> He makes it sound like anyone in the "know" understands this. That's quite telling, isn't it? There's this huge debate, but his argument is that it's obviou

> He makes it sound like anyone in the "know" understands this. That's quite telling, isn't it? There's this huge debate, but his argument is that it's obvious? From a 'tradition' or 'legacy' standpoint, the blocksize limit was never meant to be a persistent thing. It was not present in the whitepaper or the initial implementations, and when Satoshi added it, it was explicitly stated that the limit should be raised as needed and as computer hardware and network capacity increased. When Satoshi left, the project was handed to Gavin, who has gone to great lengths to support blocksize increase. Similarly, other early developers like Jeff Garzik and Mike Hearn strongly support on-chain scaling. I think it is quite fair to say that the original intention for the limit was simply an anti-spam measure, and was not intended to limit network usage, for whatever that's worth. Larger blocks require more cost to run a node. It costs more network capacity, more storage, more time to spin up a node, etc. The argument is that these costs will decrease node count to a dangerous degree. More network use could, however, create incentive for more people to run nodes. Clearly there's a trade-off between extremes, so a careful consideration must be made about where the best trade-off occurs. Larger blocks are a liability to miners as they increase the chance of orphaned blocks. This has two primary effects: it places a natural limit to the block size miners are willing to mine, and it creates incentives for miners to have excellent network access to the rest of the network. This later issue can mean mining centralization. This is one of the core arguments against increasing block size, but it's a very complicated issue. One of the primary confounders is ASIC development, which means rapid capital turnover and short investment timelines for mining. As ASIC development reaches state-of-the-art node sizes, and also as the Bitcoin price stabilizes, mining investment looks very different. This is very difficult to predict. Small blocks greatly limit what Bitcoin can do. It places a hard cap on transaction throughput. The formation of a transaction fee market greatly discourages small transactions, making Bitcoin inaccessible to a wide range of people. It makes any address with Bitcoin less than the txn fee effectively unspendable. In the extreme, it limits Bitcoin to the role of digital gold, a speculation vehicle with little use as a medium of transaction. There are two big open questions: * Will larger blocks lead to dangerous centralization? * Will 2nd layer solutions provide effective scaling solutions? In the end, all of this deals with trade-offs, but worse, trade-offs with difficult-to-quantify risks at either extreme. I simply would not trust anyone that is dogmatic about their favored approach. My personal view is that the conservative approach with least risk is to permit moderate blocksize increases while 2nd layer solutions are being developed, and to continue the debate as the network develops. This is, in effect, exactly what the '2X' compromise does.