Quote from: smooth on September 11, 2015, 01:23:12 AM

Quote from: TPTB_need_war on September 11, 2015, 01:19:46 AM

The fundamental problem is the mining is done for profit. For as long as that is the case, ASIC farms (or Larry Summers' 21 Inc. economies-of-scale) and subsidized, industrial/government/utility scale electricity will rule.

Also due to bandwidth issues and that every full mining node has to validate every transaction, scaling transaction volume will force centralization.

Maybe correct, but depends on assumptions about future costs and technology which are hard to make soundly with certainty. Bandwidth, etc. may become too cheap to matter. The new iPhone has 300 megabit (peak; ideal) wireless capability. Comcast is promising to upgrade their entire network to 1-10 gigabit within three years without needing fibre to the home.

In any case, that's not the same delegation argument r0ach was making. Pool mining and centralized mining is not the same thing.

You know I am all for not conflating separate concerns, so I concede.

But again the 51% attack (if you believe it is realistic the G7 or G20 and other government unions can regulate that much mining) renders the distinction almost entirely effectively invalid.

But I will repeat the point I've made to you this year (recently) that as you raise the bandwidth requirements, you reduce the diversity of ways that mining can be connected to the network which thus increases the State's ability to regulate you (i.e. effectively centralize mining w.r.t. to censorship resistance and KYC) and also decrease network fault tolerance issues such as network slowdowns w.r.t. to orphan rate which then impacts double spends, etc.. You can't win the argument that big O (or theta O) scaled higher bandwidth requirements beats a design with lower big O bandwidth requirements scaling, all other factors being equal. I know you did not argue higher bandwidth is superior rather you seem to be arguing it might be a mitigated issue. And I am arguing that the failure potentials are significant and that we need to improve on all aspects to maximize probability of retaining some attributes of resilience such as censorship resistance.

Also there is the issue of network fragmentation, which none of the consensus network designs deal with at all. They just collapse into chaos of double-spending.

Also on that chart you posted upthread about electric consumption and hashrate declining over time due to block reward halving, that excludes the expectation that transaction rates and fees will be increasing and assuming the market is competitive then hashrate should be higher than it would be due to value of block rewards alone.

r0ach's overall point is we are headed for centralized mining any way.

On that thesis he is correct, unless someone presents a superior consensus network design.