Quote from: monsterer on January 10, 2016, 01:58:23 PM

Here is another question, sparked by TPTB (and partially aimed at Iota, but not exclusively so):

* If there is no reward at all to performing POW (neither block reward nor transaction fees), what incentive is there for consensus to diverge vs converge?

One incentive for it to converge is the desire to have your transaction confirmed in a timely manner - this encourages transaction submitters to chose the chain with the most work and submit their transaction there.

Why would it diverge?

Thanks for raising that issue for clarification. Without a block reward, there is nothing at risk in terms of not being on one longest chain, other than that your transaction isn't truly confirmed if there is no convergence. But remember I was proposing there will be a block reward, it just won't be profitable. So the miner will still risk losing the compensation of the block reward same as for Satoshi's design if they don't follow the LCR. So this is a major difference in my design and Iota. Iota has no blocks so it can't offer a reward (which also means the coin supply will shrink to 0 asymptotically in Iota).

Without a block reward, you are correct that the only incentive to be on a longest chain is so your transaction is confirmed unambiguously. But the problem is that multiple chains can have conflicting double-spends ex post facto and thus merging partitions becomes impossible in come scenarios without reversing transactions. I don't think a DAG (Iota) can force convergence in all scenarios. I would need to spend more time writing down attack scenarios and motivations. I assume I would do that some months later if I had already invested developing my design. I don't want to go off on that tangent right now.