05:53
MrMonkey — Today at 9:38 PM
you wanna know what I really think? you don't have to see the code for that....but it's complicated 😛 but here goes.
As you know, bob just burns cycles while bil does real work, looking for hashes. The main difference from a network perspective is that actually doing that type of intensive work (looking for hashes basically behaves like an infinite loop and is very intense for the cpu) tends to bring the network to its knees.
But isn't looking for hashes better? More fair somehow, or something? Like bitcoin?
Well, on BTC the reason hash searching means it's fair is indeed that anyone can do it. You can use your gfx card or a super computer, or pen and pencil, and the only thing that affects is your chances of finding it first, in such a way that you can know that anyone on earth who comes running with the solution to the hash, must indeed have spent some rough amount of what it should cost to find it, on average (they could be very lucky, with pen and paper...once.)
BTC PoW is really a way to create a fair method for proving /cost/ - there is no way to cheat, to keep finding hashes you must spend the money on the energy, and the cool part is that it is a totally permissionless system that anyone can join with whatever new method they cook up for finding hashes faster (which is why BTC also drives energy efficiency innovation)
But back to BIL. So, here's some FUD (that I will kind of disprove).
"Since BIL actually solves hashes, I could set my own computers at mining hashes more efficiently and win every block!" - in principle, this is true!
But BIL would block that, in this way: They only accept solutions from their well known miner canisters, that they control. Which in turn means they know they code on those canisters, and can make sure they don't have code that allows someone to "cheat" and feed a solved hash to that miner that the miner can report to win the block.
So, in the end, BIL's model 100% boils down to the fact that they control all the miners, otherwise this exploit would be possible and obvious.
But then, given that the whole model is based on controlling the miners, and remembering that we talked about how actually searching for hashes brings the network to its knees...couldn't we force all the miners to work a little less hard, in a fair way where they all have to work less hard? Couldn't we maybe hash, like..a bit? And maybe burn some gas a bit? In some kind of BOB-BIL hybrid?
But then you go more and more in that direction and realize the hashing isn't really needed. Like I said with BTC, what you're trying to do in the end is create a system with Proof of Cost. For BTC, hashing is a good Proof of Cost via PoW that also allows anyone to participate with whatever hardware. On ICP, BOB is Proof of Cost via Proof of Burn, which is a fair system and the main difference against BTC is that it doesn't allow anyone to participate with their own hardware. And neither does BIL.
So BIL doesn't really gain an advantage over BOB by doing it's PoC via PoW instead of PoB. Both projects have the same reach (ICP users) and the same level of PoC fairness, just that one brings the network (needlessly, one could argue) to its knees and the other doesn't.
I suspect, btw, that BOB launched as PoW, saw the network go down, thought "oh shit", went through roughly this reasoning ^^ and decided PoB was as good as PoW for this use case. And that's why the network survived 😛
Anyway, this means that I actually think between BOB and BIL, maybe BOB > BIL, but ofc BTC >> BOB > BIL