Bitcoin is torn by opposing impulses: to be a store of value, like digital gold, or the internet equivalent of cash, used for everyday purchases like cups of coffee. It’s the tension at the center of bitcoin’s long running “civil war“—to date, the notion that bitcoin should be more like digital gold has prevailed.
But one solution that could turn a clunky store of value into a means of cheap and quick payments hit a major milestone last week, tilting bitcoin’s future towards one where people use it pay for things instead of hoard it. The fix is called the Lightning Network, and the bitcoin world has awaited its release for years. It takes an approach that neatly sidesteps the intractable fight over lifting the cap on the number of transactions that the bitcoin network can handle.
Bitcoin has struggled with growing transaction volumes. It currently processes about three transactions per second, in contrast to more than 3,600 transactions that the Visa network completes every second. One way to increase bitcoin’s transaction capacity is to increase the size of its “blocks,” which are bundles of transactions that bitcoin miners add to the bitcoin blockchain every 10 minutes. But that approach has been contentious because the block size was set at the start by bitcoin’s mysterious creator, Satoshi Nakamoto, and critics of the approach say changing it would fundamentally alter bitcoin’s character, opening it up to greater centralization.
Building on top of bitcoin
Instead of joining the fight to change the size of bitcoin’s blocks, the Lightning Network is a separate protocol that sits atop the bitcoin blockchain. It’s one type of “Layer Two” solution that observers believe will help remove the bottleneck in bitcoin’s transaction capacity. One analogy used by the Lightning Network’s developers is that it’s akin to the hypertext transfer protocol, or http, which delivers this article to you. While http handles the delivery of web content, it sits atop the TCP/IP protocols, which take care of routing all the internet’s bits to their appropriate destinations.
Last week, developers working on the Lightning Network claimed a victory: they sent two live transactions over the bitcoin blockchain using three independently written pieces of software that validated their tests to date. The result showed that Lightning transactions are interoperable, and can work across different software implementations. Lightning experiments to date were performed in a testing environment, so performing a live transaction was a milestone for the technology’s proponents.
One gripe in the bitcoin world is that if the cryptocurrency remains like a form of digital gold, it will be too slow and expensive to use at a coffee shop. Fittingly, one of the Lightning test transactions was sent to a mock-coffee shop called Starblocks. “What if in order to send an email you had to download every email that anyone had ever sent?” said Elizabeth Stark, who heads Lightning Labs, one of the startups developing the protocol. “That would be super inefficient, right? You’d never send any emails from your phone. That’s how blockchains work.”
Bitcoin’s infrastructure currently depends on users who download the cryptocurrency’s entire transaction history, or blockchain. These users are bitcoin miners, who earn a reward for their work adding new transactions; or full-node operators, who volunteer to check those new transactions. Full-nodes act as a sort of check and balance against miners: if miners produce blocks that contain erroneous transactions, for instance, nodes will reject them, preventing them from being added to the blockchain. As the bitcoin blockchain grows in size—it’s currently about 145 gigabytes—the disk space and bandwidth requirements to operate a full node increase.
The Lightning Network’s big promise is that it greatly reduces the cost of operating a node, meaning that practically any Lightning user becomes a node operator, making transactions a lot cheaper. Stark says operating a node will be as cheap as “sending a tweet,” but doesn’t provide specifics. Node operators will also be able to charge fees for routing payments, but Stark says these will be negligibly low.
Lightning works by exploiting an overlooked facet of the bitcoin program, something called hashed timelocked contracts. This is essentially a turbo-charged escrow function that only releases funds held in a particular bitcoin address after some conditions are met, such as multiple sign-offs from interested parties, and after a certain amount of time has elapsed.
To send a Lightning payment, two parties would deposit the funds at one bitcoin address, in what’s called a “channel” in Lightning parlance (pdf). They can then debit or credit the funds between them as necessary. These debits and credits wouldn’t be added to the bitcoin blockchain—they would happen “off-chain” on the Lightning Network. This gets rid of the need for dozens of small payments between the two that have to be added to the bitcoin blockchain, but still benefits from the security of the bitcoin blockchain.
The Lightning “network,” then, would just be the sum of all these payment channels. The complicated part is devising a way for payments to be routed through multiple channels, much like a packet of data is routed through the hubs and spokes of the internet. If that works, a huge volume of payments could take place “off-chain,” shielding the bitcoin blockchain from a deluge of transactions, while increasing its utility.
On-chain or off-chain?
The counterpoint to the Lightning Network and similar solutions is to simply increase the size of bitcoin’s blocks. That’s what offshoots like “bitcoin cash“, which offers an eightfold increase in the block size, have done. One vocal supporter of that approach is Ryan X. Charles, founder of a startup working on micropayments for content called Yours, and a former engineer at Reddit tasked with working on blockchain tech. He has argued that Layer Two solutions like the Lightning Network will suffer from centralization, and raise the possibility of increased fees. He contrasts that to bitcoin cash: “Bitcoin cash will have low fees indefinitely. On-chain scaling is fundamentally better than off-chain scaling in every way,” he says.
The rejoinder from Lightning proponents is that on-chain scaling is like simply adding more lanes to a highway: It seems like an obvious and straightforward way to increase capacity and speed, until someone builds a plane and bypasses the gridlock entirely. That plane is the Lightning Network, from Stark’s point of view.
Yet even the Lightning solution doesn’t completely get away from the question of bitcoin’s block size. That’s because payment channels still rely on at least one transaction on the bitcoin blockchain to be activated. That means if Lightning takes off as a payments technology, eventually bitcoin’s blocks could still become increasingly crowded—the very situation it faces today.
The arguments over on-chain and off-chain scaling get to the polarized nature of the debate around expanding bitcoin’s transaction capacity: There is a rebuttal for every new solution. This has confused developers hoping to build on top of cryptocurrency protocols. One of them is San Francisco-based Peter Kim, who finally cast his lot with ethereum. He built an app called Cipher that lets people browse decentralized applications like CryptoKitties on their phones. “I have my doubts and I’d like to see them answered. However, different people give completely different answers and all of them seem equally reasonable,” he said.
The Lightning Network was first described in a white paper in 2015, but software implementing its ideas remains at the “alpha” stage. Stark wouldn’t say when more advanced software will be available, noting simply that her team is working as quickly as possible on a beta release. She wouldn’t provide an estimate for when Lightning payments would be generally available. The years-long development of the protocol means that the project faces serious skepticism about it ever alleviating bitcoin’s transaction bottleneck.
For critics of the Lightning Network’s pace of development, Stark has a final comeback: “People vastly underestimate the amount of time it takes to build new cryptographic peer-to-peer protocols,” she said, noting that there are fewer than 10 full-time developers working on the project. “The truth is building these protocols is no small feat and takes time.”
Correction: An earlier version of this article said bitcoin cash has an unlimited blocksize; its blocksize is capped at eight megabytes.