I think this can actually explain a lot of past hard fork dynamics. Without an explicitly agreed process for agreement, people have to predict the implicit process that others are most likely to agree on and resort to. (1/7) https://social.that.world/@wei/106336302936496207
What is the most common thing that a decentralized community share? The software they're running! It's often no problem for people to download software upgrades from the same developer, but there are a lot of issues (trust, coordination, etc) for people to switch repos. (2/7)
This is why in the past hard fork dynamics, Bitcoin2x didn't happen; Ethereum Classic, as well as Bitcoin Cash and later Bitcoin SV have to resort to different tickers. You can also use the same guideline to predict the outcome of the "EIP-1559 vs. miners" war. (3/7)
On one hand, it makes hard fork relatively stable in that you always know who'll continue to "be the original coin". On the other hand, every controversial hard fork becomes a mess because the implicit "agreement process" will always be doubted by everyone, including devs. (5/7)
Hard fork governance is like having a ruler that you cannot change. For the majority of time, the ruler will act according to people's will, because he wants to stay in power. However, the only way to replace that ruler, is to throw by force. (6/7)
In this sense, on-chain governance may not be perfect, but in most cases, including for base-layer blockchains, it is better than hard fork governance, because you can always improve the rules and draw a much larger opinion sample to reach an enforceable decision. (7/7)
Many of the past hard fork dynamics are depicted as coin holders vs. miners, but if you look closely, a more accurate assessment would be developers vs. an opposing party. (4/7)