[ANN] profit switching auto-exchanging pool - www.middlecoin.com - page 451.

coinGeek

full member

Activity: 221

Merit: 100

Quote from: nitrocs on September 18, 2013, 10:02:13 AM

Why not split pool resources and mine the top two profitable coins concurrently
Split all proceeds based on N shares accepted per user?

Is "N shares" different from all the shares a user submitted? Because that's one of the things being discussed.

nitrocs

newbie

Activity: 47

Merit: 0

Why not split pool resources and mine the top two profitable coins concurrently
Split all proceeds based on N shares accepted per user?

coinGeek

full member

Activity: 221

Merit: 100

Quote from: h2odysee on September 17, 2013, 09:37:56 PM

There is too much overhead in switching coins to switch every single share. Or even every few minutes.

It seems like if the mining software cooperated you could do it. In cgminer you can use "Load Balance" to mine on more than one worker efficiently. What if each user had a bunch of workers on one pool and each could be rotated to different coins as needed? Require users to load balance X workers to mine on your new pool and Bob is your mothers brother.

Don't forget whos idea this was Shocked

h2odysee

full member

Activity: 238

Merit: 119

Quote from: coinGeek on September 17, 2013, 08:35:46 PM

I was trying to think of cheats that work if the pool just divided total pool income by shares submitted. My examples might not be the best... Most cheats don't harm the pool as a whole so much as does the steady 'set it and forget it' miners.

Back to the subject of dividing up the hash rate into smaller units though. I never looked at pool software but it seems like it's all about routing shares for miners to solve as they need them, plus accounting. If you could break the multiple coin mining jobs into single shares and and pass them out at random it would be fair and cheat proof. "If You Could". I'm guessing that there's an overhead to switching coins that makes that less efficient. Right?

Ah yeah, that would cheat the other miners out of money, if my pool divided total pool income by total shares of every coin.

There is too much overhead in switching coins to switch every single share. Or even every few minutes.

coinGeek

full member

Activity: 221

Merit: 100

Quote from: h2odysee on September 17, 2013, 08:03:55 PM

So "hard" coins are ones with high reject rates?

I was trying to think of cheats that work if the pool just divided total pool income by shares submitted. My examples might not be the best... Most cheats don't harm the pool as a whole so much as does the steady 'set it and forget it' miners.

Back to the subject of dividing up the hash rate into smaller units though. I never looked at pool software but it seems like it's all about routing shares for miners to solve as they need them, plus accounting. If you could break the multiple coin mining jobs into single shares and and pass them out at random it would be fair and cheat proof. "If You Could". I'm guessing that there's an overhead to switching coins that makes that less efficient. Right?

h2odysee

full member

Activity: 238

Merit: 119

Quote from: coinGeek on September 17, 2013, 07:06:05 PM

Of course one cheat would be to detect the when you switch to a "hard" coin that has lower shares per time and go elsewhere until it switches to an "easy" coin. Then come back full force. Then get paid off per share at the same rate as the guys that stayed the whole time.

So "hard" coins are ones with high reject rates?

I don't consider that cheating. People should feel free to do that. It won't lower anyone's profits. In fact, it might increase them.

But there's really no reason to, unless that particular miner has much higher reject rates than others in the pool.

coinGeek

full member

Activity: 221

Merit: 100

Quote from: h2odysee on September 17, 2013, 06:04:41 PM

If I split our profits as a whole pool, proportional to shares submitted, and not for each coin individually, there are ways people can cheat.

Of course one cheat would be to detect the when you switch to a "hard" coin that has lower shares per time and go elsewhere until it switches to an "easy" coin. Then come back full force. Then get paid off per share at the same rate as the guys that stayed the whole time.

You could come with tricky schemes that try to punish people who do that but they also can ding the "honest" people that just have to drop off line for a while or whatever. Honestly, one of the things I like about Middlecoin is that it doesn't need that bs.

Damnsammit

sr. member

Activity: 406

Merit: 250

You should just release the source code... competition is already there, why not just let the world see? Cheesy

h2odysee

full member

Activity: 238

Merit: 119

Quote from: hendo420 on September 17, 2013, 03:19:54 PM

Is the pool currently splitting shares of each coin mined? or is it spiting shares after the coins have been traded into bitcoin?

Once the pool changes to a new coin are the current open shares lost? Or do they roll over to the next coin mined?

How long do shares last? 1 round or more?

The shares are split for each coin mined, individually. If I split our profits as a whole pool, proportional to shares submitted, and not for each coin individually, there are ways people can cheat.

When the pool switches coins, the shares are not lost. They are held until the next time we go back to that coin.

Shares last one block.

coinGeek

full member

Activity: 221

Merit: 100

Quote

I really want to split up the hashrate like you say. It's going to be hard to figure out how to make the payouts fair though. Do I randomly choose the people who get to mine on the smaller pool, or is it who has the lowest reject rate? And do their extra profits get distributed into the main pool?

I think I see the dilemma. If I understand middlecoin as it is now: Everyone gets credit for the shares they submit in the coin that's being mined at the time, and those are sold and paid off in BTC. You get what you mined and there aren't any easy ways to cheat. Not any cheats I know of anyway

I think most people feel they should be paid for the time they mine and the hash rate and I think people could get behind a system where you share results in some fair way.

I think h2odysee has the problem that you can't see hash rate very well from the pool. All you have is shares per time which varies from coin to coin and time to time. I mine crap coin for one hour with tons of rejects and you mine something better, how do you split it up fairly if you can't tell what my hash rate is very accurately? How do you know I didn't load balance %20 to another pool? -- for example?

hendo420

sr. member

Activity: 420

Merit: 250

Quote from: Damnsammit on September 17, 2013, 02:57:24 PM

Quote from: h2odysee on September 17, 2013, 02:46:21 PM

Yeah. 1 GH/s is way too much for some of these coins.

Part of my calculations include looking at past results of profits. So if there is a coin with some problem, that will be accounted for. I don't think pre-mining like you say would be good, because you'd lose out on that 2 minutes of profit. And past results are probably a good enough indicator.

I really want to split up the hashrate like you say. It's going to be hard to figure out how to make the payouts fair though. Do I randomly choose the people who get to mine on the smaller pool, or is it who has the lowest reject rate? And do their extra profits get distributed into the main pool?

It's a pool, so all of the profits would be distributed into the pool regardless of the coin that particular miner is mining.

2 hour period:

Miner A is mining FTC
Miner B is mining CAP
Miner C is mining DCG
Miner D is mining CRAP
Miners E-Z are mining LTC

For that 2 hour period they all get shares that are distributed amongst all of the coins being mined. I am sure there are flaws in that idea since I have never operated a pool.

Regardless of the coin mined it is being transferred into bitcoin so its easy math to split it up.

TotalBTC*(UserShares/TotalShares)=RoundPayout
or something along those lines.

It doesn't matter who mines what, its a pool, no one gets special treatment or higher payouts per share.

Is the pool currently splitting shares of each coin mined? or is it spiting shares after the coins have been traded into bitcoin?

Once the pool changes to a new coin are the current open shares lost? Or do they roll over to the next coin mined?

How long do shares last? 1 round or more?

What needs to happen is the shares should be counted over all per round no matter the coin mined, then each round is paid out in bitcoin split out amongst the whole. Shares should roll over to the next coin mined if a block is not found and the shares are still open.

TheQuin

hero member

Activity: 2576

Merit: 883

Freebitco.in Support https://bit.ly/2I9BVS2

Quote from: h2odysee on September 17, 2013, 02:46:21 PM

Yeah. 1 GH/s is way too much for some of these coins.

Don't worry your customer service policy should soon solve that problem for you.

Damnsammit

sr. member

Activity: 406

Merit: 250

Quote from: h2odysee on September 17, 2013, 02:46:21 PM

Yeah. 1 GH/s is way too much for some of these coins.

Part of my calculations include looking at past results of profits. So if there is a coin with some problem, that will be accounted for. I don't think pre-mining like you say would be good, because you'd lose out on that 2 minutes of profit. And past results are probably a good enough indicator.

I really want to split up the hashrate like you say. It's going to be hard to figure out how to make the payouts fair though. Do I randomly choose the people who get to mine on the smaller pool, or is it who has the lowest reject rate? And do their extra profits get distributed into the main pool?

It's a pool, so all of the profits would be distributed into the pool regardless of the coin that particular miner is mining.

2 hour period:

Miner A is mining FTC
Miner B is mining CAP
Miner C is mining DCG
Miner D is mining CRAP
Miners E-Z are mining LTC

For that 2 hour period they all get shares that are distributed amongst all of the coins being mined. I am sure there are flaws in that idea since I have never operated a pool.

h2odysee

full member

Activity: 238

Merit: 119

Quote from: Damnsammit on September 17, 2013, 02:35:44 PM

Quote from: Wipeout2097 on September 17, 2013, 02:23:37 PM

Either way, I have the impression that any of these pools are a waste of time. Maybe these pools should have a rig with at least a single 7950 pre-mining the so called "most profitable coin" for 2 minutes before making the switch, to make sure that stalls and rejects don't eat up hash rate.

Blindingly switching according to calculations similar to what seen on coinchoose, doesn't cut it. My rig is stable, cgminer reports the same 870 Kh/s, yet HashCows mining KrugerCoin, the pool reports 160 ~ 200 for the full round. Where is the remaining hash rate going? Rejects, stalls, lag and God knows what more ...

Agree 100% with this... as I have stated for a while, I think this pool was a great idea but the hashrate is too large for it to be effective unless he could split the hashrate up and have 80% mine LTC and 20% jump around and mine the alt-coins. 1GH/s is just too large to throw at these smaller coins even if they are the most profitable.

Yeah. 1 GH/s is way too much for some of these coins.

Part of my calculations include looking at past results of profits. So if there is a coin with some problem, that will be accounted for. I don't think pre-mining like you say would be good, because you'd lose out on that 2 minutes of profit. And past results are probably a good enough indicator.

I really want to split up the hashrate like you say. It's going to be hard to figure out how to make the payouts fair though. Do I randomly choose the people who get to mine on the smaller pool, or is it who has the lowest reject rate? And do their extra profits get distributed into the main pool?

Damnsammit

sr. member

Activity: 406

Merit: 250

Quote from: Wipeout2097 on September 17, 2013, 02:23:37 PM

Either way, I have the impression that any of these pools are a waste of time. Maybe these pools should have a rig with at least a single 7950 pre-mining the so called "most profitable coin" for 2 minutes before making the switch, to make sure that stalls and rejects don't eat up hash rate.

Blindingly switching according to calculations similar to what seen on coinchoose, doesn't cut it. My rig is stable, cgminer reports the same 870 Kh/s, yet HashCows mining KrugerCoin, the pool reports 160 ~ 200 for the full round. Where is the remaining hash rate going? Rejects, stalls, lag and God knows what more ...

Agree 100% with this... as I have stated for a while, I think this pool was a great idea but the hashrate is too large for it to be effective unless he could split the hashrate up and have 80% mine LTC and 20% jump around and mine the alt-coins. 1GH/s is just too large to throw at these smaller coins even if they are the most profitable.

Wipeout2097

sr. member

Activity: 840

Merit: 255

SportsIcon - Connect With Your Sports Heroes

Quote from: Hydroponica on September 17, 2013, 01:27:19 PM

Go here instead;

https://bitcointalksearch.org/topic/annpoolprofit-switchoptional-auto-exchange-per-coinvardiff-hashcows-293872

VARDIFF, better interface, and, the operator, is actually very fast at answering questions, and addressing concerns, rather than just telling his miners to go fuck themselves

Actually we are only complaining because the payouts are below expected. If they were fine, nobody would bother with technical details.

Either way, I have the impression that any of these pools are a waste of time. Maybe these pools should have a rig with at least a single 7950 pre-mining the so called "most profitable coin" for 2 minutes before making the switch, to make sure that stalls and rejects don't eat up hash rate.

Blindingly switching according to calculations similar to what seen on coinchoose, doesn't cut it. My rig is stable, cgminer reports the same 870 Kh/s, yet HashCows mining KrugerCoin, the pool reports 160 ~ 200 for the full round. Where is the remaining hash rate going? Rejects, stalls, lag and God knows what more ...

I suspect I'll stick with my setup of switching to the most profitable WITHIN a subset of "good" pools/coins, now that I had the hassle to set this crap up.

Wipeout2097

sr. member

Activity: 840

Merit: 255

SportsIcon - Connect With Your Sports Heroes

Quote from: Damnsammit on September 17, 2013, 01:43:19 PM

Difficulty doesn't really matter if the connection to the pool is fast and reliable, but it is good to see someone implement VARDIFF on a similar server since it will obviously keep people happier!

Damnsammit

sr. member

Activity: 406

Merit: 250

Quote from: Hydroponica on September 17, 2013, 01:27:19 PM

Fuck this guy.

Go here instead;

https://bitcointalksearch.org/topic/annpoolprofit-switchoptional-auto-exchange-per-coinvardiff-hashcows-293872

VARDIFF, better interface, and, the operator, is actually very fast at answering questions, and addressing concerns, rather than just telling his miners to go fuck themselves

lol

Not sure why, but your attitude reminds me of Jonah Hill. Gotta agree that the pool op is rather slow and incredibly short when it comes to responding to questions.

Difficulty doesn't really matter, but it is good to see someone implement VARDIFF on a similar server since it will obviously keep people happier!

Also, not sure what your avatar is, but it is pretty cool.

Hydroponica

full member

Activity: 182

Merit: 100

fml

Fuck this guy.

Go here instead;

https://bitcointalksearch.org/topic/annpoolprofit-switchoptional-auto-exchange-per-coinvardiff-hashcows-293872

VARDIFF, better interface, and, the operator, is actually very fast at answering questions, and addressing concerns, rather than just telling his miners to go fuck themselves

Epeus

member

Activity: 95

Merit: 10

Makes no sense, does it ?
I did my work, payed a 4% fee, I left to do some solo-experimenting and now my remaining coins are being processed so slow that it'll be 2025 before I receive the last payout, if ever.
I had an employee quitting his job the other day. If I donot pay him his last day at the office, he'll sue me for sure.
So far I've had no explanation whatsover as of why this is happening. An estimated BTC 11+ mined by idle workers are still stuck in the pool.
This could be a mechanism to prevent people from leaving the pool. For me it has the opposite effect.

So, mister pooloperator, please explain, so I can stfu.

Topic: [ANN] profit switching auto-exchanging pool - www.middlecoin.com - page 451. (Read 829908 times)