The long, long history of the discussion about the scalability of Bitcoin
The dispute over the question of how Bitcoin should scale did not begin in 2015. He didn’t start in 2013 either, and also not in 2011. He is as old as Bitcoin and was already there before Satoshi published the first version of the cryptocurrency. And the arguments have not changed so much since then.
Sometimes debates develop, sometimes not. In the case of the dispute over Bitcoin scaling, the debate has been rotating for years around the same pipe. I climbed into the archives for you to research how the discussion started.
a:hover imgbox-shadow:0 0 20px 5px rgba(255,0,0,0.6);
2008 James A. Donald
Many people know that the dispute over the blocksize did not come out of nowhere. However, few people know that the debate goes back to November 2008 when Satoshi has announced Bitcoin in Cryptography Mailing List for the first time.
The first answer, the first idea, which was expressed to Bitcoin, is the first spark ever that the Bitcoin paper inflamed-that was literally, the statement “It does not seem to be scaled.“The Scalability was at the very beginning of the Bitcoin discussion. James A. Donald, a Canadian Cypherpunk, weapon and anarchist, was the first person who replied Satoshi. He wrote:
a:hover imgbox-shadow:0 0 20px 5px rgba(255,0,0,0.6);
We need such a system immensely, but as I understand your proposal, it doesn’t seem to scale the size it takes. […] In order to recognize and reject a double spending in good time, you have to know all the past transactions of the coin, which, simply implemented, demands that every knot have the most past transactions. If hundreds of millions of people run transactions, it means a lot of bandwidth ..
To date, Satoshi’s answer is often cited as an argument that Bitcoin can very well scale:
a:hover imgbox-shadow:0 0 20px 5px rgba(255,0,0,0.6);
Long before the network is even as large, it should be sure for the users to use Simplified Payment Verification to recognize double spends. This only needs the header of the block chain, which is about 12 kb a day.
In Satoshi’s vision, only miner full nodes operate. These full nodes are profitable because they create bitcoins. Satoshi had no problem with the prospect that these nodes become large data centers:
In the beginning, most users will operate network nodes, but if the network grows beyond a certain point, this will only make specialists with server farms with specialized hardware.
a:hover imgbox-shadow:0 0 20px 5px rgba(255,0,0,0.6);
In contrast to James A, Satoshi also looks at the requirements for the bandwidth. Donald not as a basic problem:
The bandwidth doesn’t have to be as prohibitive as you think. A typical transaction needs about 400 bytes (ECC IS pleasantly compact). Every transaction must be propagated twice, so let’s say 1 KB per transaction. Visa processed 100 million transactions a day in 2008. That would need 100 GB of bandwidth, which corresponds to the size of 12 DVD or 2 HD films, or about $ 18 in bandwidth at current prices.
a:hover imgbox-shadow:0 0 20px 5px rgba(255,0,0,0.6);
If the network is getting so big, it will take a few more years, and until then it will probably no longer be a big deal to send 2 HD films over the network.
Usually the quote ends at this point. It is gladly used to show that Satoshi intends to scale Bitcoin Onchain, and was optimistic that this is possible up to a visa-like level. However, it is interesting to read on at this point. Because James A. Donald’s answers take a lot from the one, which is still being discussed in 2016.
a:hover imgbox-shadow:0 0 20px 5px rgba(255,0,0,0.6);
First, he begins with a somewhat confusing statement about governments and economic actors to express his dissatisfaction with the fact that Satoshi plans to have Bitcoins mines through large server farms:
If a small number of entities generate new coins, this is more resistant to attacks by a state as a single publisher. But the government also regularly attacks financial networks. The financial collapse that we are currently experiencing is a direct result of one of the youngest of these attacks.
a:hover imgbox-shadow:0 0 20px 5px rgba(255,0,0,0.6);
For James A. Donald was the only subsequent solution to keep the requirements for mining nodes as low as possible:
I think we have to ensure that the data load and bandwidth minimize money editors. The smaller the requirements for data storage and bandwidth for money editors, the more resistant the system against the type of attacks by the government, the financial networks recently suffer.
The comparison with credit cards that Satoshi pulled went. Donald not far enough:
You have to go where bank cards don’t go. About File Sharing. This requires extremely cheap transactions, several seconds per second, all day ..
a:hover imgbox-shadow:0 0 20px 5px rgba(255,0,0,0.6);
Donald saw only one way to operate such payment models with bitcoins:
… we need layers of money on Bitcoin, which enable the tiniest transactions, a hundred thousandth of as small as the smallest coins, and enable anonymity, just as Chaum’s e-money is a layer of bank accounts.
Today this method is called Payment Channel. It was about 21.Co and Satoshipay realizing, as far as I know not with great success. The great hope of scaling Bitcoin sustainably is currently on the Lightning Network that is supposed to be a network of such payment channels.
a:hover imgbox-shadow:0 0 20px 5px rgba(255,0,0,0.6);
As for the supporters of the Lightning Network, Bitcoin for James A is. Donald more of a settlement network than a means of payment:
We can build a private layer on it-money to pay, and Chaum’s money, both based on bitgold coins, similar to the US banking system before 1915 layers of money and seem to have built on gold coins, and in fact we have to build up such layers on Bitcoin in order to press the transaction costs to a level that enables microtransactions and to control the bandwidth ..
a:hover imgbox-shadow:0 0 20px 5px rgba(255,0,0,0.6);
With these quotations, more or less, we have the essence of the entire scaling debate. A few months before Satoshi sent the first Bitcoin transaction to Hal Finney.
2010 Hal Finney
In December, Hal Finney, another long-term cycle and solid cryptographer, spoke out for Bitcoin banks:
In fact, there is a very good reason that Bitcoin banks exist and issue their own digital cash that can be exchanged for Bitcoins. Bitcoin itself cannot scalate far enough to broadcast every financial transaction in the world to everyone and to add to the block chain block. There must be another level of payment systems that is easier and more efficient.
a:hover imgbox-shadow:0 0 20px 5px rgba(255,0,0,0.6);
For Hal Finney, Bitcoin was less a means of payment than a Fareness. And like any other currency, the payment of middlemen, which he called banks, could be made:
Bitcoin banks solve these problems. You can work as the banks did before the currencies were nationalized. Different banks have different policies. Some are more aggressive, others are more conservative. Some would go to partial reserve, while others can have a 100 percent cover by bitcoins.
Bitcoin, as the network of miners, confirm the transactions on the blockchain, can be like an anchor in a network of banks:
a:hover imgbox-shadow:0 0 20px 5px rgba(255,0,0,0.6);
I think this will be the ultimate fate of Bitcoin-to be a “high-power money”, which serves banks as a reserve currency to hand over your own digital cash. Many Bitcoin transactions will take place between banks to settle NET transfers. Bitcoin transactions between individuals will be as rare as … Well, how purchases with Bitcoin are today.
The answers that Hal Finney received on this comment give a little foretaste of how poison the topic can be. A Bitcoint talk user said:
That sounds like a fractional reserve system, one of these problems of the Fiat money that is supposed to avoid Bitcoin.
a:hover imgbox-shadow:0 0 20px 5px rgba(255,0,0,0.6);
But at that time nobody could say how big and hot the debate would soon be.
2013 Gavin Andresen, Peter Todd, Mike Hearn, Gregory Maxwell
At the end of January 2013, the topic of scalability began to cook on BitcoinTalk. During this time, some people began to see the hard blocksize limit of 1 MB as a problem and suggested lifting or dissolving it with a hardfork.
For example, one user called for the limit to change the limit as soon as possible:
We now have to discuss how we change the limit before we come across it. A quick look at the blockchain shows that many blocks already exceed 300 kb. Nobody can seriously deny that Bitcoin grows rapidly and that we will meet the 1 MB roof relatively soon.
a:hover imgbox-shadow:0 0 20px 5px rgba(255,0,0,0.6);
The 1 MB limit was less a practical problem, because one of the expectations and promises: With 1 MB blocks, only 7 transactions fit on the blockchain. In fact, it’s more 3-4. Such a number of transactions is far too little for one thing that was gladly advertised in early 2013 as a world currency of the future.
But already in this thread that radical opposition to the increase in blocksize, which has succeeded in preventing them, has been visible to this day. The user DA2CE7 explains why a hardfork that increases the block size is a fundamental problem:
a:hover imgbox-shadow:0 0 20px 5px rgba(255,0,0,0.6);
This was discussed again and again. There is a hard limit in the protocol, and it is as difficult to change the total number of coins: more or less impossible.
Many people have invested in Bitcoin assuming that the hard limits in the protocol will not be changed.
Even if a super majority wants the change. A significant number of people (enclosed me) will reject this.
At this point, Gregory Maxwell joins the debate. Maxwell is now one of the pioneers of those who vehemently reject a Hardfork to increase the blocksize. He presents the arguments that he should present again and again in the following years:
a:hover imgbox-shadow:0 0 20px 5px rgba(255,0,0,0.6);
Without a sharp limitation of the maximum block size, there is currently no reasonable reason to believe that Bitcoin will be sure in any way if the subsidies have disappeared for miners.
Bitcoin is valuable because it is scarce. One of the important scarcities is the limited amount of coins, another the limited amount of block space: Limited space on the blockchain creates a market for transaction fees that are financed with what is necessary to protect the chain robust from enemy attacks.
a:hover imgbox-shadow:0 0 20px 5px rgba(255,0,0,0.6);
But Maxwell has further concerns:
With gigabyte-sized blocks, Bitcoin can no longer work sensibly in a decentralized way: Only a small, selected group of a few thousands of large banks would have the means and motifs to participate in the validation.
The fear that larger blocks will destroy, which is considered the most important quality of Bitcoin – decentralization – was already there in 2013.
This year the capacity of Bitcoin was still defined by soft limits. These limits could be raised by the miners by a Softfork. When the transaction volume rose in early 2013, the developers raised the limit of 250 kilobytes.
a:hover imgbox-shadow:0 0 20px 5px rgba(255,0,0,0.6);
During this time, Gavin Andresen began to explore and discuss ideas how to deal with the blocksize in the long term. One of his suggestions was to replace the hard limit with a flowing limit that ensured that the size of the blocks could adapt to the demand. Peter Todd vigorously warned of this idea.
A lot was recently said to raise the Block Size Limit, and I am afraid that only a few people understand the perverse incentives that the miners have with regard to blocks that are so great that not everyone can process them in the network in a special way that they lead to centralization inevitable inevitable.
a:hover imgbox-shadow:0 0 20px 5px rgba(255,0,0,0.6);
The scenario that the miners increase the blocksize again and again in order to push the weakest nodes from the network, until it is a completely centralized network, has already been thrown against Bitcoin Unlimited’s approach to solve the blocksize problem.
Either way, the miners, especially the largest miners, make the most profit when the blocks they produce are large enough that fewer than 100 percent of the network, but more than 50 percent they can process.
According to Todd, this leads to “to the centralization of the mining capacit.”
a:hover imgbox-shadow:0 0 20px 5px rgba(255,0,0,0.6);
At this point of the debate, Gavin Andresen explained his view of things. His comment may show more clearly than anything else, as the blocksize debate divided the Bitcoin community into two camps, which at that time already existed, but unlike today, it was still able to talk to each other properly:
I am firmly convinced that we should not aim to make Bitcoin a “high power money” that only creates 7 transactions per second.
I agree with Stephen Pair – that would be an extremely centralized system.
a:hover imgbox-shadow:0 0 20px 5px rgba(255,0,0,0.6);
One side wants to avoid the centralization of nodes and miners, the other the centralization of transactions. Andresen made it clear where his priorities were:
I mean we should take care of the users first. What do the users want? They want low fees and quick confirmations. Let us plan for this goal, because it is the users that give Bitcoin a value in the end.
After that, the discussion that we know today in the bigger. On the one hand, Gavin Andresen, Mike Hearn and many users, on the other side Peter Todd, Gregory Maxwell and others. Both sides have excellent arguments and still discuss with each other at that time. Pieter Wuille, who occupies a middle position, as well as many other users and developers. At that time it seems to be largely consensus that you have to raise the limit sooner or later, but how and when – there was no agreement on that.
a:hover imgbox-shadow:0 0 20px 5px rgba(255,0,0,0.6);
Maybe a user got to the point:
Basically, it’s about whether Bitcoin is a payment network or a value memory.
It is the old, long, never -ending story: What are the priorities? Bitcoin as a means of payment – or as a value memory? Should be nodes decentralized – or transactions? Are there any contradictions between these two goals – and where are they? And what can a compromise between the two positions look like?
These questions were discussed long, controversial and intelligently on Bitcointalk in early 2013. The corresponding thread came to an end after 26 pages. The debate about it has not yet come to an end. It was only louder and heated.
a:hover imgbox-shadow:0 0 20px 5px rgba(255,0,0,0.6);