Crypto History

The Block Size Wars: The Civil War That Nearly Tore Bitcoin Apart

The Block Size Wars: The Civil War That Nearly Tore Bitcoin Apart

Bitcoin has survived exchange collapses, state bans and software flaws that could have destroyed its supply cap. But the crisis that came closest to breaking it was internal: a three-year argument, roughly 2015 to 2017, over how many megabytes a block should hold.

Stated that way it sounds trivial. It wasn't. Underneath the number sat the real question — who gets to change bitcoin? — and nobody knew the answer until the fight was over.

The one-megabyte problem

Satoshi Nakamoto added a one-megabyte cap on block size in 2010, as an anti-spam measure, with little fanfare. For years it was irrelevant because blocks weren't close to full. By 2015 they were. Blocks filling up meant transactions competing for space, which meant fees rising and confirmations becoming unpredictable. Bitcoin as everyday digital cash was straining.

One camp — often called "big blockers," including many businesses, payment processors and large miners — saw an obvious fix: raise the limit. Two megabytes, then eight, then more as the network grew. Keep fees low, keep bitcoin usable for payments.

The other camp — many of the Bitcoin Core developers and a large body of users — saw a trap. Bigger blocks mean a bigger blockchain, which means running a full node costs more in bandwidth, storage and time. Price enough people out of running their own node and verification concentrates into a handful of data centers, at which point bitcoin's decentralization — the thing that makes it censorship-resistant and trustless — quietly disappears. Their alternative was to make transactions smaller and move volume to layers built on top, like the Lightning Network.

How ugly it got

This was not a polite technical disagreement. It consumed the community for years and turned genuinely nasty: mass bans and censorship accusations across the major bitcoin forums, which drove a permanent split into rival communities; coordinated harassment campaigns; sustained denial-of-service attacks against nodes running software from the opposing side; developers doxxed and threatened.

Several attempts were made to force a resolution. Bitcoin XT, Bitcoin Classic and Bitcoin Unlimited each shipped software with larger blocks, hoping miners would adopt them; each gained attention and then stalled. In 2017 came the New York Agreement, in which a group of major companies and mining pools representing a large majority of hashrate signed a deal to activate SegWit and then hard-fork to two-megabyte blocks — the "SegWit2x" plan.

On paper that should have settled it. A supermajority of miners and most of the industry's biggest names had agreed. What happened instead is the most important thing in this story.

The users won

Opponents organized around a mechanism called a user-activated soft fork (UASF, specifically BIP 148): a date after which participating nodes would simply refuse to accept blocks that didn't signal support for SegWit. It was a bluff with teeth. Miners' blocks are worth nothing if the economic network — exchanges, merchants, holders running nodes — won't accept them. The nodes were telling the miners: you may produce the blocks, but we decide which ones count.

Faced with a possible split in which their blocks could be orphaned by the users who give bitcoin its value, miners moved. SegWit activated in August 2017. The second half of the New York Agreement, the two-megabyte hard fork, was called off that November for lack of support.

The big-block camp went its own way, forking the chain in August 2017 to create Bitcoin Cash, which later fragmented further. These forks persist as separate coins; none has come close to displacing bitcoin.

What the war actually settled

The lasting outcome wasn't the block size. It was the discovery of where power in bitcoin actually sits.

Before 2017 it was genuinely unclear. Miners produce the blocks — perhaps they decide. Developers write the software — perhaps they decide. Big companies hold the users — perhaps they decide. The answer turned out to be: none of them, individually. Rules are enforced by the people running nodes, and a change that the economic majority won't accept doesn't happen, no matter who signs what. Miners order transactions; they don't get to rewrite the rules. Developers propose; they can't impose. An industry agreement among executives isn't binding on anyone.

That answer is why bitcoin changes so slowly now, and why the community treats that slowness as a feature rather than a failing. It's also the backdrop for every upgrade debate since, including the eventual question of migrating to quantum-resistant cryptography. The war was exhausting, divisive and in places genuinely vicious. It also proved that bitcoin cannot be captured by its own most powerful participants — which, for a system whose entire pitch is that no one is in charge, was worth finding out.

← All articles