[Update: an potential adjustment of block rewards - diff]Initial designWe have to go back to review the initial design.
https://bitcointalksearch.org/topic/m.8755719http://cryptomagic.com/misc/images/magi-thread/PoW1-21.pngPredicted consequence is fair distribution, avoiding big mining farms and maintaining the network hash under a certain level (unlike BTC network keep increasing). It will frustrate mining farms (if existed) when network hashrate is so high that the mining rewarding is at the decline side. Mining XMG is less profitable compared to mining others; only belivers keeping mining will receive the most out of it. Same frustration applies to GPU miners if they existed.
According the above image, sorry for big miners, my initial thinking is to exclude the (personal) big hash as much as possible. Since we are initially based on M7, GPU mining is allowed at that time. The situation I am trying to avoid is the GPU farm, that is the reason of using diff-dependent rewards. As to M7M, allowing big hash in network means allowing GPU miners too if they exist. I do believe CPU miners (even with farms) would hate to see GPU farms online, same applied to individual CPU miners. This initial direction must be complied. IssuesNetwork hash keeps increasing; we haven't see the arrival of rewarding peak yet while we already noticed the coming of big miners; Current mining status in two pools:
Suprnova Workers: 244
Nonce-Pool Active Workers: 183
total works: 427
Total network hash is about 105 MHps.
It would be conclusive that individual miners have reached saturation, further increase in hash is mostly due to big miners. We haven't got into the 500 XMG/block stage. Provided the general user base won't grow (most likely), what do we expect once net hash hitting a point where gets 500 XMG/block?
Solutions1) We intend to get 500 XMG/block into the net, but conditional
2) This is the idea; let's say we have three mining groups
Group A - CPU individual miners
Group B - CPU farms
Group C - GPU miners & farms
Though we may expect nonexisting of GPU miners, we can't exclude the possibility (individual). This must be taken care of.
The most importance of this solution is to maintain the #A:
a. Mining hash of #A must remain the same or above during the lifetime of mining.
b. Moving the optimum diff to a point where the net hash equals to the hash of #A; optimum diff is where we get maximum rewards.
http://cryptomagic.com/misc/images/magi-thread/PoW1-22.pngIt will be clear that with the arrival of #B, #A+#B leads to less rewards; Both #A and #B suffers low rewarding. #A is a faith miner and never go way in spite of zero reward or 500 XMG, and what happened to #B. #A is a set of many individual miners; for one who mines with 1 cpu, it won't be much loss if nothing coming out in one or two days (in an extreme case, we can assume at #A+#B rewards is zero). But for #B, since #B is a big farm and running costly, it won't be acceptable for them to mine without any return. We can image the same thing for #A+#B+#C. Obviously, the only way of getting rewards for #B or #C is to lower their hashing power, so that the block rewards are available for mining; it may get them to the same hashing level as #A.
If we can go this way, we will will need to select optimum diff and reward-diff pattern. The standard can rely on the mining history so far.
Any suggestions are appreciated!
Please suggest if this is the way we can go.
Am amazed no one has replied to the DEV yet, a shame to see the thread disintegrate into discussions of how much coins are selling for on exchanges already. You guys need to take a long term view not look for instant payouts !
Joe, I like your thinking which is enforcing the original design such that a high overall hash cripples the payouts. There are only two problems I see;
1) This would probably extend your phase 1 POW past the intended 2 months because it will take a while for the CPU farms/botnets to abandon the mining, after all they are doing nicely right now so they don't have much incentive to give up quickly. This means coin rewards will be disbursed more slowly than intended until the mega-miners abandon it.
2) How do you distinguish between (a) a lot more solo miners joining your network as word spreads, and (b) one high hash-rate farm ?
If you can identify(2) and don't mind the extra mining duration of (1) then go ahead and tighten the block reward for excessive network hash.
I have a few ideas for how to prevent the mega-miners, but unfortunately they would have to be baked into the coin design at the start, your suggestion above is the best I can see at this stage.
cause somebody know there's more coin comeing, they can make profit even at 5100sat
