{"type":"rich","version":"1.0","author_name":"npub1m230cem2yh3mtdzkg32qhj73uytgkyg5ylxsu083n3tpjnajxx4qqa2np2","author_url":"https://nostr.ae/npub1m230cem2yh3mtdzkg32qhj73uytgkyg5ylxsu083n3tpjnajxx4qqa2np2","provider_name":"njump","provider_url":"https://nostr.ae","html":"📅 Original date posted:2015-06-23\n📝 Original message:On Tue, Jun 23, 2015 at 04:12:17PM -0400, Gavin Andresen wrote:\n\u003e \u003e In particular, if bandwidth scaling doesn't go according to your plan,\n\u003e \u003e e.g. the exponential exponent is too large, perhaps due to technological\n\u003e \u003e growth not keeping pace, or the political realities of actual bandwidth\n\u003e \u003e deployment making theoretical technological growth irrelevant, what\n\u003e \u003e mechanism will prevent centralization? (if any)\n\u003e \n\u003e \n\u003e Simulations show that:\n\u003e \n\u003e Latency/bandwidth matter for miners.  Low latency, high bandwidth is\n\u003e better. However, miners with bad connectivity can simply create smaller\n\u003e blocks...\n\nPieter Wuille showed with simulations that miners with bad connectivity\nare negatively affected by other miners creating larger blocks.\nSimilarly I showed that with equation-based analysis. I've seen no\nresponse to either argument, and it's a centralization pressure.\n\nNote how propagation times are important enough to miners that they\nalready mine on top of unverified headers from other miners to increase\nprofitability, a grave threat to the security of the Bitcoin network.\n\n\u003e ... until transaction fees become significant.  But by the time that\n\u003e happens, protocol optimizations of block propagation will make the block\n\u003e size an insignificant term in the \"how profitable is it to mine in THIS\n\u003e particular place on the Internet / part of the world\" equation.\n\nThese block propagation improvements are both already implemented (Matt\nCorallo's relay network, p2pool) and require co-operation.\n\nFor instance, notice the recent full-RBF debate where Coinbase said\nthey'd consider getting contracts directly with miners to get\ntransactions they desired mined even when they otherwise would not be\ndue to double-spends. This is one of many scenarios where block\npropagation improvements fail. Thus for a safety engineering\nanalysis we need to talk about worst-case scenarios.\n\nEqually, I don't see any analysis from anyone of that % of non-optimized\ntransactions need to fail for what kind of centralizing pressure.\n\n\nIn any case, this ponts to the need for your proposal to explictly talk\nabout what kind of resources are needed by miners for what kind of\nprofitability, including the case where other miners are sabotaging\ntheir profitability.\n\n-- \n'peter'[:-1]@petertodd.org\n000000000000000008c0be16e152f86ab3a271a13c3f41c56228d72990abf7bd\n-------------- next part --------------\nA non-text attachment was scrubbed...\nName: signature.asc\nType: application/pgp-signature\nSize: 650 bytes\nDesc: Digital signature\nURL: \u003chttp://lists.linuxfoundation.org/pipermail/bitcoin-dev/attachments/20150623/97ccb0af/attachment.sig\u003e"}