By David Seaman
The Bitcoin block size debate and ensuing community chaos seem to have come out of almost nowhere, but if you’ve been following the math, such a discussion was arguably overdue.
According to data collected by TradeBlock and cited on CoinDesk, average block size has jumped from approximately 125KB to 425KB since 2013.
And more important, since the start of 2015 about 3% of the time normal, non-malicious transactions experience confirmation delays due to not fitting in the next available block.
So clearly this is not a completely frivolous topic that the community has become fixated on, instead it’s a crucial infrastructure decision.
Some in the community believe Bitcoin is becoming too visible and important–especially now that the R3 blockchain exploration deal with 9 of the world’s largest banks has been made public — to let questions like “will our blocks soon be full and transactions rejected?” linger in the air, with complete uncertainty as to the answer.
Easiest way to get your first bitcoin (Ad)
The Bitcoin XT fork, favored by Gavin Andresen, would raise Bitcoin’s existing 1 MB limit to 8 MB limit, with additional limit increases every two years into the 2030s to account for expected advances in computer hardware and transaction volume.
It is worth noting Bitcoin’s block size limit as originally designed was 36 MB, scaled down to 1 MB during its second year as an anti-spam and anti-DDoS measure. So there doesn’t appear to be anything ideologically impure about raising the block size, most of the community bitterness springs from whether the move is technically impure… in other words, is this a bad engineering decision?
Leaving the network alone, with no changes, would seemingly lead to an upwardly competitive market for transaction fees — if you want your transaction pushed through, you’ll have to pay more than small faucet transactions and dust on the network. To those who like the idea of preserving Bitcoin’s psychological scarcity, this could conceivably lead to an environment where Bitcoin is like a gold vault or deposit box, used only when needed, with large sums involved. This could be good, if people start using Bitcoin for their big ticket transactions all the time. Or it could be bad, if people stop using Bitcoin altogether in favor of a “cheaper” network that only charges a few cents for everyday transactions to go through quickly.
Other options include increasing the community’s reliance on tools like the Lightning Network, so that more small value transactions can be “pre-settled” offchain and then broadcast onto the chain at certain times to make those transactions public and permanent on the blockchain.
Is this debate a good sign for Bitcoin?
It may seem counterintuitive to think of all this as really good news, since choosing the wrong path here could lead to a currency that collapses because its network is too congested at some point in the future… or to a currency that collapses because its network becomes too expensive for everyday users to interact with, limiting Bitcoin’s future network effect.
First of all, I have some confidence that there’s enough intellectual capital within the Bitcoin community at the development level to figure this challenge out, and enough money at stake in terms of interested third parties like Coinbase, Circle, 21 Inc. and Tim Draper wanting to see the right outcome… No one’s going to break Bitcoin here, at least not intentionally. That’s not in anybody’s interest.
And where this is also good news: they’re finally stampeding over each other to our doorstep, aren’t they? While users may not want Bitcoin to be known as the transaction system where 3% of good transactions get bumped because the network is too congested right now, that is kind of a resounding ENDORSEMENT of how popular and widely used Bitcoin is becoming.
Transaction volume keeps rising. So this problem, of how do we expand the currency to let more transactions in, is obviously a better problem to have than near empty blocks.
A related debate: block times
As the Bitcoin block size debate rages on, a separate although related debate has to varying degrees consumed the cryptocurrency community. And that debate is not about block size, but block times.
A number of alt community members have tried to “improve” upon Bitcoin’s 10 minute block time by offering shorter block times — Charlie Lee’s Litecoin which was introduced in 2011, for example, has a 2.5 minute block time, and some newer proof-of-stake currencies have block times as fast as 30 seconds.
I asked Nick Szabo about this, and he wrote back saying that block time is an important security parameter. He likened a longer block time to more armor on an armored car. “It goes slower, can carry fewer pieces of paper, but is more secure.”
I’d never thought about it like that before, nor had I thought of block time itself as a core security feature of cryptocurrency, but of course it is. The longer things are in the air before the transactions are confirmed and permanently written onto the blockchain, the lower the odds of various kinds of attack.
“Altcoins with short block times are often taking too many risks,” Szabo also warned.
In a way, these views on block time may hint at the right decision for block size as well — maybe a small block size, like a slower block time, is part of what keeps cryptocurrency flexible and impervious to attack.
Either way, it’s exciting to be a user of Bitcoin in the early days, while these pioneering technical questions get hashed out. We will probably look back on these discussions fondly in the future, either because we massively overestimated how big the Bitcoin network would some day become, or because we’ve all massively underestimated how popular the blockchain will some day become.
David Seaman is a journalist covering the emergence of cryptocurrency. He writes for Coinivore and many other publications. Follow him on Twitter or YouTube.