Jeff Garzik and Gavin Andresen: Bitcoin is Being Hot-Wired for Settlement
Editor's Note: This piece was originally published on Medium by Co-Founder of Bloq Inc. & Bitcoin Core Developer Jeff Garzik and Chief Scientist of the Bitcoin Foundation & Bitcoin Core Developer Gavin Andresen.
Disclaimer: The following is personal opinion and does not reflect an official position by Bloq, MIT or any other organization or group.
The proposed roadmap currently being discussed in the bitcoin community has some good points in that it does have a plan to accommodate more transactions, but it fails speak plainly to bitcoin users and acknowledge key downsides. The roadmap summary most relevant to bitcoin users is:
Bitcoin is shifting to a new economic policy, with possibly higher fees.
Core block size does not change; there has been zero compromise on that issue. In the face of rising transaction volume — it has doubled over the past year — getting stuck at 1M results in higher fees, notable economic changes, and suffers from increased political risk by embracing an accidentally-created economic policy tool.
Change By Design
Higher fees and reshaping the fee market impact all bitcoin users, yet it is only mentioned obliquely in paragraph 18 of the roadmap:
These proposals help […] prevent defection between the miners from undermining the fee market behavior that will eventually fund security.”
and in the middle of paragraph 4 of another BIP by the same devs:
The development of a fee market and the evolution towards an ecosystem that is able to cope with block space competition should be considered healthy. […] However, the purpose of such a change should be evolution with technological growth, and not kicking the can down the road because of a fear of change in economics.”
Notable devs think it necessary to change bitcoin to a different economic system with “healthy” competition for block space. In the field today, that is accomplished by maintaining the core block size in the face of rising transaction volume — an outcome the current dev consensus has produced, and the roadmap continues.
In an optimal, transparent, open source environment, a BIP would be produced, covering a change in bitcoin’s economics to a “healthy fee market.” This would be analyzed through the lenses of technical, economic, hard fork etc. risk. This has not happened.
There would also be a related BIP describing the basic requirements for a full node in terms of RAM, CPU processing, storage and network upload bandwidth, based on experiments — not simulations — done on a platform like planet-lab.org. This would help determine quantitatively how many nodes could propagate information rapidly enough to maintain Bitcoin’s decentralized global consensus at a given block size.How Satoshi Avoided a “Visible Hand”
Getting stuck at 1M core block size transforms a historic DoS limit into an accidental policy tool. Satoshi added the 1M consensus limit in 2010, intentionally set above the free market fee range. This artificial ceiling acts against network DoS, raising the cost of attack. Setting the limit above free market range resulted in a safety limit reasonably free from politics.
The update process was also described by Satoshi in 2010.
As average block size approaches the 1M limit, the game theory picture changes. The accidental, artificial 1M limit becomes a Visible Hand in the market. Competition occurs not only for block space, but for developer consensus — because in this new economic system, the ability to freeze or move the 1M limit produces a system where humans — not the free market directly — wield oversize power.
By accident or design, Satoshi managed to create a working free market and push this Visible Hand years into the future by setting the limit high, well above the free market range for transaction fees. The limit served for years as a DoS limit, exponentially increasing cost-of-attack, while a free market equilibrium range established itself.
This block size debate ultimately comes down to competing economic and system survival theories. One theory is that a free market range exists for block size, in absence of a hard limit. Another theory is that a hard limit is required to forcibly constrain the free market. Stalling on core block size changes the former to the latter — uncharted territory for bitcoin.
A System-wide Upgrade To Avoid A System-wide Upgrade
The resultant bitcoin user and market view is muddled: From 2010 through Scaling Bitcoin:Montreal, it appeared that the core block size would see an increase. Following Scaling Bitcoin:Hong Kong, the roadmap abruptly switches direction to Segregated Witness (SW).
SW avoids an ecosystem-wide hard fork through ecosystem-wide upgrades to bitcoin transactions, blocks, addresses, scripts, full nodes, miners, wallets, explorers, libraries, and APIs. All to provide partial relief to core block pressure assuming users upgrade — 1.6M if 100% upgrade, based on current usage.
SW roll-out requires extensive software modifications just to maintain current functionality in the face of rising transaction volume. SW complicates bitcoin economics by splitting a “block” into a basket of two economic resources — core block and extended block — each with unique price incentives and (heavily intersecting) sets of actors.
In contrast, increased core block size is compatible with existing bitcoin software; Some wallets will work seamlessly with no change at all. The total number and scope of changes to wallets, databases, libraries etc. is very minimal. The high hurdle is the hard fork itself.
One of the explicit goals of the Scaling Bitcoin workshops was to funnel the chaotic core block size debate into an orderly decision making process. That did not occur. In hindsight, Scaling Bitcoin stalled a block size decision while transaction fee price and block space pressure continue to increase. Scaling Bitcoin was useful in surveying consensus on core block size. 2M appears to be the consensus most common denominator.
Skipping Hard Questions Until Too Late
The roadmap skips the short term issues of:
- When are fees too high?
- What is the process for changing core block size then?
- Why do we need high fees at this early stage of bitcoin’s life?
Rather than an automated software system, a fixed core block size puts an economic policy tool in the hands of humans. Humans are making subjective decisions about “healthy” fee levels, what miner income should look like, and the relative expense of bitcoin transactions, rather than the free market.
Users have concerns that this roadmap and new economic direction dances obliquely around a shift of bitcoin from a network for P2P cash payments to a settlement system for as-yet-incomplete technology such as side chains or payment channels, pushing out businesses that bought into the original “P2P electronic cash” vision of bitcoin. As the RootStock white paper notes:
If Bitcoin block size is not increased via a hard-fork, when the next Bitcoin reward halves, Bitcoin transaction fees may become prohibitively high for certain applications.”
Maybe that’s inevitable. However, in the short term, we have a disappointing situation where a subset of dev consensus is disconnected from the oft-mentioned desire to increase block size on the part of users, businesses, exchanges and miners. This reshapes bitcoin in ways full of philosophical and economic conflicts of interest. As noted here, inaction changes bitcoin, sets it on a new path.
Bitcoin is not an academic science project. Stalling on hard questions produces tangible market changes. Few have the luxury to pause until a new payment layer is developed on top of bitcoin-1’s emerging settlement layer. Stuck-at-1M risks reversing bitcoin’s network effect by pricing users out of the core blockchain, forcing them onto centralized platforms.
A better way forward includes leadership on a definitive short term core block size decision, plain talk with users about exploring new fee market economic theories and system survival theories, and plain talk with users about the risks and possible negative consequences of getting stuck at 1M.
Core block size resolution and validation cost edge cases are the top priorities. A positive outcome of Scaling Bitcoin was a consensus of 2M, assuming some validation DoS fixes. Segregated Witness can proceed in parallel, sans the assumptions that it’s an easy change or that it mitigates economic issues described above.
And finally, to remove long term moral hazard, core block size limit should be made dynamic, put in the realm of software, outside of human hands.
Bitcoin deserves a roadmap that balances the needs of everybody who has worked hard over the last six years to grow the entire ecosystem.
—Jeff Garzik and Gavin Andresen