| Thanks for the article. But I don't see where it says that Bitcoin can't scale 20x without decentralization problems. I see that "the block size should not exceed 4MB for X=90%; and 38MB for X=50%", but I don't think that 50% reduction of number of nodes would be a serious decentralization problem (we've seen a similar reduction already due to the implementation of SPV clients). Mining centralization could be a bigger problem, but with headers-first mining the block size wouldn't contribute to it. Also the authors of the article assume the use of Bitcoin’s current peer-to-peer overlay network, that can be drastically optimized. I've chosen 20x because Gavin had conducted 20Mb tests more than a year ago, before many optimizations took place. There is nothing bad in the low fees. Low fees is the marketing strategy to boost the usage of Bitcoin and create the large enough ecosystem. Miners are being subsidised by bitcoin holders for a reason. Larger the blocks while preserving a reasonable degree of decentralization -> lower the fees per transaction -> more people use Bitcoin -> more valuable the bitcoin ecosystem is -> higher the price of bitcoin -> mining is more profitable -> stable and secure infrastructure for the new global economy. |