Re: [ANNOUNCE] paccoin a new SHA256D coin is being introduced
I totally posted badcode I just noticed, typing too fast, sorry:
Wanted to fix before anyone tried to discredit for b/s reasons.
Code:
if (pindexNew->bnChainTrust > bnBestChainTrust && pindexNew->nHeight > nBestHeight)
Wanted to fix before anyone tried to discredit for b/s reasons.
I'm not familiar with the crypto tech, does this mean that who has the highest pile of coins gets the staking power? IE., large holder gain pos coins and such coins over write those pos of smaller coins owners?
Very likely, a large holder will get the next 1-3 block rewards corresponding to the large chunks of ( and higher-weighted) coins assigned for the POS, immediately after he/she starts staking (minting), but once rewarded, the large portions of coins will no longer be eligible for staking until next time of days or weeks of holding periord for that portion of coins to be used again. There should be less input for staking as time passed, to the point that if you (" with the other wallet") just start staking at that point, you might have a higher staking weight to get the next block.
It is a matter of chance; therefore, your chance of getting rewards immediately starting staking is much, much lower, but even if both start at the same time, there might be a time say one hour later, the one having only, say 1/10 of total coin balance of the other wallet client, would have almost the same amount of coin weight to get the next reward which you are supposed to get anyway.
It might takes longer if there are many, many higher coin-weight wallet starting the staking during the time you are in.
As you can see, the one with extremely small amount of coin most not get a reward for quite a few months but by the time, you would have much higher amount of holding-time and staking power.
If I understand correctly, that people with big pile will have lower coins/wattage (which is fair per POS), but is the main concern like this?
1. Smaller miner starts at height 1000, he's digging away
2. Big miner finds the block at 1000 and 1001
3. Smaller miner is now mining 1001, which has been mined by the big miner
4. So the smaller miner is always doing catch up
My apologies for not being a smart person...
I think its worse than this. I will start with I'm not a Dev, but what they are talking about appears to be visible in the explorer.
I think it appears that large Coin holders are overwriting MANY blocks. Not allowing many mined/staked coins to count. I once noticed my stake sometimes appears then disappears with a "generated but not accepted" flag. I assume this may have been what was happening.
I think you can see it in the Explorer..
http://paccoinexplorer.com/
Just look at the blocks in order. At some point they get a posted block that is earlier than the previous. If the block with the older Block Number is Confirmed to the chain the chain loses ALL blocks that were after that block and already posted to the chain.
Example, I just scrolled down the last 300 blocks.. Check these out..
2625860 came before 2625810. Because the 2625810 was posted by a large coin holder it RESET the chain to that point. Consequently dropping ALL blocks from the previous 2625810-2625860.
Here are the recorded blocks:
This one is the last of the orig 2625810-2625860 chain (and now block is not confirmed)
http://paccoinexplorer.com/block/000000000028c00c9e96d42aaaaad7396054f382ca29f2051fd912bc738e09be
Then 2625810 comes along and RESETS the chain back 50 BLOCKS...
http://paccoinexplorer.com/block/e6122643109608381a8486d0fac2af9b0fbeb8212020f411dddd1a009bdedede
Seems those blocks are then lost..
Like I said, I'm no dev. But this does appear to be a pretty big issue.
A chain that changes isn't really a chain?
TL;DR: This wallet has some major flaws and needs a hark fork to correct them.
^ This is exactly right and the problem I was explaining -- the chain is constantly forking which is a huge problem. The forks are caused by a flaw in the SetBestChain logic. PoS blocks of a lower height (aka. blocks that should be less trustworthy and rejected) are being accepted as valid and erasing valid PoW, PoS, and transactions. The net effect is that work is being lost / wasted while older PoS account are benefiting from a reduced coin supply and the ability to wipe out, in some cases, hours or work by other miners. Worse, transactions are being lost into the void. If you scan the block chain manually you can find plenty of transactions that were sent on one chain, then overwritten, and are currently sitting in the chain with 0 confirms despite being valid transactions because they were eaten on a fork. I've seen some for well over a million PAC get lost to the void -- which is a MASSIVE problem.
Compounding the issue is another flaw in the wallet which won't let you recoup those TX I noticed (I've been experimenting for a few days with it). Anyway, I'll post a change set to the two repositories on github and if it's not accepted I'll post a hard fork here for anyone who doesn't want their transactions, stake, and work subject to being overwritten and lost.
Hopefully I can get that submitted today, trying to wrap up some other work first.