A group of prominent developers backing a plan to expand the capacity of the Bitcoin network threw in the towel on Wednesday. "It is clear that we have not built sufficient consensus for a clean blocksize upgrade at this time," wrote developer Mike Belshe in a Wednesday afternoon email.
In the short run, the announcement averts a looming crisis for the Bitcoin community, which faced the possibility that the payment network could be split into two rival, mutually incompatible versions. But the announcement also prolongs the capacity problems that have been plaguing Bitcoin for years.
The Bitcoin network is no longer able to accommodate all the users who want to make Bitcoin transactions. As a result, the price to make a Bitcoin transaction has soared. In Bitcoin's early years, the cost to make a Bitcoin transaction was $0 or close to it. Today, you often have to pay a fee of more than $5 to get your payment accepted by the Bitcoin network.
Technically speaking, the fix for this is easy. A hard-coded limit in Bitcoin software—1 megabyte per blockchain block—prevents the network from processing more than about a dozen transactions per second. Bitcoin developers could simply change this limit to a higher number.
But Bitcoin purists worry that raising the block limit will raise the cost of participating in Bitcoin's peer-to-peer network, effectively shifting power to larger players in the Bitcoin ecosystem.
Over the last two years there have been several failed attempts to raise the block size limit. At this point, it looks like Bitcoin's block size limit could remain permanently stuck at 1 megabyte. With its capacity permanently limited to around a dozen transactions per second, Bitcoin's ability to disrupt conventional payment networks could be greatly diminished.
A compromise that fell apart
We covered this debate back in July when the Bitcoin network accepted a compromise that was supposed to clear the way to expand the capacity of the Bitcoin network. Then, as now, the Bitcoin community was split into two rival camps.
In one camp was Bitcoin pragmatists who worried that the 1 megabyte block size limit and rising transaction fees would hamper the network's future growth. They wanted to increase the block size—to at least 2 megabytes at first, and eventually a lot more—and they'd been floating proposals to do this for the last two years.
Opposing them were Bitcoin traditionalists who argued that raising the block size limit could permanently change the network for the worse. For them, the most important thing about the Bitcoin network was its decentralization—a consequence of the fact that thousands of people around the world have computers that participate in Bitcoin's peer-to-peer process for processing transactions.
To run such a "full node" on the Bitcoin network, you need to download and store a complete history of every Bitcoin transaction ever made—now about 140 GB and growing by about 5 GB per month. Double the block size limit and you double the amount of storage and bandwidth required to run a full node, traditionalists say, and that will price ordinary users out of the market.
Pragmatists counter that 5 GB per months of bandwidth and storage just isn't very much. At current cloud computing prices it would be well within the budget of ordinary Bitcoin hobbyists to continue operating nodes, even with a substantially higher block size.
While superficially a technical argument, the deeper disagreement is about what kind of network Bitcoin is going to become. Big blockers envision Bitcoin becoming a major global payment network, perhaps eventually challenging platforms like Western Union and Visa. The small-block crowd cares less about this. Their top priority is keeping Bitcoin out of the reach of governments and big companies.
While traditionalists opposed a higher block size, they favored a technical fix called segregated witness that moved part of each transaction outside the main blockchain. That would effectively give the network more capacity without raising the barrier to entry. Big blockers weren't opposed to the idea in principle, but they worried that accepting it without a block size increase would cement the one megabyte limit forever. As we'll see, that worry was well-founded.
Things fall apart
In May, a broad group of Bitcoin developers announced a deal that was supposed to bundle these two changes together. Under the terms of the deal, the network would implement segregated witness first. Then, three months later, the network would start allowing two megabyte blocks.
By July, a critical mass of Bitcoin miners had signaled support for the compromise. The plan called for segregated witness to go into effect first. Then the two megabyte blocksize increase would go into effect three months after that.
Segregated witness went into effect as scheduled in August. But because the Bitcoin network is fully decentralized, there was no way to make the second half of this deal—doubling the block size limit—legally binding.
Once segregated witness went into effect, the apparent consensus started to evaporate. Opposition to doubling the block size—a proposal that has come to be known as "2x"—started to harden.
"There’s some opposition to the actual technical changes of the plan," 2x opponent Alex Morcos said in a recent conference call. "There’s a much, much larger opposition to the process that 2x has undertaken."
Many opponents of the block size increase came to saw it as a Silicon Valley coup. Support for larger blocks is strongest among well-financed Bitcoin companies who want the network grow so they can build successful businesses on top of it.
Opponents worried that the 2x change would set a precedent that would give Bitcoin's business community outsized influence over future changes to the Bitcoin platform. They painted a lurid picture in which future changes to Bitcoin were hashed out by major Bitcoin company CEOs in smoke-filled rooms, then presented to the public as a fait accompli.
Big block supporters, by contrast, see opponents as impossible to please. They point out that this May was far from the first time the Bitcoin community considered raising the block size. Indeed, the community has been debating the issue for years, and several previous efforts to raise the block size have floundered based on the same concerns. A prominent developer named Mike Hearn left the Bitcoin world in disgust almost two years ago after a previous proposal to raise the block size limit failed to gain traction.
There's a certain amount of poetic justic in the fact that leading Bitcoin companies trying to upgrade the Bitcoin network were foiled by a populist backlash. Bitcoin is as much a political movement as it is a technology project, and the core idea of the movement is a skepticism about decisions made behind closed doors.
For a lot of people, Bitcoin's original selling point was its potential to undermine the power of the financial establishment in the United States and around the world. It was probably inevitable that once Bitcoin developed its own homegrown business establishment, this same populist fervor would turn on them.
The irony, of course, is that the Bitcoin community is never going to pose a serious threat to the global financial establishment if it continues behaving this way. Successful movements need pragmatic leadership, and that inevitably means allowing some people to broker compromises that move the community forward. There's a real danger that the increasingly deep divisions within the Bitcoin community will permanently hobble the platform's growth.