Bitcoin is on the road to increasing its capacity – but it’s not there yet.
Despite the claim bitcoin has been upgraded to support Segregated Witness-style transactions, there are more steps that need to occur before its distributed network can process additional data.
In fact, due to the complexities of keeping bitcoin’s distributed network in unison, the change might not take hold until late August.
The timeline
That said, with BIP 91 activated, bitcoin is now headed down a path toward that possible outcome.
In context, this means that BIP 148, a proposal that seeks to upgrade the network to SegWit using different code, is no longer likely (though it’s still possible). As such, one avenue that would find bitcoin splitting into two blockchains has been diminished, though it would be wrong to say it has been eliminated entirely – even in the short term.
Rather, SegWit now needs to follow a timeline designed to give miners and node operators ample time to upgrade their software, one that must also operate within the larger constraints of bitcoin’s design.
Here’s a best-case scenario of how this might unfold:
As illustrated above, miners must now signal support for SegWit until the code achieves “lock in” and “activation,” two separate milestones with unique requirements.
Short-term complications
Still, things could go awry.
For example, it’s possible that bitcoin’s miners could stop signaling support for SegWit ahead of the BIP 141 “lock in” deadline (although they’d risk having their blocks rejected by the network, losing the rewards).
While all nodes appear to be signaling correctly, it’s hard to know exactly who is running the software – this means that miners could fail to reject a block that wasn’t signaling for SegWit, continue adding new blocks on top of that block, and ultimately, produce an alternative chain.
Adding to the argument is that miner support for the idea has notably wavered in the past, with some asserting it’s largely the threat that users could push through a change that could lead to a split that has kept miners in check.
Litecoin’s attempts at integrating SegWit adds context to this theory.
As much as 75% of litecoin’s mining hashrate, for instance, was signaling for SegWit in April, meaning it reached the necessary threshold to lock in SegWit. But miner support dropped off soon after. This sparked protocol creator Charlie Lee and other litecoin users to threaten to code an alternative user-activated soft fork (UASF) proposal.
Shortly thereafter, a roundtable was held that united major miners, at which time they agreed to signal support for SegWit.
Should similar psychology come into play here, the idea is that (after averting a UASF on bitcoin on August 1), miner support for SegWit may wane, though the economic stakes here are arguably higher.
Long-term complications
But even if SegWit passes, that still only enacts one-half of the Segwit2x proposal.
An informal agreement among businesses and miners reached in May, the goal of the initiative was to push the capabilities of the software even further (some would argue too far beyond what should be advocated by prudent development).
As developer Jeff Garzik put forward in an interview this weekend, he still plans to help push forward an effort that would seek to upgrade the block size to 2MB, and to enact the upgrade via a process known as a hard fork.
The argument against that approach is that research in this area is still underdeveloped, though Garzik has argued SegWit could fail to introduce the capacity changes promised by those who have championed its integration.
This determination (as well as the acrimony Garzik’s outlook has caused in the developer community that argue this claim and others he’s made are false) could make a split later in the year more likely.
And it’s quite possible this chart will become more complex as this date approaches and different groups attempt to influence the outcome.
Disclosure: CoinDesk is a subsidiary of Digital Currency Group, which acted as organizer for the Segwit2x proposal.
Old compass image via Shutterstock