> The core problem is scalability It is indeed the limiting factor for any public blockchain based system. The rest of your comment describes capacity however,

> The core problem is scalability It is indeed the limiting factor for any public blockchain based system. The rest of your comment describes capacity however, which is something else. We can increase capacity without increasing scalability. > Miners are very incentivized to fix it, This is false. Miners are the ones who stand to gain from high fees. And before you say more capacity would lead to more transactions and higher fees, run the numbers. Compare with less developed chains. It is quite clear that there is a cliff beyond which fees per block quickly fall off. There is more to this power conflict than two sides. Explanations that are trivial and wrong will not help us govern this system towards the future. > (I still don't exactly get what segwit does) It removes the hard block cap and replaces it with a flexbile limit. Inputs counts as one fourth the size of outputs, so the more inputs a block have the bigger it is allowed to get. This factor was chosen because it could be fit in the same backwards compatible transaction style that was utilized to deploy P2SH, where non-compilant nodes only see anyone-can-spend transactions. The reason inputs were made to be cheaper than outputs was to disincentivize UXTO fragmentation. This is an increasingly serious problem which will soon rival the capacity problem. It was recently shown that a large Bitcoin exchange have completely unnecessarily fragmented millions of dollars worth into non-movable dust, since there have been no financial incentive to avoid it in the past. Finally, since the new style transactions breaks compatibility anyway, the opportunity was taken to change the format into one that is not malleable, and avoids the worst case verification times of the old format. The latter is a true scability improvement. One limiting factor for the number of transactions in a block is the latency from verifying the signatures, and it would not be good for the network if people started cheating even more on this.