KrLos
|
|
April 03, 2013, 10:01:02 PM |
|
I think.
1. No CPU friendly, for keep botnets away. 2. Only GPU Friendly, not FPGA nor ASIC.
or a duble parameter.
GPU + CPU = SHA(xxx)+sCrypt(x) ? so this way only "personal pc's" can mine. (and tons of good pc's)
well, maybe i'm telling a dumbass thing but might work.
|
|
|
|
nwbitcoin
|
|
April 03, 2013, 10:16:15 PM |
|
That's totally perfect, some people want to hold their coins, some want to spend. people can hord it if they want, but the algorithm of this altcoin insures, that in the long run at least 1-2% per year of the coins is newly generated and distributed to all participants in an as fair and as decentralized way as possible.
So people can buy goods with it, hord it, make a loan on it, or just dump it to zero. That's real freedom.
OK, we are sort of on the same wavelength, if you want to share to all participants, how about a tax on miner fees of a percentage e.g. 1% that gets paid into all wallets once a year? At least the rules are easy to implement and understand, and it would explain where the money is coming from. Working the practical details, it would pay a figure into every individual wallet which had been used in the previous year. Obviously, some people would game it so they would use a new wallet for every transaction, but that's the nature of socialism - it encourages people to cheat the system for their own benefit! But, in all seriousness, using that method, you would get a fairer method of wealth distribution in an algorithm. Feedback?
|
*Image Removed* I use Localbitcoins to sell bitcoins for GBP by bank transfer!
|
|
|
sangaman
|
|
April 03, 2013, 10:21:35 PM |
|
I'm a software developer as well, but haven't worked directly with the nitty gritty of cryptocoins.
I've thought about ways that you could distribute coins fairly to users at the beginning, and I don't think a X coins per person plan is feasible at all. My best guess now is that a POW system that favors average consumer computers is the best way to distribute coins, with rewards that decrease slowly and steadily over time.
Is there anyway to choose an unspent input in the blockchain randomly and transparently? I've wondered if it would be possible to do this and weight unspent inputs by coin-age and choose one randomly using a previously agreed upon value as a seed. If an unspent input gets chosen, the owner of that input has the ability to create a new block and sign it with the private key that received the chosen input (but only if the owner is online and connected to the network) and receive a reward. If the owner of the input is not online, a new unspent input is chosen.
This system (or something like it) would be a proof-of-stake system where no user can unilaterally increase their chances of creating a POS block or could burn up a lot of coin age in a short period of time to make several blocks in a row. The chance for an online user to make the next block would always be equal to [user's total coin-age] / [total coin-age that is online].
I haven't gotten the time yet to really look through the bitcoin or ppcoin code to see how nodes verify that a blockchain satisfies the protocol or how easily they can read from unspent inputs. The above might not be possible at all but it was something that came to mind, if this makes sense to anyone please share your thoughts.
Yes, a PRNG would work, however this encourages people to have unspent outputs which will bloat the blockchain. Also, should ideally allow it to generate a block in the next 24 hrs - so people who have normal computers instead of a 24x7 box would not be at an disadvantage. That wouldn't be the case if unspent outputs were appropriately weighted according to coin age. Having 100 inputs of .01 would be no more likely to generate a POS block than a single input of 1.00. I don't see how offline block generation would be possible though, since there needs to be an active node that signs the block with the private key.
|
|
|
|
chriswen
|
|
April 03, 2013, 10:27:33 PM |
|
I'm a software developer as well, but haven't worked directly with the nitty gritty of cryptocoins.
I've thought about ways that you could distribute coins fairly to users at the beginning, and I don't think a X coins per person plan is feasible at all. My best guess now is that a POW system that favors average consumer computers is the best way to distribute coins, with rewards that decrease slowly and steadily over time.
Is there anyway to choose an unspent input in the blockchain randomly and transparently? I've wondered if it would be possible to do this and weight unspent inputs by coin-age and choose one randomly using a previously agreed upon value as a seed. If an unspent input gets chosen, the owner of that input has the ability to create a new block and sign it with the private key that received the chosen input (but only if the owner is online and connected to the network) and receive a reward. If the owner of the input is not online, a new unspent input is chosen.
This system (or something like it) would be a proof-of-stake system where no user can unilaterally increase their chances of creating a POS block or could burn up a lot of coin age in a short period of time to make several blocks in a row. The chance for an online user to make the next block would always be equal to [user's total coin-age] / [total coin-age that is online].
I haven't gotten the time yet to really look through the bitcoin or ppcoin code to see how nodes verify that a blockchain satisfies the protocol or how easily they can read from unspent inputs. The above might not be possible at all but it was something that came to mind, if this makes sense to anyone please share your thoughts.
Yes, a PRNG would work, however this encourages people to have unspent outputs which will bloat the blockchain. Also, should ideally allow it to generate a block in the next 24 hrs - so people who have normal computers instead of a 24x7 box would not be at an disadvantage. That wouldn't be the case if unspent outputs were appropriately weighted according to coin age. Having 100 inputs of .01 would be no more likely to generate a POS block than a single input of 1.00. I don't see how offline block generation would be possible though, since there needs to be an active node that signs the block with the private key. He said offline wouldn't work. If they were offline the lottery rolls again. Maybe something like timekoin system could be implemented?
|
|
|
|
chriswen
|
|
April 03, 2013, 10:28:26 PM |
|
That's totally perfect, some people want to hold their coins, some want to spend. people can hord it if they want, but the algorithm of this altcoin insures, that in the long run at least 1-2% per year of the coins is newly generated and distributed to all participants in an as fair and as decentralized way as possible.
So people can buy goods with it, hord it, make a loan on it, or just dump it to zero. That's real freedom.
OK, we are sort of on the same wavelength, if you want to share to all participants, how about a tax on miner fees of a percentage e.g. 1% that gets paid into all wallets once a year? At least the rules are easy to implement and understand, and it would explain where the money is coming from. Working the practical details, it would pay a figure into every individual wallet which had been used in the previous year. Obviously, some people would game it so they would use a new wallet for every transaction, but that's the nature of socialism - it encourages people to cheat the system for their own benefit! But, in all seriousness, using that method, you would get a fairer method of wealth distribution in an algorithm. Feedback? If 1% is given to everyone. Think about it like ripple. Destroying it would give the same increase in buying power to everyone. And then that's just demurrage.
|
|
|
|
BBQKorv
|
|
April 04, 2013, 12:01:00 PM |
|
I had some ideas about a true CPU friendly- GPU hardened coin that would adjust as both hash rate and GPU technology evolved.
It involves using a radically different type of difficulty adjustment system that I do not believe has been seen before.
CPU friendly, so the large botnets can easily attack it, right? If it's requiring the very latest CPU extensions those botnets with usually lowspec hardware and outdated CPUs won't do much. Pure CPU coin would be nice, and being highly RAM intensive could be nice too. Of course this brings the question if there is enough people with the latest high-end CPUs and plenty of RAM (8GB or 16GB and up) to support that kind of coin.
|
|
|
|
Vladimir
|
|
April 04, 2013, 12:04:45 PM |
|
I think.
1. No CPU friendly, for keep botnets away. 2. Only GPU Friendly, not FPGA nor ASIC.
or a duble parameter.
GPU + CPU = SHA(xxx)+sCrypt(x) ? so this way only "personal pc's" can mine. (and tons of good pc's)
well, maybe i'm telling a dumbass thing but might work.
Yep good start. Something like Sha256(nonce+Sha512(nonce+Tiger(nonce+Whirpool(nonce+GOST(nonce+scrypt(nonce+bcrypt(nonce+data)))))) perhaps? or some proof of work based on HDD space as one guy proposed elsewhere
|
-
|
|
|
Mageant
Legendary
Offline
Activity: 1145
Merit: 1001
|
|
April 04, 2013, 12:08:36 PM |
|
|
cjgames.com
|
|
|
Anon136
Legendary
Offline
Activity: 1722
Merit: 1217
|
|
April 04, 2013, 02:34:03 PM |
|
i have an idea for how you could set yourself apart that may actually be a legitimately useful innovation. instead of relying on one algo for proof of work use a bunch of them. 10 20 maybe even 30 different algorithms. This would almost entirely eliminate the idea of any sort of specialized hardware like asics being developed which would cause this coin to be radically decentralized even compared to bitcoin. You could simply cycle through the list 1 at a time and when you get to the end start back at the beginning.
|
Rep Thread: https://bitcointalk.org/index.php?topic=381041If one can not confer upon another a right which he does not himself first possess, by what means does the state derive the right to engage in behaviors from which the public is prohibited?
|
|
|
SaltySpitoon
Legendary
Offline
Activity: 2590
Merit: 2156
Welcome to the SaltySpitoon, how Tough are ya?
|
|
April 04, 2013, 02:41:40 PM |
|
I would also be on board with a new type of coin, I couldn't help much software dev wise, but I am rather good at coming up with feature ideas (just not so much implementing them)
I've actually downloaded just about every coin client that wouldn't give me a virus, so I would have personal knowledge when it comes to various coin topics on the alt currency board. I have a definite opinion of what people seem to like, and not so much seem to like. So if you have any questions for me, I'd be happy to answer.
|
|
|
|
johnyj
Legendary
Offline
Activity: 1988
Merit: 1012
Beyond Imagination
|
|
April 04, 2013, 04:36:16 PM |
|
There is one change that really make difference: Late adopter advantage, means the later you join the game, the better you are positioned.
For example, all the mined coin can only exist for 100 days, they have the highest value when it first mined, and they get less and less each day
But it seems this is close to the character of all the non-lasting goods like food, cloth etc... Then no one will care about it
Out of this, I don't see any new ideas being fundamentally different
|
|
|
|
Joe_Bauers
|
|
April 04, 2013, 10:45:42 PM |
|
I've had an idea floating around for a while, either for Bitcoin, or an alt coin, in which a Lamport one-time signature scheme is incorporated into the flow, which would (should) make the coin future proof.
|
|
|
|
🏰 TradeFortress 🏰 (OP)
Bitcoin Veteran
VIP
Legendary
Offline
Activity: 1316
Merit: 1043
👻
|
|
April 04, 2013, 10:50:41 PM |
|
i have an idea for how you could set yourself apart that may actually be a legitimately useful innovation. instead of relying on one algo for proof of work use a bunch of them. 10 20 maybe even 30 different algorithms. This would almost entirely eliminate the idea of any sort of specialized hardware like asics being developed which would cause this coin to be radically decentralized even compared to bitcoin. You could simply cycle through the list 1 at a time and when you get to the end start back at the beginning.
There are no 30 different, suitable and tested algorithms.
|
|
|
|
Anon136
Legendary
Offline
Activity: 1722
Merit: 1217
|
|
April 04, 2013, 10:55:11 PM |
|
i have an idea for how you could set yourself apart that may actually be a legitimately useful innovation. instead of relying on one algo for proof of work use a bunch of them. 10 20 maybe even 30 different algorithms. This would almost entirely eliminate the idea of any sort of specialized hardware like asics being developed which would cause this coin to be radically decentralized even compared to bitcoin. You could simply cycle through the list 1 at a time and when you get to the end start back at the beginning.
There are no 30 different, suitable and tested algorithms. ok well you know, how ever many there are, that many.
|
Rep Thread: https://bitcointalk.org/index.php?topic=381041If one can not confer upon another a right which he does not himself first possess, by what means does the state derive the right to engage in behaviors from which the public is prohibited?
|
|
|
|