the download synced the blockchain up to 1180002
The longest blockchain is ~2.2M+ blocks (see below).
Community.
In October '15 my host got crashed and I had very big problems in real life so when I got some time to look into the thing the network was already down for several months. One of the latest supporters repeatedly asked me in PMs for almost a year to bring up the network back. He kindly sent me his blockchain file that was a bit longer than mine. As I stated already it has 2.2M+ blocks with total supply ~1.2bln coins. I will bring up a block explorer with that chain shortly, so you could explore it.
At the same time, I found many discussions on BCT regarding possible vulnerabilities of PoS and hybrid coins, please read these two threads:
https://bitcointalksearch.org/topic/nothing-at-stake-in-proof-of-stake-1709776https://bitcointalksearch.org/topic/security-analysis-of-powpos-hybrids-with-low-pow-reward-551861Although I am skeptical about "nothing-at-stake" issue, I see a real problem with low PoW rewards: the emission schedule made it really not profitable to mine Graincoin anymore. There is a real risk of double-spend possibility with low PoW difficulty (i.e. one produces a row of PoW blocks with transaction in a first block and then publishes a PoS fork that overrides the original transaction)
Taking to account that the protocol targets to produce at most four PoW blocks each twelve blocks we still have to mine ~30k blocks with average reward of 6 coins a block, then 65k blocks (259k/4) with average reward of 3 coins and 65k blocks with average reward 1.5 coins. Then the reward would be 1 coin a block forever. Roughly speaking, the mining would produce approximately 500k coins in the following 7 months counting from now and then it would generate 21900 coins a month.
So, I am for switching to pure PoS now (looking at the BlackCoin example that behaves just fine for more than 2 years being pure PoS).
There is a couple of additional topics I would like bring to discussion. The first one is target time. Presently it is 30 seconds a block and produces ~0.5Gb of blocks data a year and ~1.2Gb of indexing database a year. IMO this makes no real advantage: my experience shows that most exchanges and payment processors picks such a number of required deposit(payment) confirmations that your account get credited (the payment get confirmed) roughly one hour after the transaction is mined. Thus, I propose to increase the target time to 2.5 minutes - this would lower disk and memory usage.
The second thing is the existing blockchain, i.e. history: most blocks starting from July '14 are empty (have no transactions apart from coinbase and coinstake) and only consume RAM. So I propose to swap the chain. To make the swap fair and verifiable and to bootstrap the pure PoS network I am going to proceed as follows:
The minimum coinage to stake is 10 days. That means that any TXO is eligible to stake after it get 10 days old. In case of a fresh chain the network has to produce somehow 5760 blocks (if the target time is 2.5 minutes) before the first stake block could be minted. Currently the chain has approximately 1M unspent transaction outputs. These 5760 blocks are enough to evenly distribute 1M unspent transaction outputs in coinbase transactions thus reproducing the existing total supply belonging to original owners. To allow anyone to (manually) verify that any particular TXO in the new chain exists in the original chain and the amount matches, the scriptPubKey of each TXOs in these first 5760 PoW blocks would be composed this way:
OP_PUSHDATA OP_PUSHDATA OP_DROP2 DUP HASH160 EQUALVERIFY CHECKSIG
here
is an id of transaction in the original chain, is an output number in that tx and is a destination address from the original tx.
All these blocks will be signed with a keypair the public part of which will be hardcoded into CheckBlock function. Also, CheckBlock function would check at block 5760 that total supply is equal to the one of original chain and will not accept PoW blocks after height 5760. These 5760 blocks will be PoW-mined myself (with no extra reward) and then the PoS should drive the network.
Please share your thoughts, objections and ideas.
bee7
