Billed as a potential venue for debate on more heated issues surrounding the long-term viability of the bitcoin network, Scaling Bitcoin saw a who’s who of developers decamp to Montreal to talk about the underlying technical issues facing bitcoin yesterday.
Often reduced in conversation to mentions of its price or market cap, Scaling Bitcoin succeeded at showcasing the breadth of challenges posed by bitcoin’s approach to incentivizing disparate parties to maintain a common and equally beneficial distributed resource.
Held in Montreal, day one of the much-anticipated event focused more on how incentivizes for network participants should be balanced and less on the positives and negatives of any of the competing proposals – whether Bitcoin Core, its alternative BIP proposals or Bitcoin XT.
The more vitriolic parts of the discussion were confined to afternoon workshops on topics including the level of trust and privacy the bitcoin network requires between users and how relations between miners and developers should be managed.
Held under Chatham House Rules, the discussions were transcribed, though no names were able to be assigned to the content. Nonetheless, some of the more direct statements to the significance of bitcoin’s challenges were made during the presentations on these roundtables.
There, one prominent academic appealed to the overall spirit of the event, stating:
“In bitcoin, technical decisions actually have real economic consequences. Science compels us to hold these conversations to a higher standard, but those technical decisions should be made by actual measurements and not appeals to emotion.”
Others spoke to the value of the day’s denser, earlier topics including the need for a greater knowledge of how often inefficiencies in the mining network – such as orphan blocks not ultimately included in the chain – are produced, and the communications delays that result in miners in different locations globally receiving delayed information about the network state.
Elsewhere, the need for continued communication among stakeholders at events such as Scaling Bitcoin was addressed.
“It’s not just about changing the blocksize,” a participant said. “It’s about proving that when critical issues arise we can resolve them.”
Perhaps the most newsworthy event of the day came during the “Testing, Simulation and Modeling” section of the day’s content when Cornell computer science post-grad Ittay Eyal presented Bitcoin-NG, a new proposed solution to scaling the bitcoin network.
Developed by Adem Efe Gencer, Emin Gün Sirer and Robbert Van Renesse, Bitcoin-NG seeks lower latency, higher throughput and better security on the bitcoin network by proposing changes to the bitcoin mining process.
The proposal recommends breaking up the process by which miners are provided both a reward for finding a “nonce”, the arbitrary number that decides who wins the 25 BTC reward distributed every 10 minutes, and the process by which those winning miners determine the transactions added to the blockchain.
Bitcoin-NG would create two types of blocks: key blocks, which contain no content but elect a “leader”; and microblocks, which would contain only transaction content.
“Only the leader can generate the private blocks,” Eyal explained. “The interval between the key blocks would be 10 minutes, while ‘microblocks’ come in every 10 seconds.”
Under the system, keyblocks would be given the rewards from the mining block, while 40% of fees would go back to the leader and 60% to those who submit microblocks.
The proposal is still in its early stages and no white paper has yet been released.
Economics and incentives
The economics and incentivizes portion of the day saw three talks by BitTorrent creator Bram Cohen, physicist and entrepreneur Peter R and researcher Miles Carlsten.
All three talks focused on how miners – the parties on the bitcoin network that process transactions – should be compensated, and how the larger network can expect these parties to behave when as existing incentivize structure may alter, whether that’s as expected under the current bitcoin design or under an alternative proposal.
Cohen’s talk, entitled “How Wallets Can Handle Real Transaction Fees”, focused on exploring how the bitcoin network might be shaped should the number of transactions sent to miners routinely passes the 1MB cap for information included in new blocks.
Overall, Cohen was supportive of the idea that the community allows the cap to remain at 1MB on the basis that this would help ascertain whether fees could one day replace block rewards.
“If we hit blocksize limit and get real transaction fees, the idea is you should have a market where some transactions never happen because they don’t pay the fee and you get an equilibrium price. We don’t know what fees should be until we run the experiment and see what prices end up happening.”
Cohen advocated for upgrades to bitcoin wallets that would allow users to more flexibly interact with the bitcoin network by setting minimum and maximum fees and a length of time before the transaction would be canceled should the fee be too low to incentivize its inclusion in a block.
Finding a free market
Another interesting take on issues surrounding the network came during Peter R’s talk on how the economics of the bitcoin network could be expected to work should the community decide to remove a blocksize limit, essentially restoring the network back to its state before the cap was introduced as a way to combat spam.
“Most people think miners find the nonce,” Peter R said. “Miners have another job as well, they produce a new type of digital commodity called block space, or ‘room for transaction data.'”
From there, Peter R delved into the economics of supply and demand, suggesting that the market could be expected to a find an equilibrium even without a cap on block size.
“Economists struggled with this problem, too,” he said. “They postulated a new law called the law of supply. It says producers will only plant more apple trees if they make more money for doing so. Supply and demand intersect at free-market equilibrium. Even though demand can be considered infinite, we still get finite amount of production.”
With this as a backdrop, Peter R ended by calling the blocksize a “political measure” that no longer had a clear benefit for the network other than reducing potential productivity.
He ended the session by announcing the forthcoming launch of Ledger, a peer reviewed bitcoin journal that will seek to highlight the best white papers and research produced by the technology’s enthusiasts.
The imbalances that could be created by poorly aligned incentives in the network were touched on most directly by Miles Carlsten, who presented research conducted jointly with fellow Harry Kalodner and Arvind Narayanan.
Carlsten discussed the non-network-related issues those who process transactions on the network face, including the cost of electricity and hardware. He evoked the idea of a gap that will occur as miners begin to selectively contribute to the network at times when the expected reward outweighs the cost.
“This issue is made worse by the fact that hardware is becoming commoditized, which increases vulnerability to attack,” he said. “With the majority of miners mining at some gap, if attackers start mining right away, what fraction of the hashrate do they need to perform 51% attack? The fraction of hashpower quickly drops from 50%. This is a real threat to the security of bitcoin, as blocks must be immediately profitable to mine.”
Most notable was Carlsten’s contention that these issues could occur as quickly as in 2016 when the reward the network produces every 10 minutes will decline from 25 BTC to 12.5 BTC.
“We think a gap will be profitable at next block halving,” he said.
Anthropology of open source
Rounding out the day’s session was a talk by activist and author Gabriella Coleman, who provided an overview of how open-source groups have historically responded to governance challenges.
Coleman’s work includes studies on universal operating system Debian and hacktivist group Anonymous.
“I was surprised to hear this was one of the first meetings where developers have come together,” she told the crowd. “It’s pretty rare to have an open-source community that doesn’t meet. Those that don’t meet often crumble and fall away.”
Coleman suggested that, based on historical precedent, bitcoin is unlikely to function on a governance model in which a “benevolent dictator” can sufficiently oversee the project.
“Benevolent dictators can work if there’s a founder, but usually that style of governance is matched with other types of governance,” she said, adding:
“Something like bitcoin where there is this mythical character, you don’t have that kind of person.”
Images by Pete Rizzo for CoinDesk