No, I was interested in what you had to say about it. I do have a couple kimono things around, so I might have been thinking about that, but made a mistake and put them in the text instead of kimoto, which is indeed the proper spelling for this 'kimoto gravity well' item.
I wonder if this would be adjustable enough to allow ordinary users to "dial it down" so that, in accordance with the capacity of their machines, nearly anyone could mine on some similar system such as this kimoto gravity well? When you are using BCN, the amount of resource it needs is so low you can mine it using a laptop, but there is essentially no variation (nor user direction that would indicate some adjustment needed) for adaptive-ness in such a system. It just does it as far as I can tell. But in bitcoin (BTC) it is another matter entirely, and would be very interesting to see if users with limited equipment (say, those who only have laptops or towers) could make certain adjustments and run it as we used to some years back (2010 - 2011). That would spawn a whole lot of new interest to be sure, although I have about zero idea as to how technically it might work, I certainly found the visualization (graph) here interesting:
https://bitcoin.stackexchange.com/questions/21730/how-does-the-kimoto-gravity-well-regulate-difficultyThen imagine at the same time all this is happening, which introduces a lot of new diversity to the system, which I suspect is good, in come various approaches to sidechains. Regardless what anyone thinks of them, in they come, and one can imagine that there will be a system in which there will eventually be zerocash, or something like it, with its own consensus system being managed potentially on a sidechain independently from bitcoin, but still being able to be exchanged for bitcoin if someone were to sit at a table and say, "hey, let's exchange this zerocash token for some bitcoin" or something to that effect. Zerocash would have its own way of ensuring no double spends. Not to say sidechains are necessary for this to work, just turning over ideas in my brain randomly, and considering some questions to e-mail the zerocash developers, who have not (yet) released the zerocash code.
I like your thought patterns here.
It occurs to me that the amount of money individuals spend buying computer and network resources is huge (in total), the idle potential is huge, and the 'distribution' is huge. For a LOT of reasons tapping into this support base as much as possible seems like a good idea. To my dismay it seems that Bitcoin is pressured in the opposite directions (which is part of the reason I'm so down on the Bitcoin Foundation.)
One idea that I re-remembered is to have multiple proof-of-work algorithms simultaneously. One block may be solved by sha256, the next by scrypt, the next by some variation of scrypt, etc. If there is a dearth of blocks solved by one type of algo, a the reward for that type increases. The goal, of course, would be retard the propensity for specialized ASIC farms to develop, though it doesn't necessarily mean that mega-farms won't or that mining would not switch over to more general corporate cloud infrastructure.
When I first read the whitepaper I came away with the misunderstanding that simply playing a role in the transmission of transactions was rewarded (being a full node who doesn't necessarily mine.) Had I not, I probably would have been much less interested in the solution...so I'm glad I misunderstood! Anyway, I consider it critical to the long term success. Here is where I see a great value in rewarding diversity. So, say, you are operating in Iran, or are on a radio link, relaying transactions nets a much bigger reward than if you are one of 100,000 people running a node in Amazon's AWS.
Since I am mostly interested in 'reserve currency' duty, the exact block frequency doesn't matter so much. High and variable latency systems are vastly more difficult to attack partly because the systems they support are built anticipating this. Bitcoin excels here already.
I'd like to see rewards modulated with a varying coinbase as a factor. So, to play on a previous example, if you are a mining node behind a radio link, you might get a much bigger coinbase reward than a block mined in mega-mining datacenter. For the benefit of the economy the inflation rate needs to be roughly anticipatable, but it does not need to be precise. This could be achieved through an interplay between coinbase rewards and block frequency.
A massive problem to be solved is that if one has a supercomputer it is relatively easy to pretend you have a laptop. The reverse is not true. It is also fairly straightforward to emulate a variety of network connection types or to pretend to be located in a jurisdiction you are not. To solve that problem (partially) I would imagine a web of interconnected nodes which cross-check and verify one another. A system must anticipate that everyone will always cheat if they can.
Getting back to the Kimoto thing, the glancing I've done at it seems to indicate that it is nothing more than a sort of a dampening and weighting algorithm. And a fairly simple one (simple is good!) Such constructs would be abundantly necessary in the solution to some of the problems I mention above, and it is not surprising that it popped out as a solution to the problem you seem to have recognized. Again though, you gotta consider the problem I mention in the paragraph above.
My 'paracoin' idea was almost exclusively for the purposes of having some way to safely play around with some of these ideas. Sidechains are a vastly better way to solve the same problem. I implemented (or more accurately, hypothesized) the 'two way peg' in a much more clumsy way. (edit: And in a way which does nothing to help and/or support Bitcoin where the exact opposite is true of sidechains.)
edit: minor/syntax
ok 