Quote from: cswolfe on April 13, 2014, 05:20:36 PM

True CGA developers:

I suspect the ASIC resistance might prove somewhat true. I know guys with the 250 meg titans coming in. To my knowledge, you cannot "split" the hashing. Correct me if I'm wrong, but especially when mining firepower dedicated to CGA is low, anybody dropping 250 meg onto CGA would pretty much burn themselves (granted, along with everybody else) with the payout system.

Mea culpa, but too busy working on rigs to truly read more today on CGA.

Things I don't understand.

1) Why even when mining levels remain consistent, difficulty tends to jump all over the board. You could see a < 1 difficutly, then a 6, when hash power dedicated has not changed one iota.

2) when some block payouts, out of nowhere, are odd amounts, even on sub 3 difficulty, like ".15," or "2," not .33. No time to look up the payouts charts to get exact numbers, but I will, if you don't understand what I'm saying.

Excellent questions:

True, at first if a few titans were to jump into the CGA network it would pretty much "burn out" their rigs for almost no profit, but say that they jumped in not to make a 'quick profit', by dumping the mined coins immediately after mining and move on, but sat in the network for the "long haul", this would raise the difficulty significantly thus making less reward per block and making the CGA at hand more precious to the miners. We would probably see a boom in price per CGA, thus raising its profitability to mine. This has been proven to be the case, look at DOGE for example, they started very low in value (10-20 satoshi I believe) then all the miners started to jump into the network making the mining profitability of DOGE much less, which made the DOGE on hand more valuable, then the 'boom' happened and the price shot up to 300+ satoshi, thus making it one of the most profitable coins once again. I believe this is true for all coins and to run away from progress (ie. ASIC) is running away from your investment.

1) Difficulty isn't exactly directly effected by just the networkhash, the difficulty is determined by the time it takes to solve a number of blocks (in the case of CGA 2 blocks), the software will determine how long it has been between blocks and adjusts the difficulty to try and get close to the 'target time' for block solving (for CGA this is 40 sec). So if someone were to get lucky and solve a block 5 sec after the last block solved then the difficulty will spike, if it takes longer then 40 sec between blocks then the difficulty will drop, so it is more or less 'random' what the difficulty will be but the overall average is effected by the network hash rate. KGW and others are used to "smooth out" the difficulty jumps, but we have found that these have serious exploits and we feel we want to wait until a well tested version of the code is out before we make any changes.

2) Payouts (1/diff and .33333333) are effected by the previous block, not the current one. So if you see a block in the explorer that is worth .123... and the difficulty for that block is < 3 then look at the block before that one and you will see the difficulty that decided the reward for the current block. As for the blocks that claim higher then .33333333, most of the time you are seeing transactions within that block, sometimes (if it is just over .33333333) it is because of a transaction fee that was added to that block. The one thing I have yet to wrap my head around is the blocks that are worth 0, it is probably an integer overflow problem with the difficulty but all the clients seem to agree on the value of the block so it seem not to be a major problem. I have a fix in mind but it would require a hard fork and want to wait a bit before we do such a thing. Besides, it just adds to the 'elusive' factor of this coin so I am not complaining Tongue

.