The Great Scaling Debate
Bitcoin and its critics
In our What is Bitcoin article we explain how Bitcoin serves as both a store of value and a means of value transfer. We argued that, for bitcoin to continue serving it’s unique purpose, it must remain decentralized. Decentralization does not always lend itself to efficient and cheap value transfer, and so there is tension between these two principles.
This tension exploded in 2017 into a bit of a crypto civil war, between those who believe bitcoin should be primarily a cheap and efficient means of transfer, and those who believe it’s decentralized nature is more important for its long term survival. Tempers flared. Egos raged. Some launched new coins and claimed they were the true bitcoin. One of those people bought bitcoin.com, turned it into a propaganda outlet, and hired an army of ‘bitcoin birds’, paid to broadcast supportive messages on twitter, reddit and the like. It all got a bit confusing and Trumpian. In this article we try to explain what happened.
Block limits
In 2010 Satoshi Nakamoto, who had not yet taken his George Washington-esque retreat from the world of bitcoin, instituted a 1 MB limit to the amount of data included in new blocks added to the blockchain. This was an effort to limit the amount of spam that was being pumped into the blocks, which accounted for the vast majority of blockspace in those early days. That 1MB limit, somewhat arbitrarily implemented, has remained engaged until our present day.
With growing demand for bitcoin, 2017 was the first year in which the 1MB limit was no longer enough to support all transactions. A fee market began to take shape, and what was once an extremely cheap method of value transfer – transactions were often sent with no fee at all and were included in the next block – began to get very expensive. (As an aside, we should note that big blockers supporters would allegedly send spam transactions with unnecessarily high fees on the reg to push out normal transactions, increase fees required, and thus amplify the issue.)
The obvious solution would seem to be simply removing the 1MB limit and replacing it with, say, 8MB. This is called “on-chain scaling”. However, there were, and still are, three major issues with this solution.
- Implementing the change requires what is known as a “hard fork”, a protocol change which requires wallets, businesses, and miners to begin using a new bitcoin program. This requires consensus, because otherwise some percentage of the market will continue running the old program, leading to two different chains.
- There are serious concerns that raising the limit would have a centralizing effect on the bitcoin network.
- None of the best developers agreed with the solution, nor did the bitcoin holding community.
Centralization and Block Size
Bitcoin nodes, users who set their computers to download the bitcoin protocol, verify new blocks, and distribute the blockchain across the internet, initially also acted as miners. Miners are rewarded with bitcoin for successfully processing a block, and so these nodes were compensated for the computer space used. Over time, two things happened:
- Advances in mining technology made GPU mining redundant – computers no longer had the processing power to successfully mine a block.
- The size of the blockchain consistently increased (As of January 1st, 2018 it weighs 157 GB)
In other words, the economic incentive for a node no longer exists, while the real world cost of serving as a node has increased. The 10-12k active nodes at any one time then, are acting out of altruism or enlightened self-interest – i.e. they have substantial bitcoin holdings and know that the health of the network depends on a distributed node base. However, this indirect incentive is not as strong as a direct economic incentive and could easily lead to a tragedy of the commons situation, particularly if the cost of hosting a node continues to increase.
Off-Chain Scaling
The most prolific and accomplished bitcoin developers have taken a stark view of raising the block size due to the centralization effects on nodes as well as long-term unsustainability. When is the block too big? Is it ever too big? Additionally, without overwhelming consensus, a hard fork is too risky a ploy to coerce a $200 billion market into enacting.
The community seems to agree with these developers. The big block coalition, which consisted of some CEOs, some miners, and some second rate developers, were unable to unite the market behind their proposed hard fork. When it launched anyways it was a spectacular, public humiliation for the developers, as it collapsed upon takeoff due to elementary coding errors.
The bitcoin developers have focused on different solutions and we will briefly describe their characteristics.
Segwit is a bit of a Schnoor
The first two scaling technologies are called Segwit and Schnorr Signatures. Segwit is a backwards compatible technology that launched in June of 2017, and is currently being utilized in approximately 15% of network transactions. Segwit essentially removes digital signatures from the blocks and compiles them in a different thread. Transactions using Segwit weigh approximately 50% less than non-Segwit transactions, and so twice the amount of transactions could fit within the existing blocks, should 100% of the network adopt Segwit.
Schnorr signatures are made possible by the introduction of Segwit. The structure of digital signatures are changed, allowing a more efficient verification. It is estimated that Schnorr signatures would reduce the size of transactions by at least 25%.
It bears noting that these scaling solutions provide additional benefits as well. Schnorr signatures increase privacy on the network while providing for a new type of multisig transaction which should have multiple use cases. Segwit finally solves the issue of transaction malleability, a necessary requirement for the third and most interesting scaling solution, the lightning network.
The Lightning Network
The lightning network is a layer two protocol, meaning that it sits on top of bitcoin. The lightning network facilitates transactions that are effectively instantaneous, effectively free, and effectively limitless. The transactions take place between two users from lightning-enabled wallets. Transactions are settled off chain, and the users may, as required, periodically settle final accounts on the main blockchain. Connections need not be direct – transactions may “hop” between wallets, a six degrees of separation effect. Over time, the vision is that the bitcoin blockchain will be utilized for larger transactions and the settlement of accounts, while the lightning network will be utilized for day-to-day activity. This will reduce congestion, but will also make fees less prohibitive, as on-chain transactions are not required for every interaction.
The lightning network white paper was published in January of 2016, and a fully compatible version has now been released. It is an entirely open source project.
Making Changes is Hard. Good.
Throughout the big block v. off chain scaling argument, the refrain that scaling-is-a-pressing -issue-and-must-be-sorted-fast(!) could be heard from big blockers at an escalating crescendo. They additionally went outside the recognized method for making changes to an open source protocol and attempted to dictate a solution, a clear case of the means not being worth the goal.
Big blockers reaction is understandable – fees are getting ridiculously high – but a fundamental misunderstanding of both bitcoin and the community, in our humble opinion. Bitcoin’s decentralization makes making changes hard, as it should be. If it were easy to change, who could have faith in the immutability of their holdings? Who could say 21 million bitcoin is truly a hard cap? If a group could meet in a hotel room and dictate a block increase, that group is a who is to say that single point of failure will not be exploited?
Everyone is united in wanting lower fees. Currently, bitcoin is economically unusable for smaller transactions. However, the decisions made now will have a lasting effect on bitcoin’s future. Scaling must be achieved in a way that does not undermine bitcoin’s decentralized nature or we are fated to become paypal, or worse, Ripple.