Finally getting the chance to start into this thread. Interested to learn that I have already been quoted several times. Something downthread may obviate what I am about to state. We shall see.
So why is BU preaching equilibrium with unlimited blocks as the most prominent item in its FAQ.
Because they want to sustain the illusion of decentralization. Their supporting white paper is meaningless and doesn't model anything that can exist in reality (it models a perfectly uniform distribution of propagation and hashrate where all miners experience the same orphan rate, but if that were the case then no miner would make any profit because in that impossible scenario marginal cost = lowest cost).
Well, let's think about how progress is made in academia. Most research is started upon a simple model. The implications of this model are followed unto their logical conclusion. Later, same or another author will build a more sophisticated model, and apply the same techniques to it to their logical conclusion. And on and on. Truth is approached incrementally.
Peter R has freely stated that he has not modeled all of reality in his first supply/demand paper. It does not make the paper _false_, nor _useless_ -- just limited in applicability.
So with that said, I challenge you to justify the leap from '
the aggregate S/D curve is not identical to the sum of individual miners' S/D curves, because each miner's S/D curve varies from other miners' to '
BU's emergent consensus is unworkable and broken'. I don't mean to put words in your mouth, but I think that is your claim? Certainly, miners already need to deal with a panoply of nonuniform externalities (e.g., $/KJ, $/employee-hour, land costs...). Their orphaning probability curve is just one small contribution to their overall cost model, which all must be factored into the point where marginal-orphan-cost = marginal-transaction-fee.
I think you are fundamentally missing the expected dynamics of LN. ...
Only big hubs can survive in the fee competition on the LN.
I just wanted to acknowledge that this is the most simple, cogent, and concise explanation of the 'LN centralization problem' that I have yet seen. Kudos.
No Bitcoin can't remain Satoshi's 1MB without SegWit because
I already explained that "no change" is not an option.
Your linked post essentially stated a case that there are interests within Bitcoin whose will is irresistible. But you did not go on to make the implied inferior antecedent case that their will includes a change to Bitcoin.
Let alone the fact that I have serious doubts about MP's unverified self-claim of omnipotence - aren't he and his acolytes running a patched version of bitcoin-qt 0.63 or somesuch? Hardly seems like a demand for changing the protocol.
the existence of a mining cartel is a big deal, as it threatens the 21 million coin supply limit.
Firstly, any change brought on by a majority of miners -- all acting quite obviously in their individual self interest -- does not imply the collusion of a cartel. Some market behaviors develop without communication between the principals. Of course, this is not to say the existence of a cartel is an impossibility, but it is to say that the existence of a change beneficial to miners is insufficient evidence of cartelization.
Second, even if a cartel would develop, that cartel is not free to enact whatever change they so desire. The 21M limit is one of the very few parameters that are all-but-universally held to be sacrosanct. If the miners were to enact an inflationary change, I would expect Bitcoiners to abandon Bitcoin en masse, thereby leaving the miners destitute with exorbitant unrecoverable sunk costs in hardware - unfit for any purpose other than transmuting expensive electricity into a worthless Bitcoin.
I don't know whether the 1MB limit was SOLD as a spam measure but was conceived as something to guarantee the fees, or whether it was just "idiot leadership". I don't know whether the slow difficulty readjustment was conceived to avoid people forking, or also just "idiot leadership". But in any case, these measures shaped bitcoin in what it is in a far more important way than one would have naively thought when they were proposed.
Indeed, the 1MB limit has had a huge impact upon the users of the system.
I think you are relatively new here. At least having arrived after the 1MB production quota started having a real effect upon the system. But to at least a subset of the participants, it was a given that the abolition of this limitation would be enacted before it became significant. To many, this assumption was indeed part of our 'contract'.
I guess it is our own fault for not having believed 'what it said on the tin'. However, whether it was this particular change, or the widespread notion that Bitcoin could adopt any innovation cooke up in the lab of altcoins, pretty much nobody thought Bitcoin immutable. At the time. I guess we've all got egg on our faces.
And indeed, I used to be of the opinion that the long difficulty retargeting was a stupid hack only used to make the code implementation easier than a smoother curve. I now at least understand that it has important implications related to changing the system. If this be intentional, this has me even more in awe of Satoshi's prescience.
Uh, I stop you right there. If you don't have the full block, you cannot verify the correctness of the block summary. Building upon a non-verified block header is just as risky as waiting for the full block to arrive.
You misunderstand XThin. In XThin, all transactions are individually validated by validating nodes
before the block they are contained in is solved. Upon solution, the only thing requiring propagation is the list of transactions within the block (and the solution of course). If any validating node is missing a given transaction, it sends back a bloom filter including specifiers for the missing blocks. This drastically reduces propagation time for the block solution. See
https://medium.com/@peter_r/towards-massive-on-chain-scaling-presenting-our-block-propagation-results-with-xthin-da54e55dc0e4 for a better explanation of the design, the theory, and experimental results.
And yes, this reduces the propagation impedance by a significant amount. But it does not reduce it to zero. Accordingly, while it shifts the point of intersection of the BTC/blocksize curve and orphan/blocksize curve, it does not eliminate the intersection thereof.
ANY pile of non-contradictory transactions is good enough, they don't need to be fully ordered, they only need not to be contradictory.
What mechanism other than agreed ordering is there to ensure the non-contradictory attribute? If differing nodes have different understandings of transaction ordering, then double spends are enabled.
/* I almost fat-fingered this post into oblivion. Hence, I am posting this interim missive before I catch up to the head of discussion*/