are structural problems built into bitcoin from the beginning, and its biggest problem is what has often been hailed as its great virtues: the limited number of coins that will be in circulation when the Sun becomes a Red Giant, and the proof of work scheme that needs the biggest computing effort in history to make a cryptographically secure ledger.
This gigantic computing effort, needed to secure the ledger of all transactions ever, must be financed by something, and that is not a small amount. We are talking about about half a billion dollars a year to keep bitcoin's cryptography up and running. The early years of bitcoin, this has been financed essentially by inflation: the bloc rewards. It is still largely financed by inflation, but since the last bloc halving, fees have become a significant contribution to the financing of security by proof of work of the ledger.
The block halvings are needed to obtain a finite amount of bitcoin when the Sun will swallow the earth ; and to finance the huge computational effort to secure the ledger with proof of work, something else has to pay for it: the scarcity of transactions. If one finds "a solution" to the scarcity of transactions (be it large blocks, or off-chain transactions), bitcoin's economical model crumbles, because there is no scarcity any more that will pay for the huge proof of work.

Because the (poor) cryptographic security of proof of work is simply that it is too expensive to break it. So bitcoin's security is tied to wasting a lot of value. That value has to come from something. As the inflation tax is shrinking, only transaction scarcity can take over. Hence, transactions must remain scarce and expensive, or bitcoin becomes insecure.
And hence, there shouldn't come a "solution" to the problem of transaction scarcity.

The code must not work.  Well, it must work like the old core code, and say it is "BU".  That's all that it must do.  There doesn't need to be any really working BU code inside.  Just old core code, that doesn't signal that way.