The Block Size War | LearnBitcoin<br>Rabbit Hole · 13 min<br>The Block Size War<br>The 2015-2017 fight over Bitcoin's block size limit, from Bitcoin XT to the UASF to the forks that split away, and how it settled who actually controls the rules.

Where you're going: For about two years Bitcoin tore itself apart over how big a block should be. The argument was nominally about a megabyte of block space, but the real fight was over who gets to change Bitcoin's rules: the miners, the big companies, or the ordinary people running nodes. This chapter walks through how the war started, the failed deals and the alternative software, the standoff over SegWit, the moment users forced the issue, and the two forks that broke off. The ending is the payoff, because it decided how Bitcoin has handled every upgrade since.

A fight about one megabyte

Bitcoin blocks have a size limit. Satoshi added it quietly in 2010, capping each block at one megabyte. Blocks were nearly empty back then and cost almost nothing to fill, so the cap was a cheap defense against someone stuffing the chain with junk. At the time nobody treated it as a sacred number. It was a temporary guard rail.

By 2015 the guard rail was starting to bind. Blocks were filling up during busy periods, transactions were waiting longer to confirm, and fees were creeping up. The question was obvious: do we raise the limit?

Answering it turned out to be the hardest thing Bitcoin has ever done. The argument ran from roughly August 2015 to November 2017. It split companies, developers, miners, and online communities into hostile camps, and it ended with two permanent forks off the network. Jonathan Bier later wrote a whole book about it, The Blocksize War, and the title gets the tone right.

The war in one picture: moves by miners and companies above the line, moves by users and developers below it. Bitcoin Cash left in August 2017; the SegWit2x fork was called off before it happened.<br>The two sides

The dispute sorted people into two camps that disagreed about what Bitcoin is for.

The big-blockers wanted to raise the limit, soon and substantially. Their reasoning was straightforward. Bitcoin should work as everyday electronic cash, fees should stay low, and if blocks are filling up then you make them bigger. Disk space and bandwidth are cheap and getting cheaper, so why accept artificial congestion? This camp included early heavyweights like Gavin Andresen, who Satoshi had handed the project to, the developer Mike Hearn, the entrepreneur Roger Ver, and the large Chinese mining operations led by Bitmain's Jihan Wu.

The small-blockers wanted to keep blocks small and grow capacity other ways. For them the issue was decentralization. The whole point of running a full node is that you verify the rules yourself instead of trusting someone else. The bigger blocks get, the more expensive it becomes to run a node, and the fewer people bother. Push that far enough and validation ends up concentrated in a handful of data centers, which is the kind of choke point Bitcoin exists to avoid. Better to keep the base layer lean, improve how efficiently it uses the space it has, and move everyday payment volume to systems built on top, like the Lightning Network. This camp included most of the Bitcoin Core development group.

It is worth being fair here, because the caricature from the time was that one side wanted Bitcoin to grow and the other wanted to strangle it. That is not what was going on. Both sides wanted Bitcoin to scale. They disagreed about how, and behind that, they disagreed about who should get to make the call. The second disagreement is the one the war ended up settling.

The alternative clients

The big-block camp's first move was to route around Bitcoin Core by shipping rival software with a bigger limit baked in.

Bitcoin XT came first, in August 2015, built by Mike Hearn and Gavin Andresen around BIP-101, which would have raised the cap to 8 MB and then kept doubling it. The idea was that if enough miners and users ran XT, its rules would simply become Bitcoin. It got a burst of attention, then faded.

Bitcoin Classic followed in early 2016 with a more modest proposal, a straight bump to 2 MB. Bitcoin Unlimited , the most ambitious of the three, tried to remove the fixed limit altogether and replace it with something called "emergent consensus," where node operators and miners would each set their own preferred size and the network would supposedly converge on a number on its own. It had first appeared around the start of 2016, and by 2017 it had become the big-block camp's flagship.

Bitcoin Unlimited is also where this approach hit a wall in the most embarrassing way possible. In March 2017 a bug in its code was exploited, and most of the Bitcoin Unlimited nodes on the network crashed at once. The count dropped from about 780 to around 250 before patched nodes came back online. The cause was a debugging check that the developers had left active in production software,...