If the first day of Scaling Bitcoin focused on fostering constructive dialogue, day two moved the conversation toward defining how that dialogue would proceed following the event and how such discussion could coalesce into a clearer vision for the open-source technology’s future.
Topics of discussion at the event still focused most directly on the larger question of how the bitcoin network could support increased transaction levels. However, these presentations were buoyed by musings on the trade-offs that will need to be weighed should the community want to honor the project’s original democratizing vision.
For example, some of the day’s talks occasionally provided a broad, analytical look at current proposals, delving into the theoretical considerations such as how much capacity the bitcoin network would need to handle at scale and how the metric of scale could be defined.
“When people talk about how bitcoin will have to scale, people throw out something about Visa processing 20,000 transactions per second or something,” Harry Kalodner, of Princeton’s Security and Privacy Research Group, said. “There are other relevant factors though.”
Kalodner went on to suggest using the bitcoin blockchain as a domain name storage system might add 294 million transactions to the network, a figure that doesn’t include expected use cases such as in the Internet of Things. “There are 245 million cars in the US,” he added.
In this light, sometimes contentious approaches to blockchain design such as the Lightning Network, which seeks to add payment channels to bitcoin, and BIP 102, a proposal to increase the blocksize to 2MB, were discussed but with the goal of providing clarity as to their intent.
Paul Sztorc, author of the Truthcoin white paper, spoke about the existential issues facing the blocksize debate, and the need for arguments to be understood as the sum of underlying assumptions that must be qualified and stated clearly.
Sztorc told attendees:
“If someone hasn’t stated what they think the blocksize does for bitcoin, you can’t tell if what they said was right, or if you agree with the purpose they expressed. Not agreeing on the objective function is almost as bad as not agreeing on the constraint.”
Elsewhere, the afternoon sessions featured 12 roundtables held to address community challenges. Some moved toward solutions, such as a proposed set of principals that seeks to help the development community grow.
“We need to have better resources for new entrants to the community, so they can be onboarded to the community without negativity,” one presenter said.
The proposed principles are just one initiative that will be developed prior to being presented at a second Scaling Bitcoin conference in Hong Kong.
To be held in December, the event is slated to feature presentations for scaling the bitcoin network, which will be evaluated against criteria developed at this weekend’s event in Montreal.
Block size considerations
One of the day’s more anticipated talks was by Bitcoin Core developer and BitPay employee Jeff Garzik, whose presentation was entitled “Issues Impacting Block Size Proposals.”
This July, Garzik introduced BIP 102, a patch to the bitcoin software that would raise the blocksize limit to 2MB via a hard fork. The short-term solution would alleviate immediate worries about the network reaching capacity, though the specific pros and cons of the measure were discussed only tangentially.
“The trend is pretty clear that we are headed towards the 1MB block size limit,” Garzik said, warning that user experience on the bitcoin network will decline when blocks hit capacity. This occurrence, he argued, that would necessitate that consumers pay fluctuating fees to execute transactions, something consumers and consumer products aren’t yet equipped for.
“A wall at 1MB creates chaos as fees shift to a new higher equilibrium. When we hit that wall, businesses or users might be incentivized away from bitcoin by high fees,” he said, adding:
“If we fail to get to consensus at all, and it’s incredibly difficult to figure out where to go, and we just sit at 1MB, then that strangles bitcoin growth and adoption because of how toxic we’re making this.”
Garzik suggested that larger financial entities may be inclined to use the bitcoin network, but that they are currently confined to experiments because doing so would instantly cause the network to hit its transaction processing capacity.
The talk perhaps best succeeded at showcasing Garzik’s viewpoint that any and all solutions to bitcoin’s current blocksize dilemma need to be considered.
“A second course correction hard fork is likely. All the world’s coffee payments won’t fit on the blockchain, you must have layer two and all the other scalability solutions,” he concluded.
Payment channels
One of the central considerations of the blocksize debate is how many transactions need to be recorded directly onto the bitcoin blockchain. For example, the Lightning Network, currently in development, proposes updating the protocol to support off-blockchain micropayment channels which would then settle on the blockchain.
Led by Lightning developers Joseph Poon and Thaddeus Dryja, the talk focused not directly on the project, but the implications for not scaling bitcoin to meet expected demand.
“The underlying assumption in these failure models is the assumption everyone would use bitcoin if they could,” Dryja said. “We’re assuming everyone wants to use this.”
Dryja suggested that the community should start first by drawing rough parameters on the size of blocks it is willing to “agree are crazy”, ruling out sizes of 1 kilobyte or as big as 1 petabyte.
In between, Dryja spoke about a “bathtub” of outcomes, hypothesizing that if the block size were extremely limited, it would grow only 50MB annually.
“[In this scenario], you can run a full node on your phone. But 10 large institutions have private keys. This is a failure because users don’t have the keys. That was the promise of bitcoin to be your own bank,” he continued.
In turn, Poon spoke about the technical limitations of the blockchain for micropayments today, suggested that the cost of using the network for a transaction is 3 cents, making smaller transactions cost-prohibitive.
The talk was followed by a presentation by Come Plooy of Amiko Pay, a project that aims to be an implementation for Lightning that deals with routing data between payment channel nodes.
Block size research
The more experimental aspects of blockchain mechanics were discussed in a series of talks that included Ethereum researcher Vlad Zamfir, Blockstream co-founder Mark Friedenbach and BitFury chief information officer Alex Petrov, among others.
Zamfir led off the section with a talk on “sharding” blockchains, a process by which nodes hold only a subset of the state and the blockchain. In bitcoin, nodes are currently expected to hold a complete version of the blockchain.
“[With sharding], instead of everyone redundantly doing the same work, we’re going to share the load but still have an economic assurance even though we’re not going to validate every transaction,” Zamfir explained. “We can get orders of magnitude of transactions per second if everyone isn’t validating everything.”
Under the system, miners are assigned shards of the blockchain, he said, before explaining some of the issues inherent in coordinating parties within such a system.
“We need to sample the mining power, which presents a problem,” he continued. “We need to split the state space into shards, we need to process transactions within shards, and then deal with attack vectors where not everyone is checking everything.”
Elsewhere, Friedenbach spoke about the benefits of having a blocksize on a blockchain network, explaining why the measure was enacted, as well as some of the problems that have resulted from the decision.
“We want to do this because in the early days of bitcoin there are often certain kinds of transactions there are ways to slow down a validator by using a non-standard transaction or filling up a block with tons of spam,” he said.
Words of warning
While mostly well-received, the conference ended on a controversial note, with remarks by Nicholas Negroponte drawing the ire of more anti-government members of the bitcoin community.
Founder of the MIT Media Lab and an investor in WIRED Magazine, he urged the bitcoin community to weigh the societal implications of the technology when making decisions.
“What is the difference between a mission and a market?” he asked. “If you’re after a market, be my guest, but what is going to make bitcoin work is when you think of it as a mission. It’s going to make the world a better place in a lot of different ways.”
Negroponte also chastised the community for its inclinations to treat the technology as a “get rich quick scheme” and to hoard bitcoin due to its expected future value.
“What’s wrong with being run by the government? If you think the government can’t run anything, go to Finland and ride a train. There are certain things that governments can run,” Negroponte said, concluding:
“Screwing it up is the people. It’s not the technology.”
Presentations from the day’s event can be found in both written and video form.
Images via Pete Rizzo for CoinDesk