20 MB blocks: an unfortunate compromise

Monday 15 June 2015

The bitcoin network is approaching saturation - but it needs to work smarter, not just harder.

There is a saying, misattributed to Einstein, that you don’t solve a problem by using the same thinking that got you into it. That’s an apt maxim to remember as we consider the problem of bitcoin block sizes.


20 MB blocks: capacious, but not something you'd want sitting on your hard drive

Read also: Spotlight on Ramchains

As things stand, bitcoin blocks are 1 MB. This gives a maximum volume of transactions that the network can handle per second. Neither is this a theoretical maximum that we don’t have to worry about for some years: the network is approaching capacity now, and if bitcoin is to scale to become a widespread payment method, the problem needs fixing as a matter of urgency. Blocks aren’t full, by any stretch - in fact, on average they’re only a third full. The problems occur at the edges of distribution, where random variation means you get a full block - perhaps because lots of transactions happen to occur at the same time. Some transactions get delayed. The issue will only get worse with greater transaction volumes.

As Gavin Andresen, bitcoin’s core developer, says, ‘If the number of transactions waiting gets large enough, the end result will be an over-saturated network, busy doing nothing productive. I don’t think that is likely – it is more likely people just stop using Bitcoin because transaction confirmation becomes increasingly unreliable.’

The most obvious solution - and the one Andresen advocates - is to increase block sizes. Not just a little: to 20 MB. That is the simplest answer given the difficulties of making changes to the bitcoin protocol. You have to be sure that miners will accept them, because if you can’t get 51% mining on the same fork, the changes won’t take. The change needs to be made soon in order to get enough miners on board by the time the hardfork takes effect, at a block height some months down the line.

This makes some sense, but it involves an increasing centralisation of the blockchain around those miners who can afford the hard disk space and bandwidth. The bitcoin blockchain is already unwieldy, at around 30 GB. It would balloon under the new proposals, putting it out of reach of regular users and meaning that only those with the biggest resources could download and maintain it.

The alternative? Work smarter, not just harder. Lightweight, super-fast blockchains are being developed. Ramchains, as they’re known, are still in their infancy. But they offer enormous benefits of speed and scalability over traditional blockchains.

Increasing block size is a temporary solution that uses conventional thinking. It will not last, and it will make the bitcoin protocol even more unwieldy. It may be necessary in the short term - in fact, in the short term, it's the only reasonable solution - but if bitcoin is going to succeed then we need some out-of-the-box thinking here.

Brandon Hurst

comments powered by Disqus