The word "decentralization" is so often used in discussions about the scalability of Bitcoin that its true meaning seems to have been forgotten. Although decentralization is an important property of Bitcoin, many people have reached the point of making this word one fetish which attributed all kinds of mythical properties. "More decentralization" is the panacea that will solve any problems that Bitcoin will face. The word is used as a whistle closing any productive discussion: just accuse your opponent of trying to "centralize Bitcoin" and immediately hordes of angry trolls will attack mercilessly, while every vestige of rationality is thrown out the window .
Since the "decentralization" is the justification most frequently used to maintain restricted the size of the blocks, it is time to have an accurate and succinct definition of this word.
Before offering my own definition, I will explore the ways in which it is usually (mis) use of the term. When it used by restrictionists artificially the block size, "decentralization" is generally understood to mean "number of nodes":
"The operation is cheaper node, there will be more nodes on the network. Therefore, there is no other more important factor to keep costs as low as possible nodes. Only then Bitcoin can achieve maximum decentralization. "
This is a grotesquely dimensional view of how the network operates, also it reveals a complete lack of understanding of what gives value to Bitcoin.
At the most basic level, this view tends to confuse "decentralized" with "distributed". While a distributed computing network can be very robust in practice Bitcoin more like a decentralized network to distributed. But this is an advantage because it allows network members to participate at all levels.
If the barrier to participate in the network Bitcoin leaves out anyone who does not hold a node, we exclude all potential participants lacking funds, access to hardware, enough free time or know - how (or inclination to invest time, money and effort) to configure and maintain a node. Such a high barrier hinders adoption .
Fortunately, Bitcoin does not work that way. Participation levels ranging from mining farms several million dollars on one end to private keys printed on paper in the other. The network gets stronger allowing each user to participate in the match level where his inclinations and capabilities . Specialization of the participants is what pushes Bitcoin to become the best version of himself.
Those with this view shortsighted of decentralization also make the assumption wrong that the best and only way that the network has more nodes is to minimize the cost of operating a node. It does not consider the possibility that if Bitcoin was five times more users, you might also have five times more nodes. Even if the total percentage of operators of nodes decreased, and have five times more users only doubled by two the number of nodes, the end result would be the Bitcoin nodes that exist now.
New users are not attracted to Bitcoin by how cheap it is to operate a node. Mostly they attracted to Bitcoin because it provides useful , and only after this initial attraction that some of these new users choose to operate nodes. The only logical to absorb the increased demand option is to maximize the usefulness of Bitcoin, which is achieved with fast and affordable transaction and dulls with slow and costly transactions. It should be noted that those interested in the operation of relevant network nodes are almost exclusively attracted to the utility users rather than speculation, and adding more users attracted by the utility also benefits the other users. It makes no sense to favor one at the expense of others.
The vision simplistic of decentralization also erroneously assumes that the value of Bitcoin depends on their degree of decentralization. "Let us make a more decentralized Bitcoin , " says this argument, "and will be more valuable." This does not withstand scrutiny because Bitcoin presumed that a thousand users each operating its own node is more valuable than a Bitcoin with one billion users but users rate operators nodes only one percent. Empirical evidence suggests the opposite: Bitcoin in 2009, when most users also were complete nodes was significantly less valuable than in 2017, with millions of users and operators rate 1% lower nodes .
I propose to use a definition of the word "decentralized" much more accurate from the operational point of view: "lacking a central point of control" . Under this definition, it makes no difference if over time it becomes more expensive to compete in mining, or increase the cost of maintaining a node. The key element is that Bitcoin has no central points of monitoring, planning and control (and therefore is less susceptible to catastrophic failures).
According to this criterion, Bitcoin is now more decentralized than ever. When MtGox handled 80% of all trading volume in exchange sites, the decline spread shock waves throughout the community and caused great turmoil in the market. Today there are probably hundreds of exchange sites, and easily tolerates specific market failures.
A persistent myth is that mining tends to be centralized increasingly, and that this poses a threat to Bitcoin. This myth is fueled by photos of large spaces dedicated to mining and the presence of a company with a majority stake in the hardware market for mining. The myth does not take into account that these large mining represent increasingly smaller percentage of the network, and obtaining mining hardware is easier and buyers are less susceptible to fraud than ever before. Mining is becoming increasingly decentralized geographically, with large-scale mining projects emerging in dozens of countries worldwide.
Sort something like "centralized" or "decentralized" is a binary view of the matter. Regardless of the number of data processing centers held by PayPal, and how geographically dispersed are, it will always be a centralized system, because the corporation PayPal itself is a central point in case of failure compromises the entire system. Bitcoin, by design, can only be a decentralized system. There may be different degrees of decentralization, but it is incorrect to say that this or that proposal Bitcoin can turn into a centralized system. We must speak of a "decentralized" or "less decentralized" system, but still, all indicators tell us that Bitcoin has become more decentralized the time .
For whom the idea of "decentralization" has become a fetish, more decentralization is always better, and less is always fatal. But the good engineer should always consider the cost-benefit ratio . If you put financial sovereignty within the reach of thousands of millions of people carries a lower percentage of users who maintain nodes, that is an acceptable cost. If only I could have ten thousand nodes, it would better serve a billion people a million.
Finally, when we consider the main points whose failure can be transmitted to the entire system, we can not forget one aspect of Bitcoin remains to be decentralized, but also becomes more decentralized every day. I am referring, of course, the development of the protocol.
The existence of a single monolithic development team (today is Bitcoin Core , but could be any), represents one of those dangerous focal points. No matter how many people contribute, and how open to programmers outside the main equipment is the development process; have one group in charge of spreading the "official" software and a network of nodes that automatically download is a big threat to Bitcoin. No matter what the group 's intentions are or how wonderful is your code, the focus is there. You can trust that PayPal will not abuse your personal information, but if PayPal suffers a violation of their databases will have to face the consequences anyway.
And MtGox resulted in hundreds of new exchange sites, and the number of mining pools has increased from a few to a few dozen, and the Bitcoin mining moves away from China and becomes increasingly global and celebrate all that we must also celebrate the current proliferation of implementations of Bitcoin. Bitcoin say should be decentralized in all other aspects as dependence on one team defends developers is patently absurd . If Segwit ends up being discarded, instead of seeing the non-adoption of Segwit by the market as a system failure we must take it as proof that the system works as it 's supposed to work. The battles daily can be frustrating, but it would be much worse if such a radical change to the code of Bitcoin could be implemented without debate.
Decentralization fetishists consider the notion of "decentralization" as separate from Bitcoin and as a goal in itself. Instead of recognizing that the goal is to be good currency Bitcoin, they assume that the goal of Bitcoin is being decentralized. But the goal of Bitcoin has always been to become good money, and decentralization -the elimination of central points that endanger the entire system- is and always will be a means to that end . Since the value of decentralization in Bitcoin is related to the ability to prevent the emergence of central points that can transmit a failure localized to the entire network, the issue of decentralization is more a "yes / no" than " how much?"
More important than maximizing the number of nodes in the network it is to be safe at dangerously centralized points, and the only non - authoritarian way to accomplish this is by promoting competition .
Since the "decentralization" is the justification most frequently used to maintain restricted the size of the blocks, it is time to have an accurate and succinct definition of this word.
Before offering my own definition, I will explore the ways in which it is usually (mis) use of the term. When it used by restrictionists artificially the block size, "decentralization" is generally understood to mean "number of nodes":
"The operation is cheaper node, there will be more nodes on the network. Therefore, there is no other more important factor to keep costs as low as possible nodes. Only then Bitcoin can achieve maximum decentralization. "
This is a grotesquely dimensional view of how the network operates, also it reveals a complete lack of understanding of what gives value to Bitcoin.
At the most basic level, this view tends to confuse "decentralized" with "distributed". While a distributed computing network can be very robust in practice Bitcoin more like a decentralized network to distributed. But this is an advantage because it allows network members to participate at all levels.
If the barrier to participate in the network Bitcoin leaves out anyone who does not hold a node, we exclude all potential participants lacking funds, access to hardware, enough free time or know - how (or inclination to invest time, money and effort) to configure and maintain a node. Such a high barrier hinders adoption .
Fortunately, Bitcoin does not work that way. Participation levels ranging from mining farms several million dollars on one end to private keys printed on paper in the other. The network gets stronger allowing each user to participate in the match level where his inclinations and capabilities . Specialization of the participants is what pushes Bitcoin to become the best version of himself.
Those with this view shortsighted of decentralization also make the assumption wrong that the best and only way that the network has more nodes is to minimize the cost of operating a node. It does not consider the possibility that if Bitcoin was five times more users, you might also have five times more nodes. Even if the total percentage of operators of nodes decreased, and have five times more users only doubled by two the number of nodes, the end result would be the Bitcoin nodes that exist now.
New users are not attracted to Bitcoin by how cheap it is to operate a node. Mostly they attracted to Bitcoin because it provides useful , and only after this initial attraction that some of these new users choose to operate nodes. The only logical to absorb the increased demand option is to maximize the usefulness of Bitcoin, which is achieved with fast and affordable transaction and dulls with slow and costly transactions. It should be noted that those interested in the operation of relevant network nodes are almost exclusively attracted to the utility users rather than speculation, and adding more users attracted by the utility also benefits the other users. It makes no sense to favor one at the expense of others.
The vision simplistic of decentralization also erroneously assumes that the value of Bitcoin depends on their degree of decentralization. "Let us make a more decentralized Bitcoin , " says this argument, "and will be more valuable." This does not withstand scrutiny because Bitcoin presumed that a thousand users each operating its own node is more valuable than a Bitcoin with one billion users but users rate operators nodes only one percent. Empirical evidence suggests the opposite: Bitcoin in 2009, when most users also were complete nodes was significantly less valuable than in 2017, with millions of users and operators rate 1% lower nodes .
I propose to use a definition of the word "decentralized" much more accurate from the operational point of view: "lacking a central point of control" . Under this definition, it makes no difference if over time it becomes more expensive to compete in mining, or increase the cost of maintaining a node. The key element is that Bitcoin has no central points of monitoring, planning and control (and therefore is less susceptible to catastrophic failures).
According to this criterion, Bitcoin is now more decentralized than ever. When MtGox handled 80% of all trading volume in exchange sites, the decline spread shock waves throughout the community and caused great turmoil in the market. Today there are probably hundreds of exchange sites, and easily tolerates specific market failures.
A persistent myth is that mining tends to be centralized increasingly, and that this poses a threat to Bitcoin. This myth is fueled by photos of large spaces dedicated to mining and the presence of a company with a majority stake in the hardware market for mining. The myth does not take into account that these large mining represent increasingly smaller percentage of the network, and obtaining mining hardware is easier and buyers are less susceptible to fraud than ever before. Mining is becoming increasingly decentralized geographically, with large-scale mining projects emerging in dozens of countries worldwide.
Sort something like "centralized" or "decentralized" is a binary view of the matter. Regardless of the number of data processing centers held by PayPal, and how geographically dispersed are, it will always be a centralized system, because the corporation PayPal itself is a central point in case of failure compromises the entire system. Bitcoin, by design, can only be a decentralized system. There may be different degrees of decentralization, but it is incorrect to say that this or that proposal Bitcoin can turn into a centralized system. We must speak of a "decentralized" or "less decentralized" system, but still, all indicators tell us that Bitcoin has become more decentralized the time .
For whom the idea of "decentralization" has become a fetish, more decentralization is always better, and less is always fatal. But the good engineer should always consider the cost-benefit ratio . If you put financial sovereignty within the reach of thousands of millions of people carries a lower percentage of users who maintain nodes, that is an acceptable cost. If only I could have ten thousand nodes, it would better serve a billion people a million.
Finally, when we consider the main points whose failure can be transmitted to the entire system, we can not forget one aspect of Bitcoin remains to be decentralized, but also becomes more decentralized every day. I am referring, of course, the development of the protocol.
The existence of a single monolithic development team (today is Bitcoin Core , but could be any), represents one of those dangerous focal points. No matter how many people contribute, and how open to programmers outside the main equipment is the development process; have one group in charge of spreading the "official" software and a network of nodes that automatically download is a big threat to Bitcoin. No matter what the group 's intentions are or how wonderful is your code, the focus is there. You can trust that PayPal will not abuse your personal information, but if PayPal suffers a violation of their databases will have to face the consequences anyway.
And MtGox resulted in hundreds of new exchange sites, and the number of mining pools has increased from a few to a few dozen, and the Bitcoin mining moves away from China and becomes increasingly global and celebrate all that we must also celebrate the current proliferation of implementations of Bitcoin. Bitcoin say should be decentralized in all other aspects as dependence on one team defends developers is patently absurd . If Segwit ends up being discarded, instead of seeing the non-adoption of Segwit by the market as a system failure we must take it as proof that the system works as it 's supposed to work. The battles daily can be frustrating, but it would be much worse if such a radical change to the code of Bitcoin could be implemented without debate.
Decentralization fetishists consider the notion of "decentralization" as separate from Bitcoin and as a goal in itself. Instead of recognizing that the goal is to be good currency Bitcoin, they assume that the goal of Bitcoin is being decentralized. But the goal of Bitcoin has always been to become good money, and decentralization -the elimination of central points that endanger the entire system- is and always will be a means to that end . Since the value of decentralization in Bitcoin is related to the ability to prevent the emergence of central points that can transmit a failure localized to the entire network, the issue of decentralization is more a "yes / no" than " how much?"
More important than maximizing the number of nodes in the network it is to be safe at dangerously centralized points, and the only non - authoritarian way to accomplish this is by promoting competition .