pull down to refresh
Every time we alter these block validation rules, don't we run the risk of losing our Byzantine General protections?
No, you just risk ending up with 2 chains and most sane people agree that that is an undesirable outcome. I think that that's the other thing we may need to do besides respecting those that disagree with us: acting for the better of the whole, including those that disagree with us.
Isn't that where many proposals that come from more respectfully operating people (that may still ragequit) get stranded? Where the benefits (for everyone) don't clearly outweigh the risks (for everyone)? I think in these scenarios too that this is not a bad concept at all and that there, to date, is some balance, and maybe it even tips just a little bit to the influential-changer, looking at taproot, which in hindsight could have used maybe one more wide iteration before proposition (though maybe that really is only in hindsight - activation ultimately happened so quickly that I was astonished by it.)
BIP-110
I don't think that Luke after so many years stopped understanding the mechanics around how chaintips work. Instead I just think they overestimated their importance outside of their echo chamber. This is a fairly common mistake when people plan for coups, and why throughout history, many a general has been imprisoned, exiled, beheaded or otherwise disposed of.
But we still have a Byzantine Generals problem when we try to talk about block validation rule changes.
If before you commence to battle, 95% of your generals are telling you they're not going to obey you, is there really a Byzantine Generals problem at all? I think that if we watch reality as if everyone doing a proposal is a retarded child that will rather fork off than just not get what they want, then yes, we're in deep shit. But I've only really seen this once before and that was even more emotional than this.
chuckled out loud at that last paragraph.
I think I need to take my Byzantine General speculations back to the drawing board. I am still quite fascinated by the way we negotiate changes to bitcoins rules, but I always find that I haven't yet found a really good model of it in my head. (This despite initiatives like BCAP and the like.)
I made a poster about this:
By underestimating the value of the already-agreed-upon nature of the current validation rule-set, I think I was trying to get at an idea I've been struggling with all summer:
Bitcoin allows us to do something pretty incredible and that is to agree on the state of a ledger though we don't trust any of the other editors (the Byzantine General thing). But we can only do this so long as we all use the same block validation rules. Every time we alter these block validation rules, don't we run the risk of losing our Byzantine General protections?
The method Satoshi developed of solving the Byzantine Generals problem cannot be used when we negotiate changes to block validation rules (I think this was actually a crucial misunderstanding on the part of the BIP 110 people -- they thought heaviest chain mechanics could be used to "vote" on block validation rule changes, and that was just completely misguided).
But we still have a Byzantine Generals problem when we try to talk about block validation rule changes: we don't trust any of the actors in the debate. So, changes to the rules are fraught with danger. Your emphasis on a very high bar for measuring consensus seems like a potential way of avoiding the problem, but there's no real way to measure such a thing. Aren't we always kinda living on a prayer when it comes to miner signalling or thigns like that?
I wondered about needing a new tune because I don't think I've really come to a much deeper understanding of this topic in the last year and I feel that writing new versions of the same thoughts over and over is offputting (even if the thoughts are sincere).