The latest twist in bitcoin’s story is a proposal by Joseph Poon, Lightning Network Developer, Stephen Pair from BitPay, Christopher Jeffrey from Purse and Fedor Indutny. They publicly announced today a soft-fork blocksize increase to “3MB, up to 9 total,” according to Poon, as well as an implementation of “a malleability fix whose individual component is smaller than segwit, as it does not use the weights and discounting,” – Poon says.
It’s not quite what Bitcoin Unlimited or Bitcoin Core wants, but it appears to give them both a great deal of their stated aims, potentially breaching the divide.
Some current and former Blockstream employees have seemingly come out against it, but other Bitcoin Core developers appear more measured in their initial reaction. Eric Lombrozo, a Bitcoin Core developer, publicly stated:
“I am happy to see work being done on extension block ideas. This has been one of the potential routes being explored for future block size increases for a long time.
However, I am concerned that trying to combine segwit and extension blocks into a single proposal will lead to further development delays since any such proposal will still require review, obligatory bikeshedding (i.e. determining DoS parameters), testing, implementation, building ecosystem support, adapting existing applications, etc…”
Replying in a different context to the argument that it would take time to test and implement, Poon stated that “it took [Christopher Jeffrey] two weeks to implement extension blocks, we can test and deploy within months.” He further said:
“Lightning is safer with some fixes on the extension block. Those changes may be too aggressive on the 1mb chain. I believe the extension block should get changes first before it goes to bitcoin, can backport in the future.
A malleability fix simpler than SW is provided on the extension block and Lightning on the extension block is superior to SW.
This proposal supersedes BIP 141, as it already provides a comprehensive framework for smart contracts such as lightning network.
The extension block proposal is a soft fork path forward for Bitcoin. It is a litmus test for corruption in the Bitcoin community. The miners have passed with flying colors in their public interest in Extension Blocks with positive intent and an ability to compromise for the betterment of Bitcoin. Core is also getting what they want (soft fork plus malleability fix), but oppose due to nonsensical reasons. We as a community need to move forward.
I encourage industry stakeholders to publicly express support to accelerate the social mandate for miners. Without the ability to make pragmatic decisions Bitcoin as a project is stagnant and dying.
We as a community are not obligated to support BIP141. I ask the community to support extension blocks with humility and hope it gives everyone’s publicly stated goals for the past two years.”
BIP141 can still be activated with small changes, but Poon’s statement that extension blocks are superior to segregated witnesses (segwit) for the Lightning Network is somewhat surprising as the main selling point of segwit has been its ability to allow the implementation of the Lightning Network and therefore potentially significantly increase off-chain capacity.
Moreover, as the proposal is a soft-fork, the main argument against bigger blocks, namely that it is a hardfork and therefore requires the consensus of literally everyone, falls away as some Bitcoin Core developers have argued that for a soft-fork you do not need widespread consensus because it is opt-in.
A similar proposal has also attracted the backing of Johnson Lau in a modified form as he suggested it earlier this year, thus indicating that a potential significant consensus can be achieved.
Has a bunny, therefore, come out of a hat? Can this really break the stalemate? Will bitcoin really reach a compromise and increase its capacity? We do not yet know, but what is obvious is that this debate cannot continue for much longer. Businesses are leaving, bitcoin’s market share keeps falling, the debate itself is slowly becoming very boring.