[ANNOUNCE] SolidCoin - new and improved block chain. Secure from pools

[Zibbo]

Don't know if this issue has been already brought up, but as I understand it, there is a problem with SolidCoin retarget algorithm that will manifest itself when the difficulty reaches the hashrate. Since it considers only the last 240 blocks when retargeting (correct me if I'm wrong), there will be huge fluctuations in the difficulty level. I ran some quick simulations, and you can expect the difficulty level to regularly fluctuate between +-20% where it "should" be, and more than that in some rarer cases. So every once in a while, you'll have situations, where the "real" difficulty according to the "real" hashrate should be 10000, but SolidCoin target is actually 12000 and drops to 8000 with one retarget (33% drop), and all this while the hashrate is static.

It's certainly not a showstopper, but hardly what I would call optimal, and something to consider if you are planning on copying retargetting from SolidCoin for your own project.

If I were to design a retargetting algo, I would find the first block that has a timestamp a certain amount of time in the past (say 2 weeks), see how many blocks were generated in that time, and adjust accordingly and often. Not the other way around as it's done in all current chains.

[spruce]

Seriously though, back this claim up:

It's bullshit is it? Considering we can number the guys who mined the first million Bitcoins on a single hand how is it bullshit when SolidCoin has over 10000 individual owners of coin already?

Your victims investors deserve to know how you came up with that "10000 solidcoin owners" number.

Couldn't one get some idea by counting the number of different accounts or IP addresses on the exchange? It wouldn't be exact but it would give a ballpark figure.

[coblee]

Seriously though, back this claim up:

It's bullshit is it? Considering we can number the guys who mined the first million Bitcoins on a single hand how is it bullshit when SolidCoin has over 10000 individual owners of coin already?

Your victims investors deserve to know how you came up with that "10000 solidcoin owners" number.

Couldn't one get some idea by counting the number of different accounts or IP addresses on the exchange? It wouldn't be exact but it would give a ballpark figure.

There's really know way to know. You can count the number of miners on all the pools or the number of accounts on the exchange. That will just give you an estimate. For example, I've mined solidcoins and sold all of them for bitcoins for a small profit. But I wouldn't call myself an owner of solidcoins since I don't own any anymore.

And I can't imagine how he can claim that he can count the number of miners of the first million bitcoins on a single hand.

[CoinHunter]

Don't know if this issue has been already brought up, but as I understand it, there is a problem with SolidCoin retarget algorithm that will manifest itself when the difficulty reaches the hashrate. Since it considers only the last 240 blocks when retargeting (correct me if I'm wrong), there will be huge fluctuations in the difficulty level. I ran some quick simulations, and you can expect the difficulty level to regularly fluctuate between +-20% where it "should" be, and more than that in some rarer cases. So every once in a while, you'll have situations, where the "real" difficulty according to the "real" hashrate should be 10000, but SolidCoin target is actually 12000 and drops to 8000 with one retarget (33% drop), and all this while the hashrate is static.

LOL, you better rerun your tests again. Could we please get a clue in here.

[FlipPro]

Don't know if this issue has been already brought up, but as I understand it, there is a problem with SolidCoin retarget algorithm that will manifest itself when the difficulty reaches the hashrate. Since it considers only the last 240 blocks when retargeting (correct me if I'm wrong), there will be huge fluctuations in the difficulty level. I ran some quick simulations, and you can expect the difficulty level to regularly fluctuate between +-20% where it "should" be, and more than that in some rarer cases. So every once in a while, you'll have situations, where the "real" difficulty according to the "real" hashrate should be 10000, but SolidCoin target is actually 12000 and drops to 8000 with one retarget (33% drop), and all this while the hashrate is static.

LOL, you better rerun your tests again. Could we please get a clue in here.

Please explain why he's wrong coinhunter.

[CoinHunter]

Please explain why he's wrong coinhunter.

Sure, since the difficulty can only raise in three bands 2%, 6% and 10%, and decrease to whatever level is necessary (upto 4x) where is he getting this 20% figure from?

If we are at target hash rate it will maintain that level within 2%. The other thing to consider is the target hash rate is more like "the maximum" wanted, it's different to BTC where 10 minutes is "Desired". "3 minutes" isn't necessarily desired in SolidCoin, it's the maximum.

[Jimmy2011]

Please explain why he's wrong coinhunter.

Sure, since the difficulty can only raise in three bands 2%, 6% and 10%, and decrease to whatever level is necessary (upto 4x) where is he getting this 20% figure from?

If we are at target hash rate it will maintain that level within 2%. The other thing to consider is the target hash rate is more like "the maximum" wanted, it's different to BTC where 10 minutes is "Desired". "3 minutes" isn't necessarily desired in SolidCoin, it's the maximum.

CoinHunter, I am wondering how did you expect "Maximum coin generation of 18.9 million (18900000) to be reached by ~2022" if you do think that 3 minutes is not necessary for a block and 3 minutes is the maximum in solidcoin.

[CoinHunter]

Please explain why he's wrong coinhunter.

Sure, since the difficulty can only raise in three bands 2%, 6% and 10%, and decrease to whatever level is necessary (upto 4x) where is he getting this 20% figure from?

If we are at target hash rate it will maintain that level within 2%. The other thing to consider is the target hash rate is more like "the maximum" wanted, it's different to BTC where 10 minutes is "Desired". "3 minutes" isn't necessarily desired in SolidCoin, it's the maximum.

CoinHunter, I am wondering how did you expect "Maximum coin generation of 18.9 million (18900000) to be reached by ~2022" if you do think that 3 minutes is not necessary for a block and 3 minutes is the maximum in solidcoin.

Because I can't tell you exactly what it will be. All I can say is at 3 minute blocks it will be roughly 2022. Even bitcoin developers can't reliably tell you their chain lifetime. Bitcoin could be dead in a week. Yet no bitcoin developer realizes it. SolidCoin will still be working, and very fast, as long as someone is hashing. Just check the average block times over it's current life vs any other chain. Then look at the massive amount of mining power variance. Miners have very little influence over the speed and reliability of the SolidCoin network, and this is what end users and businesses want. Reliability. They don't want to have to tell their customers "oh 20% of the miners electricity costs were too high this week so they stopped mining and that's why there was a 12 hour waiting period for your sale."

[FlipPro]

Please explain why he's wrong coinhunter.

Sure, since the difficulty can only raise in three bands 2%, 6% and 10%, and decrease to whatever level is necessary (upto 4x) where is he getting this 20% figure from?

If we are at target hash rate it will maintain that level within 2%. The other thing to consider is the target hash rate is more like "the maximum" wanted, it's different to BTC where 10 minutes is "Desired". "3 minutes" isn't necessarily desired in SolidCoin, it's the maximum.

CoinHunter, I am wondering how did you expect "Maximum coin generation of 18.9 million (18900000) to be reached by ~2022" if you do think that 3 minutes is not necessary for a block and 3 minutes is the maximum in solidcoin.

Because I can't tell you exactly what it will be. All I can say is at 3 minute blocks it will be roughly 2022. Even bitcoin developers can't reliably tell you their chain lifetime. Bitcoin could be dead in a week. Yet no bitcoin developer realizes it. SolidCoin will still be working, and very fast, as long as someone is hashing. Just check the average block times over it's current life vs any other chain. Then look at the massive amount of mining power variance. Miners have very little influence over the speed and reliability of the SolidCoin network, and this is what end users and businesses want. Reliability. They don't want to have to tell their customers "oh 20% of the miners electricity costs were too high this week so they stopped mining and that's why there was a 12 hour waiting period for your sale."

Could you please explain this more or less in layman terms? Why is it that miners dropping from the network have no effect on transactions processing? Why does Bitcoin have this weakness and SC is immune?

[Zibbo]

Please explain why he's wrong coinhunter.

Sure, since the difficulty can only raise in three bands 2%, 6% and 10%, and decrease to whatever level is necessary (upto 4x) where is he getting this 20% figure from?

If we are at target hash rate it will maintain that level within 2%. The other thing to consider is the target hash rate is more like "the maximum" wanted, it's different to BTC where 10 minutes is "Desired". "3 minutes" isn't necessarily desired in SolidCoin, it's the maximum.

Well, my reasoning was that the difficulty is calculated from the time it takes to generate 240 blocks, and that time varies +-20% all the time. However, you are correct that limiting increases to 2%, 6% and 10% stabilizes the difficulty, but it stabilizes it to a level below where you would get a block every 3 minutes. But apparently that's what is wanted?

I still claim, that difficulty drops can be quite big just from the variance, if not as big as in my example (33%), but say 20%. It certainly will not "maintain that level within 2%", or I'm completely misreading the source. However, since those kinds of drops are corrected very fast, it might not be a problem in the end.

[CoinHunter]

Could you please explain this more or less in layman terms? Why is it that miners dropping from the network have no effect on transactions processing? Why does Bitcoin have this weakness and SC is immune?

Ok well currently bitcoin retargets itself every 2016 blocks. At ideal speeds (10minutes) this is 2 weeks of time. If there is a 400% drop off in mining power (ie the current 10Thash becose 2.5Thash) then BTC blocks will now average 40 minutes in length instead of 10 minutes. However as we have also seen with Bitcoin recently, 10 minute ideal blocks don't always occur even when difficulty is evenly matched. We have been forced to put up with close to 90 minutes blocks.

SolidCoin on the other hand isn't trying to target staying at 3 minute blocks. If we are at 3 minute blocks and a 400% drop off in miners occurs  3 minute blocks now take 12 minutes. Very close to Bitcoin's "ideal" block speed. So in a severe mining drop off a user of the SolidCoin network is barely going to notice the fact that blocks are happening 400% slower, as 12 minutes is still relatively fast, especially compared to Bitcoin.

What SolidCoin guarantees is low block variance, that is, the amount of time a single confirmation takes will always be within an acceptable range even under sever conditions. I designed SolidCoin because as a business which uses Bitcoin I know what is currently delivered is completely subpar for the modern world.

Bitcoin is VERY susceptible to a competitor like SolidCoin "diverting" it's mining power as people become more aware of the new and better network. Basically meaning if you have holdings in BTC today, it's not known if you will be able to do anything with those BTC tomorrow. If you want a current example of this, ask anyone who owns NMC, IX or I0Coins that are all based off the original bitcoin code.

[CoinHunter]

Well, my reasoning was that the difficulty is calculated from the time it takes to generate 240 blocks, and that time varies +-20% all the time. However, you are correct that limiting increases to 2%, 6% and 10% stabilizes the difficulty, but it stabilizes it to a level below where you would get a block every 3 minutes. But apparently that's what is wanted?

That is exactly what I wanted. I do not want to "target" 3 minute blocks like 3 minutes is some magic ideal number to hit (like BTC people think 10 is). It's simply the maximum where you want to be to conduct business reliably.

I still claim, that difficulty drops can be quite big just from the variance, if not as big as in my example (33%), but say 20%. It certainly will not "maintain that level within 2%", or I'm completely misreading the source. However, since those kinds of drops are corrected very fast, it might not be a problem in the end.

Difficulty drops of small amounts will exactly match whatever the drop off is. There is only a limit of a 4x dropoff, and this is kept at 4x to make malicious mining pools have to wait before attempting another attack, thereby reducing their profits from such a thing.

If the difficulty drops "too much" so that 3 minute blocks aren't hit, and it less than 3 minutes, that is fine, and part of the algorithm. SolidCoin's api is changing so that businesses can rely not on some "confirmation" magic number, but a base of time and confirmations. Again, as a business owner and experienced programmer I know exactly what is needed for bulletproof handling of real transactions.

[Jimmy2011]

To CoinHunter:

Yes, maybe nobody can tell the real world market, nobody can tell the real lifetime of Bitcoin, ixcoin and solidcoin.

But now the first case is, someone who say he/she proposed a model and he want the expectation time for a block is 10min, 8min, 5min or 3min, and the results show that the real average time is not so perfectly matched with the expectation, and we get the conclusion that the model is wrong or the model was not correctly implemented.

The second case is, CoinHunter claimed the maximum time (could I think it is the average block time for a window of 240 blocks under the same difficulty?) is 3 minutes for solidcoin block, and he also expect that 18.9m coins to be reached by ~2022 according to 3 minutes a block generation rate. So can we say it is a "cheating"? It is irrelevant with any coins.

For my logic, it is not a "cheating". There is nothing behind solidcoin's retargeting algorithm. Nothing can support whatever 2%, 6%, 10% or 12%. Of course, CoinHunter contribute to the community of Bitcoin, who realized the problem and the importance of retargeting. However, now the problem is not well solved.

Furthermore, don't simply compare whatever phenomenological data. No one believe Bitcoin is perfect and Bitcoin is a standard.

[3phase]

 limiting increases to 2%, 6% and 10% stabilizes the difficulty,

It's actually 2.041%, 6.38% and 11.11% as per main.cpp, lines 673-678:

Code:

    if(nActualTimespan<(nTwoPercent*16)) //less than a minute?
            nActualTimespan=(nTwoPercent*45); //pretend it was only 10% faster than desired -- Actually 100/90=1.1111
        else if(nActualTimespan<(nTwoPercent*32)) //less than 2 minutes?
            nActualTimespan=(nTwoPercent*47); //pretend it was only 6% faster than desired -- and so on
        else
            nActualTimespan=(nTwoPercent*49); //pretend it was only 2% faster than desired

Not that it makes a big difference.

I will refrain from commenting on the comments' phrasing.

[Jimmy2011]

Could you please explain this more or less in layman terms? Why is it that miners dropping from the network have no effect on transactions processing? Why does Bitcoin have this weakness and SC is immune?

What SolidCoin guarantees is low block variance, that is, the amount of time a single confirmation takes will always be within an acceptable range even under sever conditions. I designed SolidCoin because as a business which uses Bitcoin I know what is currently delivered is completely subpar for the modern world.

How do you guarantee it and what is your definition for the severe condition for solidcoin or for the business?
Maybe what you guarantee is the fast block generation less than 3 minutes.

[CoinHunter]

 limiting increases to 2%, 6% and 10% stabilizes the difficulty,

It's actually 2.041%, 6.38% and 11.11% as per main.cpp, lines 673-678:

Code:

    if(nActualTimespan<(nTwoPercent*16)) //less than a minute?
            nActualTimespan=(nTwoPercent*45); //pretend it was only 10% faster than desired -- Actually 100/90=1.1111
        else if(nActualTimespan<(nTwoPercent*32)) //less than 2 minutes?
            nActualTimespan=(nTwoPercent*47); //pretend it was only 6% faster than desired -- and so on
        else
            nActualTimespan=(nTwoPercent*49); //pretend it was only 2% faster than desired

Not that it makes a big difference.

I will refrain from commenting on the comments' phrasing.

Yes or 10%, 6% or 2% of current difficulty. I agree the phrasing of comments could be done better but nonetheless it's irrelevant.

[CoinHunter]

How do you guarantee it and what is your definition for the severe condition for solidcoin or for the business?
Maybe what you guarantee is the fast block generation less than 3 minutes.

A severe condition is 75% of the hashing power dropping off very suddenly (<8 hours). SolidCoin can handle this with almost no problem, it's not ideal to be in this condition and it's assumed to be a short term phase as SolidCoin defends itself against certain scenarios. If over 75% drops off or more within a few hours (something which is unlikely to happen given global spread of mining power and time delays in moving miners) then SolidCoin still manages to correct itself within a few days, compared to the months it would take Bitcoin to correct itself, if it ever would.

SolidCoin is pretty much unbeatable when it comes to the  issues of miners moving themselves about as profitability dictates. SolidCoin can defend itself against a SolidCoin clone. We can already see a weeks worth of mining variability and the ability for SolidCoin to maintain impressive speeds.

Bitcoin is very vulnerable to something like SolidCoin which not only improves the protocol, but offers features which will drive business and users to it's chain. It is vulnerable to numerous attacks, whether malicious in intent or not, and it's very likely that we will see as SolidCoin gains traction that bitcoin cannot sustain itself. ie NameCoin, I0coin, Ixcoin, all based off bitcoin, all known failures of protocol.

How many bitcoin clones do we need to see fail before people understand there is something inherently wrong there? I give people ample warning of the situation so that if what I predict does come to pass no one can blame me for them losing their investment.

[Tasty Champa]

noticing the hash rate has dropped a significant amount in the last 36 hours, why hasn't the difficulty dropped with it? It just keeps going up very minutely, but still going up.

[CoinHunter]

noticing the hash rate has dropped a significant amount in the last 36 hours, why hasn't the difficulty dropped with it? It just keeps going up very minutely, but still going up.

Because there is still a lot of power for the current difficulty. ie We haven't exceeded 3 minute blocks.

[coined]

is there a limit on how much the difficulty cant drop in a single adjustment? sorry if it's been posted already.