Bitcoin’s user experience for transactions declined significantly in 2017 as worldwide attention caused demand to explode above the 1mb block limit.
Shifting payments on to a second layer via the Lightning Network is the main focus of the community to bring relief from high fees and unreliable transactions as bitcoin scales to meet increasing demand.
Below I explore whether Lightning is likely to:
1) Solve Bitcoin’s high fees and unreliable user experience in 2018.
2) Solve Bitcoin’s high fees and unreliable user experience within the next 2 or 3 years.
3) Successfully scale Bitcoin in the long term while avoiding centralization.
1) Is Lighting a Solution to High Fees and Unreliable User Experience in 2018?
A few weeks ago Blockstream announced they would accept Bitcoin Lightning payments for their online store.
This turned out to be controversial as people debated whether the technology was developed enough to invite the public to risk real money. Elizabeth Stark of Lightning Labs called this a bad move as testnet is still grappling with bugs which are causing devs to lose funds.
Personally, I’m glad to see the pace pick up with the shift to mainnet as others follow Blockstream’s example of offering real products and services for sale. Errors are inevitable but faster progress will be made as assumptions are tested in the real world, more minds focus on overcoming the challenges, and new knowledge and technology is created to solve problems as they emerge.
In light of these somewhat surprising developments, a wave of excitement has swept the bitcoin community.
Twitter is in the midst of a Lightning mania (reminiscent of the Segwit mania taking hold around this time last year). Enthusiasts are already campaigning large merchants to integrate Lightning and offer payments. A sizeable portion of the bitcoin community is under the impression that scaling is now solved, confidently stating that 2018 will bring mainstream adoption with pretty much free and instant transactions as more nodes join the network.
These high expectations are misguided. The scaling debate has been bitter and emotional and many have invested their hopes in Lightning for so long that they’re hailing it as a panacea. But the current hype far exceeds the reality of what kind of impact Lightning can hope to have on user experience in the short term.
Lightning is still in its very early stages and simply isn’t ready to take on a meaningful percentage of bitcoin’s day to day transactions this year.
It’s Too Complex and Burdensome to Gain Mass Implementation in 2018
Integrating Lightning and using it for safe, reliable payments is a significant technical challenge at this early stage. Lightning is clearly going to take a while to get to a level that’s secure and user friendly enough for large numbers of businesses to devote the time and resources necessary to implement it.
Merchants have been slow to accept even regular bitcoin payments and reluctant to allocate the required engineering time to implement segwit. Given how much more complicated and risky Lighting is, it shouldn’t be a surprise if the reception it gets from businesses is underwhelming.
There are many easier payment options that businesses can work with. It’s far more convenient for them to integrate other cryptocurrencies than Lightning payments. If on-chain fees return to a high level and Lightning fails to match the cost, speed, and convenience that other cryptocurrencies offer, bitcoin may continue to lose market share until these usability issues are brought back to a level that competes with the alternatives.
Early enthusiasts are setting up Lightning nodes and experimenting with commerce, but this won’t be enough to have any real impact on transaction capacity or fees which means it’s not going to bring any economic impact or alleviation of bitcoin’s underlying problem this year.
It Requires Significant Resources, Risk, and Liquidity to Grow to the Required Scale
To build out stable, reliable routes and scale the network to a large user base, channels need to be funded with a lot of liquidity.
This involves reallocating large bitcoin holdings from more secure storage and tying it up in an unproven network with dubious levels of security.
Users are likely to be hesitant to commit the necessary capital investments the Lightning Network requires for it to have any meaningful impact on bitcoin’s overall usability for transactions.
The high levels of risk and opportunity cost the Lightning Network imposes means growing out the network is a long term endeavor.
Lightning will not solve bitcoin’s high fees and unreliable user experience in 2018.
2) Is Lighting a Solution to High Fees and Unreliable User Experience Within the Next 2 or 3 Years?
The problems that will stifle meaningful adoption of Lightning in 2018 are likely to remain over the next few years. These are not trivial hurdles to overcome so 2018’s struggle for traction may drag into 2019 and 2020.
Bitcoin needs to be able to accommodate a massive amount of new users if it’s to stay on track and fulfill its potential as a global currency. Demand for transactions will surge the next time cryptocurrencies enter a period of hype, causing the 1mb base layer to be congested again.
This congestion will cause fees to skyrocket as they did in December 2017 when Bitcoin was on the front page of newspapers all over the world.
It’s hoped that users will have shifted their transactions to Lightning by the time this happens, but high demand with full blocks on the base layer would negatively affect Lightning’s usability even if it does become widely adopted in time.
A congested layer 1 causes problems for layer 2 in the following ways:
High On-Chain Fees Make it Costly to Open and Close a Channel
Regularly opening and closing channels will be necessary for many users who won’t be able to commit funds to channels indefinitely. When the money is needed outside the network they’ll have to suffer a high fee to close the channel and another one to open a new channel at a later date.
People don’t like locking up money for purchases in advance. Many around the world live paycheck to paycheck and don’t have spare cash lying around to tie up in channels with high entry and exit fees.
Reliably finding a route through the Lightning network is a tough, unsolved technical challenge that gets harder for nodes as it scales. Failures to route payments means many users will have to close their channel, wait for their funds to become available on-chain again, and use the funds to to open a new channel that can find a route. This results in more on-chain fees.
Other cryptocurrencies with low on-chain fees will look like attractive alternatives to regularly paying high on-chain fees to open and close channels.
Delays in Opening and Closing Channels
The low throughput limit of a 1mb blocksize means onboarding and getting out of Lighting will be slow and unreliable if on-chain is backlogged with demand.
Problems Interfacing With the Base Layer Would Frustrate Users, Hurt Adoption Rates, and Concede More Ground to Alternative Cryptocurrencies
If the problems of layer 1 cause layer 2 to function worse than expected, it will be difficult to get people interested in going through the hurdles beyond the hardcore Lightning enthusiasts.
Lightning faces an uphill battle to solve bitcoin’s high fees and unreliable user experience within the next 2 or 3 years.
3 – Will Lightning Successfully Scale Bitcoin in the Long Term While Avoiding Centralization?
As described above, if bitcoin is to scale to hundreds of million/billions of users, massive demand for opening and closing channels will still mean congestion on a small block limit base layer with resulting high on-chain fees.
These huge on-chain fees would price many people out of ever being able to perform an on-chain bitcoin transaction and make them fully reliant on Lightning payments.
A bad user experience setting up and closing channels tempts people to take shortcuts that well-funded and well-connected third party services can offer to alleviate pain. Users might only want to open a channel with the largest hubs if they consider this the cheapest and most reliable way to route through the network.
Merchants might deem it practical to only open one channel with the most connected hub and decline to open channels individually with customers. This would promote centralization further if customers had to go through that centralized hub to pay the merchant.
High on-chain fees make these scenarios more likely. If people aren’t priced out of making layer 1 transactions then it’s easy to avoid any malicious aspects of centralization within the network. Users can always escape back to layer 1 and easily route around central hubs.
Do Hubs Have Real Power Or Is It Trustless?
Centralized hubs can’t confiscate your bitcoins but they would be able to cause problems for you by refusing to route your payment.
You can close the channel without losing your funds but this will be an extra cost and could be a long delay if the base layer is congested. It would cause you to lose access to your money for a period of time which would be a big deal for a lot of users.
If it’s unaffordable to get back onto the base layer, users may tolerate worse behavior from nodes. It needs to be very easy to exit Lightning to make this a non-issue.
Lots of Unknowns and Potential Unintended Consequences.
The system could behave in undesirable ways if a few extremely valuable and connected nodes emerge in a system with some usability issues.
How users and businesses will interact with Lightning at scale in the real world is a big unknown at this stage.
The Route Finding Problem Promotes Centralization
As the network grows and becomes more complex, more data must be tracked and kept. This makes it harder for your node to reliably find the best route through the network.
Matthew Zietzke estimates the amount of network topology data one must keep up with in order to effectively route transactions:
- Nodes and connections, as well as connection directionality
- Useful capacity of channels in each allowed direction.
- Node routing fees
- Average distance (number of hops) between end user nodes (payment hubs can be largely discarded)
Network changes to update prior to routing a tx:
- Network topology changes, meaning activity of nodes in the route, and the opening and closing of payment channels
- Changes in fees
- Changes in useful channel capacity (an upper bound on your tx size through that node)
This is a massive amount of data to keep up with and it’s constantly changing.
As the Lightning Network grows to its intended scale it gets more difficult to maintain a node’s route map. If your node can’t keep up with the data you get “Route not found” errors.
Other fields have solved this problem by adding centralization. Unfortunately it’s still an unsolved problem for the distributed calculation of routes that Lightning is designed for.
Implications of the Route Finding Problem
Favors Routing Through Large Nodes
Unreliable route finding leads to a bad user experience and paying higher average fees through inefficient routes.
This incentivizes users to go through large well connected hubs that can offer more convenience with their superior routing capabilities.
People might prefer to only open one channel that passes payments through the easiest and most connected route.
High Rates of Channel Opening and Closing to Ensure a Route is Found
If your node fails to route you need to open a new channel to create a path to the recipient.
On a high fee layer 1 the rate of channel creation people can tolerate is limited. This pushes people towards using larger hubs that can ensure reliable routing.
Adding more channels to the network also increases the routing complexity and costs to run a functioning node.
A Breakthrough is Needed to Solve the Route Finding Problem
I don’t doubt that this technical challenge can be solved at some point, but it requires the creation of new knowledge which makes any estimate of when the problem will be solved unreliable.
Will Lightning Successfully Scale Bitcoin in the long term while avoiding centralization?
1 – Lightning will not solve bitcoin’s high fees and unreliable user experience in 2018.
2 – Lightning faces an uphill battle to solve bitcoin’s high fees and unreliable user experience within the next 2 or 3 years.
3 – Unknown if Lightning successfully scales Bitcoin in the long term while avoiding centralization.
- Payments channels are a good asset for bitcoin. A network of payment channels is a clever idea that will be useful at some point.
- Despite being a very interesting concept with the early roots of the network starting to form, Lightning isn’t close to being ready to play a major part in solving real world problems users have when transacting with bitcoin.
- Lots of people don’t realize this and have unrealistically high expectations. Lightning is not a panacea and the market is likely to be disappointed if mass demand for transactions returns and Lightning is unable to take a major part of the load away from the base layer.
- Competing cryptocurrencies are cheaper and often faster. Lightning is unlikely to change this for a while so use-cases may continue to be conceded to other coins.
- Lightning is complex and would introduce many subtle changes to the incentive structure and system as a whole if adopted on a mass scale. This has unknown implications on how people and institutions will interact with it and what kind of unintended consequences there will be.
Follow me on Twitter @leebanfield1