Pool: [ 160.32 GHash/s | 118 Users | 240 Workers ] - What happened? I got bumped off?
shouldn't have been bumped offsite lagged while 2 days of NMC shares processed
caught up now
Topic: [0Th]Ozcoin Pooled Mining |DGM 1%|Stratum+VarDiff port 80|NEW CN mining| - page 84. (Read 398185 times)
October 11, 2012, 11:56:03 AM
site lagged while 2 days of NMC shares processed
caught up now
October 11, 2012, 11:35:37 AM
I usually don't follow everything you say.
Not sure if I was just insulted, there.
October 11, 2012, 11:29:00 AM
Pool: [ 160.32 GHash/s | 118 Users | 240 Workers ] - What happened? I got bumped off?
October 11, 2012, 08:31:36 AM
What about Namecoin-Mining? Did it get stuck?
sure looks like itQuote
Namecoin Round: [ 1,250,782 Shares | 2d 9h 56m | 90.34% Efficiency ]
tonight's thing to fix
cheers
October 11, 2012, 06:20:18 AM
Are we trying to set a new record? 17hr 37minutes since last block.
I keep tweaking things.. right now I'm glad I have 1/2 pointed here and 1/2 at eclipse.
M
I keep tweaking things.. right now I'm glad I have 1/2 pointed here and 1/2 at eclipse.
M
If everyone divided their hashrate among all pools equally (and thus all pools had equal hashrate) there'd be very little variance in earnings at all and PPS wouldn't be needed.
I usually don't follow everything you say. But this one got me thinking. This implies that at any given moment, some pool somewhere is having good luck, therefore, if you were in all pools, you'd never have bad luck. I don't think that's true. All pools could be having bad luck, which means PPS would still come out ahead.
Did I miss something along the way?
M
To put it yet another way, this statement:
This implies that at any given moment, some pool somewhere is having good luck, therefore, if you were in all pools, you'd never have bad luck.
which, as you suggest is incorrect, is ignoring the fact that we're looking at an average earning over many blocks. What I'm saying (and what Meni Rosenfeld explained in another thread a while back) is that as time goes on and you submit shares to blocks, the variance of mining at a single non PPS pool is much greater than the variance of mining at all non PPS pools. There's more to it than that, but that's the basic gist of it.
I usually don't follow everything you say.
Not sure if I was just insulted, there.
/OT
October 11, 2012, 05:00:00 AM
Yes, you missed that the weighted average luck of all the pools you are mining on is ... the weighted average luck of all the pools you are mining on
So it's adding up the hash rates of all the pools you are mining on and thus reducing your pool block finding variance due to that.
Even the whole network doesn't necessarily match PPS in the short term ... it's always of course one of the 3: above, the same, below
So it's adding up the hash rates of all the pools you are mining on and thus reducing your pool block finding variance due to that.
Even the whole network doesn't necessarily match PPS in the short term ... it's always of course one of the 3: above, the same, below
October 11, 2012, 04:28:59 AM
Are we trying to set a new record? 17hr 37minutes since last block.
I keep tweaking things.. right now I'm glad I have 1/2 pointed here and 1/2 at eclipse.
M
I keep tweaking things.. right now I'm glad I have 1/2 pointed here and 1/2 at eclipse.
M
If everyone divided their hashrate among all pools equally (and thus all pools had equal hashrate) there'd be very little variance in earnings at all and PPS wouldn't be needed.
I usually don't follow everything you say. But this one got me thinking. This implies that at any given moment, some pool somewhere is having good luck, therefore, if you were in all pools, you'd never have bad luck. I don't think that's true. All pools could be having bad luck, which means PPS would still come out ahead.
Did I miss something along the way?
M
October 10, 2012, 11:29:50 AM
Are we trying to set a new record? 17hr 37minutes since last block.
I keep tweaking things.. right now I'm glad I have 1/2 pointed here and 1/2 at eclipse.
M
-ozbot/#ozcoin- BLOCK: 202682 | 13044563 shares | 18h 0m 23s | Amount: 50.0792249 | Found by slaveindebt | 427.04% of difficultyI keep tweaking things.. right now I'm glad I have 1/2 pointed here and 1/2 at eclipse.
M
record is 800% long way off
I should have said something hours ago!!
M
October 10, 2012, 11:09:04 AM
Are we trying to set a new record? 17hr 37minutes since last block.
I keep tweaking things.. right now I'm glad I have 1/2 pointed here and 1/2 at eclipse.
M
-ozbot/#ozcoin- BLOCK: 202682 | 13044563 shares | 18h 0m 23s | Amount: 50.0792249 | Found by slaveindebt | 427.04% of difficultyI keep tweaking things.. right now I'm glad I have 1/2 pointed here and 1/2 at eclipse.
M
record is 800% long way off
October 10, 2012, 11:03:34 AM
Are we trying to set a new record? 17hr 37minutes since last block.
I keep tweaking things.. right now I'm glad I have 1/2 pointed here and 1/2 at eclipse.
M
I keep tweaking things.. right now I'm glad I have 1/2 pointed here and 1/2 at eclipse.
M
If everyone divided their hashrate among all pools equally (and thus all pools had equal hashrate) there'd be very little variance in earnings at all and PPS wouldn't be needed.
And then, I suppose, there would be world peace.
Hope you don't mind if don't hold my breath.
October 10, 2012, 10:55:11 AM
Are we trying to set a new record? 17hr 37minutes since last block.
I keep tweaking things.. right now I'm glad I have 1/2 pointed here and 1/2 at eclipse.
M
I keep tweaking things.. right now I'm glad I have 1/2 pointed here and 1/2 at eclipse.
M
If everyone divided their hashrate among all pools equally (and thus all pools had equal hashrate) there'd be very little variance in earnings at all and PPS wouldn't be needed.
October 10, 2012, 10:41:36 AM
Are we trying to set a new record? 17hr 37minutes since last block.
I keep tweaking things.. right now I'm glad I have 1/2 pointed here and 1/2 at eclipse.
M
I keep tweaking things.. right now I'm glad I have 1/2 pointed here and 1/2 at eclipse.
M
October 09, 2012, 12:09:19 PM
We have run a duplicate payout for block 201731 - miners that had a score have now been paid twice for that block 
I think there's a better version of that "Who's Awesome" somewhere on the internet
October 09, 2012, 09:02:09 AM
share processing got stuck as we found block 202481
we have restarted processing but stats will look weird for the next few hours
we have restarted processing but stats will look weird for the next few hours
October 08, 2012, 04:16:01 AM
now you make me blush
thanks
thanks
October 08, 2012, 04:14:17 AM
We have run a duplicate payout for block 201731 - miners that had a score have now been paid twice for that block
October 08, 2012, 03:53:41 AM
We have run a duplicate payout for block 201731 - miners that had a score have now been paid twice for that block
The balance of the unusual fees I will put into the "buffer" - DGM has no "buffer" as such. I do however ensure I keep funds aside from times of good luck to cover times of bad, an unofficial buffer if you like
Best wishes
Graeme
The balance of the unusual fees I will put into the "buffer" - DGM has no "buffer" as such. I do however ensure I keep funds aside from times of good luck to cover times of bad, an unofficial buffer if you like
Best wishes
Graeme
October 07, 2012, 07:25:44 AM
Thanks guys, learned something on a sunday morning !
October 07, 2012, 06:51:19 AM
I think the confusion lies in the fact that OzCoin shows the found block was worth 157 BTC while the pool only paid out 47 BTC, so the questions is then where did the 110 BTC go ?
As you describe DGM; a share that is being submitted at time T is scored compared to the other shares based on the value of the transaction that is submitted to the network at time T, even if the pool does not find the block that transaction is included in. If someone includes a transaction fee of 100.000 BTC, each share submitted at that time would have an abnormally high score value whether or not OzCoin finds the block.
Seeing as this score is only to determine how much of the total generated BTC a share is worth to a miner compared to other shares submitted by miners, the question remains if the transaction fee received by OzCoin is 10.000 BTC then 10.000 BTC + blockreward should be divided amongst all all the shares submitted within the scoring period.
As I started out saying, now it looks like OzCoin received 157 BTC and only paid out 47 BTC. Regardless of any reward system, the payout on each share should have been roughly 3 times at high as it usually is (with DGM less so than with PPS, but higher over a longer period of submitted shares).
This is how I see it at least
The mistaken assumption is that the pool always pays out exactly 100% of the block reward.As you describe DGM; a share that is being submitted at time T is scored compared to the other shares based on the value of the transaction that is submitted to the network at time T, even if the pool does not find the block that transaction is included in. If someone includes a transaction fee of 100.000 BTC, each share submitted at that time would have an abnormally high score value whether or not OzCoin finds the block.
Seeing as this score is only to determine how much of the total generated BTC a share is worth to a miner compared to other shares submitted by miners, the question remains if the transaction fee received by OzCoin is 10.000 BTC then 10.000 BTC + blockreward should be divided amongst all all the shares submitted within the scoring period.
As I started out saying, now it looks like OzCoin received 157 BTC and only paid out 47 BTC. Regardless of any reward system, the payout on each share should have been roughly 3 times at high as it usually is (with DGM less so than with PPS, but higher over a longer period of submitted shares).
This is how I see it at least
Basically, PPLNS absorbs no variance (all block reward is paid), PPS absorbs all reward (payment is completely independent of block rewards), and DGM absorbs some variance - in lucky times less than the block reward is given, in unlucky times more than the block reward is given. On average it evens out.
To further complicate things, this description is correct only if the block reward is fixed (and "luck" only relates to how many blocks are found). But if the block reward is variable (e.g. due to transaction fees), true PPLNS also doesn't always pay out exactly the block reward. That's the only way to be completely hopping-proof and to make sure expected payout is always equal to expected contribution. The differences will usually be minor though.
To wait until block reward is known would involve rewriting software and making it something that is not DGM in my understanding.
More concretely, it would enable pool-hopping when tx fees become significant. Hoppers will mine at your pool when the fees are low (knowing that the low fees will not be factored into their score) and as more time passes without a block and fees increase, hop to solo or a pool that does handle fees correctly (where they are credited for the high fees). October 07, 2012, 06:48:36 AM
I think the confusion lies in the fact that OzCoin shows the found block was worth 157 BTC while the pool only paid out 47 BTC, so the questions is then where did the 110 BTC go ?
As you describe DGM; a share that is being submitted at time T is scored compared to the other shares based on the value of the transaction that is submitted to the network at time T, even if the pool does not find the block that transaction is included in. If someone includes a transaction fee of 100.000 BTC, each share submitted at that time would have an abnormally high score value whether or not OzCoin finds the block.
Seeing as this score is only to determine how much of the total generated BTC a share is worth to a miner compared to other shares submitted by miners, the question remains if the transaction fee received by OzCoin is 10.000 BTC then 10.000 BTC + blockreward should be divided amongst all all the shares submitted within the scoring period.
As I started out saying, now it looks like OzCoin received 157 BTC and only paid out 47 BTC. Regardless of any reward system, the payout on each share should have been roughly 3 times at high as it usually is (with DGM less so than with PPS, but higher over a longer period of submitted shares).
This is how I see it at least
The way I understand DGM is that the pool operator takes some of the variance from miners (depending on pool variables), thus on long rounds we pay >49BTC and on short rounds 49BTC< (allowing for 2% fee)As you describe DGM; a share that is being submitted at time T is scored compared to the other shares based on the value of the transaction that is submitted to the network at time T, even if the pool does not find the block that transaction is included in. If someone includes a transaction fee of 100.000 BTC, each share submitted at that time would have an abnormally high score value whether or not OzCoin finds the block.
Seeing as this score is only to determine how much of the total generated BTC a share is worth to a miner compared to other shares submitted by miners, the question remains if the transaction fee received by OzCoin is 10.000 BTC then 10.000 BTC + blockreward should be divided amongst all all the shares submitted within the scoring period.
As I started out saying, now it looks like OzCoin received 157 BTC and only paid out 47 BTC. Regardless of any reward system, the payout on each share should have been roughly 3 times at high as it usually is (with DGM less so than with PPS, but higher over a longer period of submitted shares).
This is how I see it at least
There is no "buffer" like some systems, if the pool runs out of Bitcoins they come from my pocket, if the pool is lucky I benefit.
The pool software does not look at the txn fee when a block is found - per DGM scoring this is already factored in.
On this occasion the pool has had a windfall, as I said in EDIT: in previous post, I am looking at making an extra payout to miners in that round to share the windfall - even though the DGM worked as it should. As a sign of goodwill - not because something is wrong with the pool software.
To wait until block reward is known would involve rewriting software and making it something that is not DGM in my understanding.
Again I would like to point out that this is an extremely unusual situation and if any other pool had found this block your score would reflect the txns fees from that time - and the pool would be paying out extra - not benefiting.
Jump to: