Scalability Debate Continues As Bitcoin XT Proposal Stalls

Should a certain controversial alternative to the mainstream implementation of bitcoin's code have gained traction, today might have marked a significant date in the bitcoin calendar. However, that was not to be.

As many industry observers know, the open-source bitcoin community remains engaged in a months-long struggle to determine how best to increase the capacity of the transaction network, and as of 11th January, that discussion is still ongoing.

Initially the date set by project developers Gavin Andresen and Mike Hearn, 11th January was to be the earliest possible time that Bitcoin XT would have begun introducing larger 8MB blocks to bitcoin users running the XT software. Others running Bitcoin Core would still process 1MB blocks, a development observers argued amounted to a split of the network.

Regardless of the proposals, there is consensus in the bitcoin community that a change is necessary because of the perceived risks to bitcoin as a payment system should daily transactions increase towards the network's 1MB limit. At this point, users would be forced to more actively choose the fee they would pay to process the transaction on the blockchain, essentially making more prominent the postage stamp charge that comes with every message.

The current block size of 1MB means that in the near future it is possible that the network could effectively become clogged up, leaving transactions delayed or even failing altogether. Such instances have already happened, as highlighted by spammers, who have in the past pushed the network to capacity.

If a sufficient number of all the bitcoin node owners had chosen to adopt XT – 75% to be precise – Bitcoin Improvement Protocol 101 (BIP101) would have become active and the block size for those running that software would also have begun to climb . This would have seen block size jump from the present 1MB up to 8MB, and doubling every two years until a block size of 8GB was reached.

But this hasn't happened. Today, just 10% or so of worldwide nodes have converted to XT. And despite support from some notable companies including Coinbase, BitPay, Circle and Blockchain, bitcoin's miners have largely not come on board.

When asked what the lack of consensus on the XT release means for bitcoin, Hearn, who now has minimal involvement with XT, told CoinDesk in an email that he still believes capacity to be a problem on the bitcoin network.

In particular, he cited the fact that bitcoin's miners have demonstrated a willingness to align with decisions made by bitcoin's Core developers, the open-source meritocracy that oversees code changes.

"Bitcoin can't be credibly described any longer as a decentralised system. In how it operates and how much influence users and merchants have, it is indistinguishable from any other proprietary payment network," he argued.

Hearn is now working with blockchain startup R3, which is working to adapt the technology for use by enterprise financial institutions.

Of course, if 75% of nodes switch to XT at some time in the future, BIP101 would still effectively kick in. But that eventuality is not looking likely, going by bot data and comments from some within the industry.

"It appears that most miners are in agreement that BIP101 is too much too fast and tries to predict too far into the future what an appropriate block cap will be," said BitGo engineer Jameson Lopp. "Bitcoin has a strong status quo – XT has shown just how difficult it is to overcome."

Core developer BTCDrak agreed, indicating his belief that BIP 101 was "too aggressive", especially for miners, though this group had earlier stated they could handle 8MB blocks.

Still, he framed a lack of support as the predominant issue:

In his comments, Andresen told CoinDesk that bitcoin users should be more proactive about what they want out of the software they run, statements that echo his call for the bitcoin network to support multiple implementations.

"The community should tell developers of the software they are using what they want. If the software developers can't or won't give it to them, then they should switch software," he said.

Notably, Andresen suggested he is open tweaking his proposal or releasing a new BIP as necessary.

Despite the differing opinions, however, it appears widely agreed that changes to the bitcoin network's transaction capacity should take place.

To supporters of such changes, it's less of a question of if it will happen, and more a question of how and when.

Other proposals include a method called 'segregated witness' (popularly shorted to 'SegWit') first proposed by Bitcoin Core maintainer and Blockstream co-founder Pieter Wuille.

This would could make transactions appear smaller to current nodes on the network, in theory making a 1MB block become equivalent to 4MB (although in practice actually more like a maximum of 2MB). The measure, roughly equivalent to reorganizing the closet as opposed to buying a bigger one, has attracted support since debuting at Scaling Bitcoin Hong Kong last year.

Yet another solution, a so-called '2-4-8' plan, has drawn interest from supporters like BTCC, a China-based bitcoin mining pool and exchange, as a more modest means to raise the block size limit.

However, it remains possible that both solutions could be pursued simultaneously due to the different approaches the proposals take to solving the issue.

Arrow image via Shutterstock