The ongoing bitcoin forking controversy over Bitcoin Unlimited vs SegWit is a clash of cultures, but not one of East (miners) vs West (developers) – rather, it’s the classic culture clash between hardware engineering and software maintenance, according to bitcoin evangelist Andreas Antonopoulos.
Speaking at a Bitcoin and Ethereum Singapore meetup event at Singapore Management University, Andreas Antonopoulos explained that the East vs West analogy stems from the fact that most miners are in China, while most developers are from the West. But the bitcoin culture gap isn’t a product of cultural geography, Antonopoulos said – a better analogy is the culture gap between people who build hardware and people who build software that has been diverging since the 1950s.
Simply put, hardware is all about spending years getting everything right before releasing it into the wild, after which it’s troublesome and expensive to repair if something goes wrong. Software is all about beta releases and instant patches to fix the bugs as you go along.
Thus, hardware culture is very conservative. Hardware requires absolute precision. A mistake cannot be fixed with a rewrite. (A contemporary example: Samsung’s Galaxy Note 7 fiasco.) Software culture is far more relaxed, as mistakes can be rectified in less than two minutes.
This basic difference in mentality essentially applies to the division over Bitcoin Unlimited vs Segwit, Antonopoulos said. Bitcoin Unlimited (backed by hardware-intensive Chinese miners) is ironically taking the software mentality of pushing out untested solutions quickly to solve the pressing capacity problem and it is the Bitcoin-Core software developers who, because of the permanence of the blockchain, are now taking a hardware mentality trying to get everything right first before deploying the software upgrade.
The emergence of trustware
Interestingly, said Antonopoulos, the emergence of bitcoin has spawned a new emergent field he calls “trustware”, where you have consensus rules that are software with the backing of hardware with a diverse set of participants.
“It has all the headaches of software, hardware and then some,” he explained.
This means that every mistake now gets baked into the blockchain and is carried forward forever. This has led to the saying, “In Bitcoin there are no bugs. There are consensus rules created by tradition”.
For example, when multisig (transactions that needed more than one signatory to spend) was introduced, the OpCheckMultiSig function had a bug that popped an extra variable off the stack – where three signatures were expected it would pop four. This was discovered only after the software had gone live and users had transferred bitcoin to multisig addresses. Fixing the bug would be worse than the workaround, as those transactions need to be valid forever. So, in order to redeem, they have to put in an extra dummy signature.
This is the essence of trustware, Antonopoulos said. We are now writing software that is backed by hardware and is deployed on a network with a set of global consensus rules, and if someone makes a mistake, someone may lose millions. This is not a game – this is a new frontier.
All of this pain is worth is as it gives the world something amazing – a decentralized platform of trust that is neutral and not controlled by anyone.
In any case, regarding Bitcoin Unlimited vs SegWit, Antonopoulos said he is firmly on the side of Segwit, as Bitcoin Unlimited’s emergent consensus algorithm is “completely untested” and opens up a new venue of attack. If the Bitcoin Unlimited camp chooses to go with a 51% attack, he said, it’s questionable whether Bitcoin Unlimited could sustain that 51% hashrate in an environment where bitcoin is under constant attack on just about every front.
“I don’t think people who are contemplating the hard fork at 51% have thought through the implications,” he said.
That said, however, it will create a scenario where bitcoin can test all the attack scenarios – transaction malleability, replay attacks and more. It may result in bitcoin being worth a lot less, but technically the community will have a battle-hardened code that is worth a lot more.
After Antonopoulos’ talk, Disruptive.Asia approached Ethereum co-founder Vitalik Buterin to ask him how Antonopoulos’ hardware/software divide model fit in with the fork that Ethereum went through earlier.
Ethereum is about to move from “proof of work” to “proof of stake”, in which the consensus mechanism would be in the hands of the developers. Ethereum Classic would continue on proof of work but with a key difference: the hardware is generic, off-the-shelf rather than custom silicon as in the bitcoin world.
Antonopoulos had in his talk noted that the circumstances in the Ethereum fork were different in that the developers had 80% of miners on their side, unlike in the current situation with bitcoin.
Buterin did not answer directly but said, “I disagree completely with everything [Antonopoulos] has just said. Everything he said was completely wrong.”
He explained that the ones advocating the move-fast-and-break-things model were the hardware-backed miners and the conservative, must-get-it-right-first-time camp of SegWit were the bitcoin-core software developers.
Later, Antonopoulos responded that it was not the first time Buterin had disagreed with him completely.
Be the first to comment