By Jameson Lopp
Source: https://blog.lopp.net/on-ossification/
Translator's note: The original title is "On Ossification". "Ossification" literally means "ossification/bone formation", which is also the origin of the original title picture. In the translation, I use "fixed shape" to translate "ossification".
In the context of network protocols, "ossification" refers to the slowing down of the rate at which a protocol evolves and changes. This seems to be a law of network science. Essentially, as a network protocol gains more and more adoption and the "crowd" of the network grows, the effort required to change direction by coordinating software upgrades among the network's users grows dramatically. Eventually, the ability to safely activate any protocol changes crumbles at the sheer size of the network, as it becomes impossible to coordinate a large number of actors without a central point of presence.
Therefore, finalization is the fate of Bitcoin . But from today's point of view, we find that we are still debating whether we should actively finalize the protocol . I personally strongly oppose this direction and believe that from this point in time, the basic protocol still has much more room for improvement.
I ask that Bitcoin, only by being willing to implement reasonable, widely adopted protocol improvements in a careful, consensus-driven environment, can it remain vibrant, interesting, and secure in the long term. Actively finalizing, freezing changes, at this time, is arrogant, short-term, and antithetical to the vision that originally created Bitcoin. Thoughtful and sustained speech is key to Bitcoin's long-term value. The reason why electronic gold is better than physical gold is that it is not inert . What I mean is: the properties of physical gold cannot be optimized, so financial innovation on gold occurs through centralized IOUs... Bitcoin's properties can be optimized to enhance its trustless usage.
Learning from history
Bitcoin is only 15 years old and has already gone through many consensus changes and upgrades. It would be premature to assume that now is the ideal end time. The protocol needs to continue to improve in order to remain attractive.
We should learn from other popular Internet protocols, such as SMTP. If the Bitcoin protocol is finalized, developers will build increasingly complex layers on top to add the desired functionality. Complexity will lead to bugs and explosions.
Recommended article: The Death of Decentralized Email
This is not a criticism of complex technical layers like BitcoinOS/BitVM/Botani’s Spiderchain/Citrea’s zero knowledge rollup — they are just trying to make the best use of the tools available to them. Even if we don’t want Bitcoin to become a kitchen sink protocol, if adding low-level functionality to the base layer can significantly reduce the complexity of features in development at higher levels, then that makes sense.
Possible Path
Many useful features, such as "covenants", "vaults", and payment pools, require upgrades to the base layer. It is much easier to develop these features in a clean way in the protocol itself than in a cumbersome overlay. A base layer with more modules will unlock a new design space for Bitcoin.
“Bitcoin L2 developers want OP_CAT. CAT is good for many projects and can accelerate trustless innovation.
Only someone who has never tried to develop other layers would say, 'We don't need to change the base layer; everything can be done in other layers.'"
——Author ’s Twitter
A careful, well-tested upgrade that has been well-discussed, has gained community consensus, and will not undermine property rights or Bitcoin’s core position as a stable currency. The upgrade reflects the will of the users, not the will of the nurse.
I believe that Bitcoin has the potential to be much greater than what we have already realized. I see the Bitcoin blockchain as a cryptographic accumulator that allows a variety of other systems to be anchored on top. But we are just scratching the surface right now. Today, developing trustless second layers is so difficult that if the protocol is finalized, we will constrain developers and greatly limit experimentation in discovering the most valuable uses of the block space.
People often say that we don’t need to change Bitcoin because we can just scale in other layers. Of course, that would be nice… and developers wouldn’t have to work so hard! But our base layer doesn’t yet have all the primitives that would allow developers to easily launch trustless second layers! We could (for example) make Bitcoin script powerful again.
Recommended Video: The Great Script Restoration Project
Note that to create the Lightning Network, we implemented 3 different forks to enable 3 different modules . Without the features brought by these forks, the Lightning Network protocol would be much more cumbersome and its game mechanism would not be as reliable.
There are other soft forks we can do, like SIGHASH_ANYPREVOUT, which would greatly enhance the Lightning Network, enable “channel factories”, and exponentially increase the efficiency of the Lightning Network. There are other forks that would enhance privacy, like “signature aggregation across inputs”. And there are still forks, like OP_CTV, that would improve the security of self-custody with things like restrictions and safes. We can see a lot of possibilities, but it’s getting harder and harder to get there.
On Sovereignty and Autonomous Custody
Scalability is another long-term issue. Greg Maxwell put it this way in 2015:
“If the system becomes too expensive, people will be forced to trust a third party rather than independently enforce the rules of the system. If the resource usage of Bitcoin blocks is too large (relative to the technology available), Bitcoin will lose its competitive advantage over traditional systems because verification becomes too expensive (driving away many users), forcing people to revert to these traditional systems. If capacity is too low and the way we structure transactions is too inefficient, using the blockchain to resolve disputes will become too expensive, forcing people to revert to the same systems.”
Decentralization of validation, which won the block size dispute, is only part of the story. Decentralization of economic participants is equally important to Bitcoin’s long-term success. Remember, it is not the will of the nodes that determines Bitcoin’s future, but the will of the economic majority among the nodes . Bitcoin’s economic participants include miners, holders, and transactors (including businesses and custodians). What I am saying is this: if the number of nodes with an economic stake is reduced due to centralization, and most Bitcoin users give up self-custody because it is too expensive, then we can expect the “governance” of the protocol to become more centralized and fragmented.
In the case of Bitcoin today, maybe 100 million people around the world can understand the basic properties of Bitcoin. But in a world with a world population of 8 billion, this will not result in a new decentralized currency that changes the landscape of value and scarcity, but only a new elite that will follow in the footsteps of historical elites over time, starting with creating prosperity and ending with bread and circuses. Bitcoin gives us the first chance to transcend the historical cycle and not replace the old elite with a new elite, but this will only be possible if we continue to drive decentralization by optimizing the protocol and making Bitcoin's basic properties more accessible.
Also, note that the argument that “we can do whatever we want on other layers” ignores the fact that no layer can provide the same security model as the base layer. Whenever you develop a layer on top of Bitcoin, you inevitably create a new set of gaming mechanisms and tradeoffs for the self-custodial users of that layer.
Ossification Steelman
Finalizationists believe that Bitcoin has already fulfilled its core functions as a sound currency and store of value. Future changes, even if well-intentioned, will bring unnecessary risks and may weaken many of the properties that make Bitcoin valuable. By finalizing the protocol, we ensure that Bitcoin will remain a trustworthy, decentralized, and immutable monetary system for the long term.
1. Protecting trust in Bitcoin’s fundamental properties
Bitcoin's primary value proposition is its fixed supply and immutability. Any change to the protocol, even a soft fork, has the potential to erode confidence in these core properties. Some claim that the ability to change the protocol reduces confidence in inflation schemes, no matter how well motivated.
2. Reduce developer control and centralization risks
Allowing constant changes to Bitcoin gives developers too much influence. This creates centralization risks, as a small group of people can quietly change the properties of Bitcoin.
3. Avoid unintended consequences
Even well-intentioned, well-tested changes can have unforeseen effects on the network. As Bitcoin grows in value and importance, the potential impact of these unintended consequences becomes increasingly severe.
4. As the network grows, changes become more difficult
As Bitcoin adoption grows and more economic value is built on it, any changes become riskier and more controversial. That’s when the protocol should stabilize, say the old guard, just as other foundational protocols, like TCP/IP and power socket standards, have.
5. Maintaining Bitcoin’s Function as Sound Money
Bitcoin’s primary purpose is to serve as a new monetary system that is immune to debasement. Finalization ensures that this core functionality is protected and that this fundamental property is not sacrificed for potential benefits that may not be critical to its primary purpose.
Common arguments of stereotypers
Stereotypers argue that Bitcoin works extremely well on its own, and that the rules of the protocol should be extremely difficult to change in order to protect its core values as an immutable, non-inflationary, politically neutral (apolitical) currency. Even well-intentioned changes can carry significant risks that could jeopardize Bitcoin’s long-term success and stability.
1. Don’t fix what isn’t broken
Bitcoin has survived with its current protocol and grown from zero to $1.4 trillion in market cap in just 15 years. There is no need to risk sacrificing that success by initiating hasty or unnecessary changes.
2. Stability First
The core value of Bitcoin is its stability and predictability. Protocol rules should not be constantly changed on a whim. Frequent patching will undermine people's confidence in the immutable nature of Bitcoin.
3. Think before you act
Changing Bitcoin is a serious matter, akin to changing a constitution, or a law that has stood for centuries. Any change should be made slowly, carefully, conservatively, and only after considering all the long-term consequences.
4. The bigger the stakes, the less anxious you should be
The larger and more successful Bitcoin becomes, the more important conservatism becomes about protocol changes. When trillions of dollars of value are at risk, we cannot afford to be wrong and take unnecessary risks.
5. The road to hell is paved with good intentions
Even well-intentioned protocol changes can have unintended negative effects, such as harming the economic model of miners, developers, and coin holders. The downside risks may outweigh any theoretical benefits.
6. Keep it simple
Not every feature or optimization needs to happen at the base layer. Many things can happen at higher layers (such as layer 2 or layer 3) without compromising Bitcoin’s core security model and stability.
7. Stable currency in a changing world
Continuous "optimizations" and protocol changes violate Bitcoin's promise to be a stable, politically neutral currency and settlement network. Finality is a property that prevents Bitcoin from being captured by special interests and subject to unpredictable changes.
8. Bitcoin should only be a currency
Non-monetary uses of block space, such as tokens/NFTs, will only encourage fraud and drive away users who use Bitcoin as a currency.
My response to stereotypes
Protecting trust in Bitcoin’s fundamental properties
Trust in Bitcoin's properties and its resistance to bad ideas comes from the governance process for negotiating protocol changes, not from the sheer impossibility of change. Bitcoin is crypto-anarchism, and a system where apathy is the norm (apathy itself is a powerful veto). Either you agree that Bitcoin's governance has worked well so far, or you believe that it's all just luck and we could fail at any time.
Recommended video: Decentralized 2018
Reduce developer control and centralization risk
Similar to the previous point, either you believe that the compilation mechanism around protocol changes is reliable, or you believe that it is all just luck. Bitcoin developers cannot force anyone to run code they disagree with.
Recommended article: Who Controls Bitcoin Core?
Likewise, the internal process of development is fraught with challenges, with the vast majority of proposed rules and code changes failing the test.
Recommended article: Bitcoin Core Contributor Challenges
An excerpt from my research four years ago:
After analyzing all rejected pull requests from Bitcoin Core, we found that:
- A total of 901 1209 lines of code were rejected for addition
- A total of 627,9435 lines of code were rejected for deletion
A total of 1,529,0644 lines of code were rejected, while only 365,1046 lines of code were accepted!
This means that, as of the time of writing, only 19% of proposed code changes have been accepted into the Bitcoin Core codebase.
Preventing unintended consequences
The fear of “unknown unknowns” and “unintended consequences” is, in my opinion, not an argument. Why? Because every decision has unknowable unknowns. Changing the protocol has its own unknowable unknowns. Not changing the protocol has its own unknowable unknowns. There are always countless unknowable unknowns, so the argument cancels itself out. The key is vigilance — this determines Bitcoin’s antifragility.
As growth increases, change becomes increasingly difficult
I don't think there's any disagreement on this point. As the network grows and becomes more valuable, it has become harder to change, and it will continue to do so.
Maintaining Bitcoin’s function as sound money
The upgrade Bitcoin is in will not be finalized. The world will continue to throw new problems at Bitcoin; if Bitcoin cannot adapt to the world and solve these problems, weird and most likely centralized solutions will emerge. This is what led to the decline of SMTP. My point is that finalization cannot guarantee the properties of Bitcoin.
Don't fix what isn't broken
No one is pushing for hasty changes, and no one really thinks Bitcoin will "break" without a certain feature. But what we can say is that it's natural for the way Bitcoin is used to change. It has changed dramatically over the past 15 years.
Stability First
Since Satoshi left the project, there have been no "spur of the moment" rule changes in Bitcoin. Satoshi used to make arbitrary changes without consulting others. I've noticed that some influential people have disparaged Bitcoin developers as "tinkerers", which couldn't be further from the truth. We can observe that Bitcoin's change toxicity has slowed down significantly over the years because developers are overwhelmingly cautious.
Think before you act
In fact, slow, careful, and carefully reviewed optimization is exactly what innovators demand.
The bigger the stakes, the less anxious you should be
Without a doubt, we should strive to avoid making mistakes. We should also have confidence that mistakes are not irreversible. For example, a large number of vulnerabilities have made their way into Bitcoin Core, but they were fixed without causing damage. Moreover, even when the Bitcoin network suffered a consensus failure, it recovered within a few hours (because the network participants remained vigilant). Antifragility is a fundamental property of Bitcoin.
Bitcoin does not exist in a vacuum. In a sense, it is a living thing. And life cannot stagnate and grow at the same time.
Recommended article: Bitcoin in the Currency Mushroom Forest
The road to hell is paved with good intentions
This is another argument that cancels itself out. The Bitcoin ecosystem is extremely complex, consisting of countless moving parts and participants. Beyond the protocol itself, the rest of the system will continue to change, and we should be prepared to deal with the unforeseen consequences of those changes.
Keep it simple
This is another point that both sides of the debate can largely agree on. We should agree that the most important changes to the base layer are those that have spillover effects, allowing innovation in other permissionless layers to experiment as much as they want without worrying about changes to the base layer.
A constant currency in a changing world
This seems to be another problem caused by not trusting the game mechanism behind Bitcoin's governance. My point is that if you don't trust the checks and balances built into the system, then you don't really trust Bitcoin.
1) Public trustless consensus systems allow you to use them without trusting any one person. However, you must trust everyone as a whole.
— Jameson Lopp, June 26, 2016
I will point out later that stereotypes themselves can also allow Bitcoin to be captured by special interest groups. Numbness cannot bring the answer, vigilance and the ability to adapt to new stressful environments are the answers!
Bitcoin should only be a currency
For as long as I can remember, there has been a debate about which uses of Bitcoin should be considered scams. I don’t find these debates interesting because the debaters always end up appealing to subjective value, while we can observe that non-financial uses of block space have objective value to some people — and they are willing to pay for it.
Technically, Bitcoin is a piece of data; a blockchain is basically an append-only log with some other interesting properties. As a result, people have been using Bitcoin for non-financial purposes for more than 10 years. As I wrote 8 years ago, it is this data's permanent existence and unchangeable reliability that attracts people to use it for non-financial purposes.
Recommended article: Bitcoin: Anchor of Trust in the Sea of Blockchain
We don't know what the long-term market for block space will look like - which uses will provide the highest utility and value. This question will become more important as the mining reward halves. I think more functionality, more layers to run experiments, means more chances of figuring out the most valuable uses. Maybe, for example, it could be to enable zero knowledge rollup. We don't know, so I think we should let innovators continue to explore the potential design space.
So, where, in my opinion, is the source of the controversy?
Stereotypers vs. innovators pic.twitter.com/tAbvORcy7V
— Jameson Lopp, October 29, 2024
Frequently asked questions
Q: Are all protocol changes that are not bug fixes “nice but unnecessary”? Are they just interesting things for ivory tower tech geeks, but not strictly necessary for Bitcoin to become a global currency one day?
A: I don't think that "necessary" is opposed to "good". I think it's a matter of finding a path to explore the design space of Bitcoin and maximize the value of the system. To be clear, we can finalize Bitcoin now, and it may continue to work well for a while until we encounter a fatal problem, such as quantum computing or timestamp overflow point. But the nature of Bitcoin itself will develop in another way due to development constraints.
Q: When do you think the change is “end point”? When will Bitcoin be good enough?
A: We should continue to optimize Bitcoin until we can’t optimize anymore. We have seen a sharp decline in the number of new Bitcoin optimization proposals since 2017, now less than 1 per month. Moreover, many proposals are not even offered activation until then because developers don’t want to face the ordeal. As a result, we are losing talent for protocol development, which further exacerbates decay.
Q: Can protocol changes be limited to "bug fixes and maintenance" instead of "adding new features"?
A: This is certainly an option, and it may be the path we will take in the end. But I don’t think we have reached that point yet.
Q: Please describe for us the circumstances under which the Bitcoin ecosystem could accept a soft fork or hard fork change.
Any critical issues that may jeopardize the continued operation of the system may occur. However, the stereotyped nature may make it very difficult to perform emergency repairs in the future.
- Imagine that Bitcoin's consensus rules haven't changed in 10 years. So perhaps the Bitcoin protocol developers working in that time have never actually experienced a consensus change. This situation is not good for us.
- Imagine there is some imminent problem, but we don't know exactly when it will become fatal. The nature of stereotypes means we keep putting off solving the problem... until it's too late. I recommend you check out my recent talk on quantum computing, which provides an exact example of this situation.
Recommended Video: Safeguarding Statoshi's Stash
Conclusion
We should be working on making changes that strengthen Bitcoin and allow more trustless systems to be built on top. I think we can all agree that Bitcoin should not become a kitchen sink protocol (like EVM-based networks), but as it stands, developers who want to innovate find it very difficult without creating complex enumeration logic.
Stereotype is complacency. Sure, we all agree that Bitcoin is awesome. But I disagree that Bitcoin has reached its full potential. I believe complacency is one of the biggest threats to Bitcoin — we can’t rest on our laurels.
Technology is deflationary by nature. Bitcoin’s consensus rules should prioritize security and keep the system decentralized in as many ways as possible. Not just node operators, but also how blockspace is used. Because, after all, if a person is unwilling to use blockspace because it is too expensive, then they certainly won’t run a full node. I also think that we should strengthen developers’ ability to develop trustless second layers so that we can continue to explore the potential use of blockspace as a cryptographic accumulator, so that there is more opportunity to find sustainable demand for blockspace and pay for Bitcoin’s thermodynamic security in perpetuity.
Do we want Bitcoin’s future to be driven by optimism or pessimism?
Stereotypes are driven by pessimism and fear of failure.
In my opinion, a person's position can be summed up as either he believes Bitcoin is anti-fragile or he believes Bitcoin is fragile.
— Jameson Lopp, October 30, 2024
To be clear, caution is extremely important. You will never see me claim that “Bitcoin will fail if it doesn’t implement feature X.” In my opinion, one of the factors that has made Bitcoin successful is actually making it very difficult to change.
The real headache comes when you realize that Bitcoin’s exchange rate will continue to soar even as its unique properties are slowly eroded.
Wealth should not bring complacency.
— Jameson Lopp, October 29, 2024
But if we are paralyzed by fear, we will likely be on the wrong path and significantly hinder exploration of Bitcoin’s design space, which I believe will naturally limit its adoption, use cases, and enhancement of its useful properties.
The pain to be prevented often outweighs the actual danger to be avoided.
—Napoleon
I issue this warning because I see a conflict coming. Think about why we would want to improve the Bitcoin protocol. The only reason to care is if you are a real user of the protocol. In other words, protocol improvements only make sense for users who have self-custody . If you use a third party to hold your funds, you don't care how they use the protocol.
Those concerned about Bitcoin TradFi adoption don’t care about optimizing the protocol to scale the network because they don’t care about self-custody.
The next big battle over the future of Bitcoin is brewing.
— Jameson Lopp, September 21, 2024
Maybe Bitcoin has already been finalized, and the above words are futile. It is impossible to know for sure until it is too late. The world never stops evolving, and we must ask ourselves whether we want Bitcoin to evolve with it, or let it fall behind.
Let us forge ahead courageously!
