Scaling Bitcoin: update Bitonic (part 1)

The debate surrounding Bitcoin's scalability, the various solutions proposed so far and the varying opinions within the Bitcoin community have helped elucidate the elements at play.

Our observations and conclusions on the current state of affairs.

As we have emphasized in previous communication: Bitcoin is especially valuable as an innovation due to its decentralized character. Therefore, as we have indicated before, we will not support Bitcoin XT; this approach to solving the blocksize issue raises many questions concerning its impact on Bitcoin's decentralization. We wrote more on this here.

This article by Aaron van Wirdum, journalist for Bitcoin Magazine, provides a good summary of all proposals and solutions that are currently being worked on.

What is the best solution? That is a difficult question involving many factors, a question the Bitcoin community is currently hard at work trying to answer. And so are we. We traveled to Montreal, Canada, on September 12 to attend the Scaling Bitcoin Workshops.

Workshops Scaling Bitcoin - Part I

A global open-source community coming together to discuss solutions is in and of itself something to be proud of and an absolute asset for Bitcoin; initiatives like this are, needless to say, of great value.

The purpose of this academic-style conference was clarifying what requirements a solution to the blocksize debate should meet. The focus was on determining what factors should be considered when designing solutions. The focus was not on presenting actual solutions, but rather provide input for future solutions — as Jeff Garzik put it during his presentation:

I look at the next workshop as output, looking to generate proposals, this first workshop is input into some of those proposals.

The approach of this initial workshop has lead to a thorough analysis of all relevant factors regarding the blocksize. We will briefly summarize and highlight some of the most interesting talks below.

Mining gaps

When the block rewards decreases, miners will become increasingly dependent on transaction fees for their income. At the same time there are certain costs involved in generating a block. When the cost to create a block is higher than what income will be generated through transaction fees it will be beneficial for miners to temporarily stop mining until enough transactions have been done, so that transaction fees do make mining a block worthwhile.

This could cause a delay in the time it takes transactions to be included into a block and receive those much-wanted confirmations. In addition, it can be considered a threat to the security of Bitcoin. When an attacker is prepared to mine sub-optimally for some extend of time, while other miners are waiting for transaction fees to increase, the attacker will need an increasingly small percentage of the total hashrate to produce more than half of all blocks. This increases the likelihood of a 51% attack. Miles Carlsten describes these periods where mining a block will only result in a monetary loss for the miner as 'gaps' and posits that this phenomenon should be taken into account when designing for scalability.

(Mind the Gap: Security Implications of the Evolution of Bitcoin Mining)

Determining blocksize

In this presentation, rather than attempt to answer whether we need a blocksize and what the limit should be, Peter R. Rizun postulates that a fee market for transactions will exist when there is no blocksize limit. This is due to the fact that miners, aside from adding blocks to the blockchain, have another task: producing a digital commodity called blockspace. The amount of blockspace that is available depends on the demand for this commodity. When demand is high, transaction fees will rise as people are willing to pay more to have their transactions confirmed. Miners might subsequently choose to produce more blockspace, as it is now (more) profitable for them to do so, causing transaction fees to drop… and so on.

Apart from the debate on what size blocks should be, it is interesting to consider the fee economy that could arise when supply and demand of blockspace are able to interact freely.

(Alternatives to block size as aggregate resource limit)

Lightning Network

Aside from various solutions that revolve around varying Bitcoin's blocksize, there are solutions that move the majority of Bitcoin transactions off-chain so that more transactions can be done while still having relatively small blocks. Worth mentioning here is the Lightning Network; a network of payment channels where numerous transactions can take place between parties in a completely trustless way, without burdening the Bitcoin blockchain and without threatening Bitcoin's trustless character. Using the Lightning Network, only a few transactions take place on the Bitcoin blockchain: one transaction required for opening a payment channel and one to close the payment channel.

While on this topic, we would also like to mention Corné Plooy's work in this area. Corné has been developing Amikopay since 2011. The methods used therein are similar to how the Lightning Network works. Corné describes it as:

A decentralized network for fast, secure, cheap and privacy-friendly off-blockchain Bitcoin transactions.

In Conclusion

Approaching the problem from various perspectives has expanded our knowledge on the possibilities and impossibilities surrounding Bitcoins's scalability. Specialists from various backgrounds have contributed to providing input for solutions; Bram Cohen, architect/invent of BitTorrent — Adam Back, architect/inventor of Proof-of-Work — Wladimir van der Laan / Gavin Andresen, Core-developers Bitcoin — Peter Wuille / Gregory Maxwell, Core-developers Bitcoin & Blockstream.

A lot of ground has yet to be covered to come to consensus and realize a fitting solution. Meanwhile, the Bitcoin community steadily continues innovating.

Have the workshops in Montreal resulted in a solution? No. However, this was not the goal. The goal was to analyze the problem and determine what factors need to be considered when designing solutions. These new insights will be used to present solutions at the next conference, this weekend in Hong Kong.

We will be travelling to Hong Kong (December 6–7) to attend the second Scaling Bitcoin workshop. Afterwards, we will report back on all the interesting talks there!