Re: [ANN][CLAM] CLAMs, Proof-Of-Chain, Proof-Of-Working-Stake, a.k.a. "Clamcoin"
Congratulations to The Great CLAM on reaching the one million blocks milestone!
It's been a while since I replied to posts here, so here goes:
You want tcp/31174 open to the world for p2p connections, and tcp/30174 open locally if you want to use RPC. Don't open the RPC port globally unless you know what you're doing.
I've been updating it semi-regularly here:
https://bitcointalk.org/index.php?topic=623147.msg9772191#msg9772191
I did just update it too.
Not sure if this would considered exploitable, but you could have gained quite an advantage by only staking selectively.
It was exploitable for sure. I don't think it was designed to be exploitable; I think it was an accident. I did point out to the developers (on Jul 23 '14) that it seemed exploitable soon after it was launched (on Jul 12 '14) but it was already live by then:
It turns out that the cheating that happened combined this with manipulating the timestamps on blocks. See how the timestamp on block 163967 is 6 minutes earlier than the that on the previous block:
so it turned out it was even more exploitable than I had previously thought.
To maximize staking, smaller is better until you get to the point that your wallet is using all available CPU checking your outputs. You need to balance that against the future transaction fees required to recombine your outputs when you want to spend them. Note that splitting won't speed up staking, it will only reduce the effect of the 500 block maturation time after an output stakes.
If you only have 30 CLAMs, you will be waiting an average of 30 days to stake no matter how you split them. If you have 3000 CLAMs and split them into 100 piles of 30, each of the piles will be waiting an average 30 days to stake, but one of them will get lucky and stake early. You can expect a few of them to stake each day on average. If instead you split them into 1000 piles of 3, each will be waiting about a year to stake, but you can still expect a few of them to stake each day on average. And each time one of your 3's stakes, it is only making 3+1 CLAMs inactive for 500 blocks, whereas when one of your 30's stakes, it is making 30+1 CLAMs inactive for 500 blocks, which is clearly less efficient from a stake-as-much-as-you-can perspective.
I recently noticed that over 50% of the Just-Dice hot wallet was unavailable for withdrawal. Upon investigation I saw this massive stake transaction. A large deposit had staked before I got a chance to split it up. It split as it staked, but by then it's too late - all the outputs are unavailable until they get 500 confirmations.
It's been a while since I replied to posts here, so here goes:
My router has upnp setup does work with the clam coin client. I saw same thing in the bitcoin core client but then it was fixed at some point. Anyhow, what are the correct udp or tcp ports that need to be opened for home router to accept inbound traffic.
You want tcp/31174 open to the world for p2p connections, and tcp/30174 open locally if you want to use RPC. Don't open the RPC port globally unless you know what you're doing.
Can we get an updated bootstrap?
I've been updating it semi-regularly here:
https://bitcointalk.org/index.php?topic=623147.msg9772191#msg9772191
I did just update it too.
I think the lottery was poorly thought out, but not intentionally exploitable.
Not sure if this would considered exploitable, but you could have gained quite an advantage by only staking selectively.
It was exploitable for sure. I don't think it was designed to be exploitable; I think it was an accident. I did point out to the developers (on Jul 23 '14) that it seemed exploitable soon after it was launched (on Jul 12 '14) but it was already live by then:
Quote
--- Log opened Wed Jul 23 07:30:10 2014
07:30 so I guess we're going to do a hard fork some time soon-ish
07:30 I wonder if we should do some math on the lottery reward before that
07:30 I suspect it's game-able at the moment
07:33 suppose I have 1000 CLAM
07:33 I could split it into 200 piles of 5
07:34 each pile would stake every few days at best, giving me at best 100 stakes per day, or 10 CLAM per day?
07:35 alternatively, I could leave it as a single pile of 1000, and modify the client to only let it stake if it wins the lotto
07:35 after 10 days, it will have a weight of 10,000 and so will be entering the lottery very very often
07:36 I only let it post if it wins a decent amount
07:36 I suspect that after 10 days it will have won more than the 100 CLAMS I could have earned by splitting and staking honestly
07:37 and if not, after 20 days it will be entering the lottery twice as often, since it will be twice as old, twice the weight
07:37 whereas my piles of 5 won't be significantly different than they stated - earning 10 CLAM per day
07:42 the issue is that the frequency at which I can "enter" the lottery goes up linearly
07:42 whereas the cost of withholding is pretty linear
07:30 so I guess we're going to do a hard fork some time soon-ish
07:30 I wonder if we should do some math on the lottery reward before that
07:30 I suspect it's game-able at the moment
07:33 suppose I have 1000 CLAM
07:33 I could split it into 200 piles of 5
07:34 each pile would stake every few days at best, giving me at best 100 stakes per day, or 10 CLAM per day?
07:35 alternatively, I could leave it as a single pile of 1000, and modify the client to only let it stake if it wins the lotto
07:35 after 10 days, it will have a weight of 10,000 and so will be entering the lottery very very often
07:36 I only let it post if it wins a decent amount
07:36 I suspect that after 10 days it will have won more than the 100 CLAMS I could have earned by splitting and staking honestly
07:37 and if not, after 20 days it will be entering the lottery twice as often, since it will be twice as old, twice the weight
07:37 whereas my piles of 5 won't be significantly different than they stated - earning 10 CLAM per day
07:42 the issue is that the frequency at which I can "enter" the lottery goes up linearly
07:42 whereas the cost of withholding is pretty linear
It turns out that the cheating that happened combined this with manipulating the timestamps on blocks. See how the timestamp on block 163967 is 6 minutes earlier than the that on the previous block:
Quote
--- Log opened Thu Oct 23 22:57:51 2014
23:04 SetBestChain: new best=66045c13690ec7a7242c height=163966 trust=1539530299233 blocktrust=21933643 target=195 date=10/21/14 06:27:48
23:04 SetBestChain: new best=ef7b492dcf436a6c5790 height=163967 trust=1539540973066 blocktrust=10673833 target=402 date=10/21/14 06:21:40
23:04 the timestamps are very wrong - but they actually arrived very close together
23:04 the 2nd had nSubsidy: 93442202719.000000
23:04 >900 reward
23:04 SetBestChain: new best=66045c13690ec7a7242c height=163966 trust=1539530299233 blocktrust=21933643 target=195 date=10/21/14 06:27:48
23:04 SetBestChain: new best=ef7b492dcf436a6c5790 height=163967 trust=1539540973066 blocktrust=10673833 target=402 date=10/21/14 06:21:40
23:04 the timestamps are very wrong - but they actually arrived very close together
23:04 the 2nd had nSubsidy: 93442202719.000000
23:04 >900 reward
so it turned out it was even more exploitable than I had previously thought.
Can anyone tell me a good size for each input in desktop wallet to maximize staking? I was told 10-15, I was thinking 25?
Any assistance would be appreciated.
Any assistance would be appreciated.
To maximize staking, smaller is better until you get to the point that your wallet is using all available CPU checking your outputs. You need to balance that against the future transaction fees required to recombine your outputs when you want to spend them. Note that splitting won't speed up staking, it will only reduce the effect of the 500 block maturation time after an output stakes.
these issues haunts me as well dooglus mention about ~30 clam input, but I think This applies only to very long-term,
because first reward you have to wait very long- 40-50 day. (in case 10-15 input , even 3-4 month)
dooglus mention about ~30 clam input, but I think This applies only to very long-term, because first reward you have to wait very long- 40-50 day. (in case 10-15 input , even 3-4 month)
If you only have 30 CLAMs, you will be waiting an average of 30 days to stake no matter how you split them. If you have 3000 CLAMs and split them into 100 piles of 30, each of the piles will be waiting an average 30 days to stake, but one of them will get lucky and stake early. You can expect a few of them to stake each day on average. If instead you split them into 1000 piles of 3, each will be waiting about a year to stake, but you can still expect a few of them to stake each day on average. And each time one of your 3's stakes, it is only making 3+1 CLAMs inactive for 500 blocks, whereas when one of your 30's stakes, it is making 30+1 CLAMs inactive for 500 blocks, which is clearly less efficient from a stake-as-much-as-you-can perspective.
Any of those are fine, it doesn't really matter. The main thing is just not to have one huge output.
I recently noticed that over 50% of the Just-Dice hot wallet was unavailable for withdrawal. Upon investigation I saw this massive stake transaction. A large deposit had staked before I got a chance to split it up. It split as it staked, but by then it's too late - all the outputs are unavailable until they get 500 confirmations.