Bitcoin Classic proposes a hard fork that would increase the blockchain's block size limit to 2 MB. It triggers when 750 of the last 1000 blocks announce support via coinbase transactions. Gavin Andr
Bitcoin Classic proposes a hard fork that would increase the blockchain's block size limit to 2 MB. It triggers when 750 of the last 1000 blocks announce support via coinbase transactions.
Gavin Andresen, the Bitcoin Classic developer, wrote about hard fork safety in the context of 75 percent network support. He argues that miners and economic participants have strong incentives to join the larger fork after a split occurs. Andresen prefers the 75 percent threshold because he wants to prevent a single large mining pool from blocking changes to Bitcoin's consensus rules.
Peter Todd, a contributor to Bitcoin Core, sees things differently. In an interview on The Bitcoin Game, Todd laid out several potential problems with activating a hard fork at 75 percent support.
Todd's first concern is straightforward: nobody can predict what happens once the fork launches. He said: "I think the problem is we don't know." He pointed to a Bitcoin development mailing list thread where Alex Morcos, a Chaincode Labs co-founder and Bitcoin Core developer, challenged Andresen's optimism. Morcos wrote: "In truth, no one has any idea what would happen if the proposed Classic hard fork activated with 75% right now. There is some chance you are right, but there is a very legitimate possibility that a concerted effort would arise to maintain a minority fork or perhaps if miners don't see nearly a complete switch over, many of them might themselves reverse the fork if they think it would be easier to achieve consensus that way. We as a community have never been in such a situation before and it behooves us to speak honestly and directly about the uncertainty of the situation."
Todd won't accept unknown risks at this scale. He said: "When you're working with billions of dollars, you don't take risks lightly." He also stated: "I think Gavin is willing to take much bigger risks than most people."
Bitcoin Core supporters argue that the economic majority, not miners, controls the network. A 75 percent threshold doesn't mean bitcoin holders support the change. Users can still reject the fork after miners activate it.
Todd identified another flaw in Bitcoin Classic's approach. An opposing miner or mining pool could announce support, then retract it after the 28-day countdown begins. Miners would activate the fork with lower hash power supporting it than claimed. Other miners might panic and revert to Bitcoin Core.
On the development mailing list, Todd highlighted the "Not Bitcoin Classic" problem. Developers first deployed this tactic in summer 2015 in response to Bitcoin XT. The idea involves nodes faking support for Classic without any intention of following through once the fork activates. It could trigger chaos, a premature hard fork, or both.
Andresen claimed that a 95 percent or higher threshold would let a single miner or pool veto consensus changes. Todd countered that a 75 percent threshold excludes the remaining 25 percent of hash power. He explained: "If you're willing to go through with something like this, even though we have a quarter of the hashing power disagreeing with you, you're disenfranchising a lot of people. You're telling a lot of people, 'We don't care about your thoughts. We're not going to try to compromise with you. We're just going to go ahead with this.' In a system that needs consensus, that's a pretty dangerous thing to be doing."
The Bitcoin Roundtable sent a letter opposing Bitcoin Classic as the reference implementation. Todd pointed to this letter as evidence that many within the community oppose the plan. He said: "Right there, that's enough to say this Bitcoin Classic plan is probably not going to happen that way."
Todd also mentioned a proof-of-stake vote on Bitcoinocracy.com. Holders of just under $12 million in bitcoin have signed a statement claiming that major bitcoin holders would crash the price if a non-Bitcoin Core implementation activates a hard fork. The site needs more voting activity before anyone can draw firm conclusions.
Throughout the interview, Todd made clear that his main objection is the 75 percent threshold, not the 2 MB block size increase. He explained his preference for a much higher threshold: "Why don't you just release it with a higher threshold? I may not necessarily agree with what the change is, but if we, for instance, had 99 percent of hashing power agreeing to a hard fork, what happens to the 1 percent chain? 1 percent hashing power basically means you get a confirmation every sixteen hours or so. That's not a viable chain."
Todd would conduct a hard fork differently to maximize safety. He concluded: "You're never going to get 100 percent, but 95 percent or 99 percent — that's a heck of a lot safer than having a quarter of all hashing power potentially disagreeing with you."
