Re: New paper: Accelerating Bitcoin's Trasaction Processing
I'm going to have to sit down and read the paper more carefully to have a solid opinion on it, but my meta-opinion is that I think it's great to see people taking scalability and blocksize seriously on an academic level. We've got some incredibly naive ideas floating around the Bitcoin space right now - like the idea that just removing the blocksize limit entirely will work out fine - and we need research into scalability solutions that considers incentives and the resulting security carefully. That kind of reasoning tends to involve a lot of math, rather than platitudes about "market forces"
Speaking of, while the paper presents a solution preserving security guarantees, a quick skim of it doesn't seem to indicate they take into account the incentives around block propagation. If you wind up with a situation well large, centralized, mining pools earn more money as a part of this high-speed propagation game, even though in theory all the work being done contributes towards 51% security, the overall result may be a serious negative due to the incentives towards centralization. Lately I've done some work (pdf) on that topic; it's a very important crypto-currency design consideration that I'd like to see other people analyzing as well.
Speaking of, while the paper presents a solution preserving security guarantees, a quick skim of it doesn't seem to indicate they take into account the incentives around block propagation. If you wind up with a situation well large, centralized, mining pools earn more money as a part of this high-speed propagation game, even though in theory all the work being done contributes towards 51% security, the overall result may be a serious negative due to the incentives towards centralization. Lately I've done some work (pdf) on that topic; it's a very important crypto-currency design consideration that I'd like to see other people analyzing as well.
I'm not a developer, so excuse my ignorance. But wouldn't 1 second blocks essentially make solo mining much more plausible? It seems at that rate even lowly cpu miners could pull of solo mining. Might not be profitable per se, but the chances of at least mining a block solo would be greatly increased.
Yeah, but if the result of the solution is such that solo-mining earns you only 50% of the revenue that hashing in a large, 25% hashing power pool got you, you're incentive would be to mine in the pool instead leading to centralization. The problem is we know that pools already have higher revenue per unit hashing power; my suspicion is the solution proposed by the paper makes that problem even worse. But I've got a flight to catch in an hour or two and will get a chance to think about it all more carefully later.
