[4+ EH] Slush Pool (slushpool.com); Overt AsicBoost; World First Mining Pool

[ewitte]

So do you think that almost 50% drop for good few days in a row is normal? I know that luck is important, our hashing power is not increasing compare to other pools and we get smaller chunk of the pie, and difficulty is increasing but would that factor for such a difference? I was expecting drop in reward but not so quick and so much.

Measuring just a few days of earnings is not enough data to make any reasonable conclusions, especially at these difficulty levels and slush's pool speed.  "Luck" takes a significantly longer amount of time to "even out" than what most people think it does.

One thing is for sure if you have 2-3 times the hashing power it should theoretically blow through the unlucky blocks 2-3 times faster.

[Lucko]

There is a problem with the pool and not sure why slush is not here to correct any issues.

What makes you say that? (Besides the chance that you may be Lucko's alter ego   )

Well from 8th of May my daily reward dropped from 0.101 to 0.052 and no I'm not anyones alter ego.

Ahh I see, yeah pool must be broke

So do you think that almost 50% drop for good few days in a row is normal? I know that luck is important, our hashing power is not increasing compare to other pools and we get smaller chunk of the pie, and difficulty is increasing but would that factor for such a difference? I was expecting drop in reward but not so quick and so much.

Yes his answer was stupid... I answer you also. Read my. Or you can read eleuthria, but then I have to add that difficulty also increased.

[KNK]

Lets say it is connection problem, but even if the pool is directly connected to BTCguild will not have an invalid, as we will simply not have found a block in this case = no end of the round.
Unless the next round is shorter than 30-40min it doesn't matter for the score method if the block is invalid or we just didn't found it at all.

Thanks to ASICminer - BTCguild and Bitminter together are ~50% of the network = if the pool is not connected to them directly, it is not connected to half of the network - yes it is a connection problem (the last 3 invalid blocks where after they have found a block first just a few seconds before us), but not because of the pool connection, but because of ASICminer

[Lucko]

Lets say it is connection problem, but even if the pool is directly connected to BTCguild will not have an invalid, as we will simply not have found a block in this case = no end of the round.
Unless the next round is shorter than 30-40min it doesn't matter for the score method if the block is invalid or we just didn't found it at all.

Thanks to ASICminer - BTCguild and Bitminter together are ~50% of the network = if the pool is not connected to them directly, it is not connected to half of the network - yes it is a connection problem (the last 3 invalid blocks where after they have found a block first just a few seconds before us), but not because of the pool connection, but because of ASICminer

I'm interested. How do you know it is seconds?

[nottm28]

I refuse to quote that lucko character any more - I think he just loves the conflict - every time he's quoted he has 5 more idiot things to say.
He says he's left the pool but he insists on posting here to troll slush.
The best thing we can do as a group is to ignore his posts and don't quote him. Eventually he'll get bored and maybe move to the BFL thread where he can troll all he likes.

[ewitte]

I feel better I've now got cg monitor running on all 3 machines with 3 different pool settings on BTCGuild/Slush

[KNK]

I'm interested. How do you know it is seconds?

Simple - comapre https://mining.bitcoin.cz/stats/ and http://blockchain.info/ for each of them
Example for round 18050:
 we should have found 236382 if it wasn't invalid at '2013-05-16 00:59:30', but the actual 236382 ( http://blockchain.info/block/000000000000010ff86768774d6c3ea12dbbee106dcfc389e8b7bd3b64e98744 ) was found at '2013-05-16 00:57:38' (timestamp based and received from blockchain at '2013-05-16 00:59:33') from Bitminter and the next one (236383) was also found from Bitminter ... we had no chance to beat that right?

[Lucko]

I refuse to quote that lucko character any more - I think he just loves the conflict - every time he's quoted he has 5 more idiot things to say.
He says he's left the pool but he insists on posting here to troll slush.
The best thing we can do as a group is to ignore his posts and don't quote him. Eventually he'll get bored and maybe move to the BFL thread where he can troll all he likes.

If you can't look truth in a face they yes I'm trolling. But if you look at my posts. They are fair. If there is a reason and way how something can happen without saying it is pool problem I did. If you can't accept simple facts then I can't help you. When you say you are stupid thinking there is a problem even if you see it this doesn't help and you are making more damage as you think.

[Lucko]

I'm interested. How do you know it is seconds?

Simple - comapre https://mining.bitcoin.cz/stats/ and http://blockchain.info/ for each of them
Example for round 18050:
 we should have found 236382 if it wasn't invalid at '2013-05-16 00:59:30', but the actual 236382 ( http://blockchain.info/block/000000000000010ff86768774d6c3ea12dbbee106dcfc389e8b7bd3b64e98744 ) was found at '2013-05-16 00:57:38' (timestamp based and received from blockchain at '2013-05-16 00:59:33') from Bitminter and the next one (236383) was also found from Bitminter ... we had no chance to beat that right?

OK but if I click on invalid block I get "Block Not Found" so I can't see timestamp to compare them. I'm interested in time when it was send to network not the time it was fond by the pool because nobody is saying the time is synchronised...

EDIT: just figure it out. It took 2 minutes not 2 seconds(if time is synchronised). This is too long not to detect a new block. For blockchain it doesn't matter but for a pool this it too long... If you have cgminer you can see new block detected msg before you see that info on blockchain... Sometime it is a big difference up to a minute or more.

EDIT2: Look at differences in time for last block.

18078   2013-05-17 21:59:35   2:25:00
Timestamp   2013-05-17 21:59:18

[sunriselad]

I refuse to quote that lucko character any more - I think he just loves the conflict - every time he's quoted he has 5 more idiot things to say.
He says he's left the pool but he insists on posting here to troll slush.
The best thing we can do as a group is to ignore his posts and don't quote him. Eventually he'll get bored and maybe move to the BFL thread where he can troll all he likes.

I think most people know, they humour him.
Surely everyone has noticed it is BTCGuild that he promotes in his signature.

[Trongersoll]

heh, can't we kick our clock ahead a minute or two? 

[Lucko]

I think most people know, they humour him.
Surely everyone has noticed it is BTCGuild that he promotes in his signature.

It is a simple reason. It is PPS. If you have Slush on a long run and someone starts mining it resets your score for all miners if you have any shares on a worker that started mining again with more then 30 minutes pause. That is why I relay wouldn't like that to happen. I'm sure you didn't know that...

[nottm28]

I feel better I've now got cg monitor running on all 3 machines with 3 different pool settings on BTCGuild/Slush

cgminer is not a gui - so more difficult to set up for new starters - but it's infinitely better than guiminer imo - good news

[Lucko]

heh, can't we kick our clock ahead a minute or two? 

It doesn't work that way. Clock is synchronised with network. And you were probably thinking punting it back  

[haveagr8day]

I'm interested. How do you know it is seconds?

Simple - comapre https://mining.bitcoin.cz/stats/ and http://blockchain.info/ for each of them
Example for round 18050:
 we should have found 236382 if it wasn't invalid at '2013-05-16 00:59:30', but the actual 236382 ( http://blockchain.info/block/000000000000010ff86768774d6c3ea12dbbee106dcfc389e8b7bd3b64e98744 ) was found at '2013-05-16 00:57:38' (timestamp based and received from blockchain at '2013-05-16 00:59:33') from Bitminter and the next one (236383) was also found from Bitminter ... we had no chance to beat that right?

OK but if I click on invalid block I get "Block Not Found" so I can't see timestamp to compare them. I'm interested in time when it was send to network not the time it was fond by the pool because nobody is saying the time is synchronised...

EDIT: just figure it out. It took 2 minutes not 2 seconds(if time is synchronised). This is too long not to detect a new block. For blockchain it doesn't matter but for a pool this it too long... If you have cgminer you can see new block detected msg before you see that info on blockchain... Sometime it is a big difference up to a minute or more.

EDIT2: Look at differences in time for last block.

18078   2013-05-17 21:59:35   2:25:00
Timestamp   2013-05-17 21:59:18

The time received by the network will be the more accurate time, which is only a 2 second difference. The timestamp can be off by up to 2 hours ahead, and the median time of the last 11 blocks behind. Because of this, miners will sometimes fudge the timestamp as somewhat of an extra nonce value.

See: https://en.bitcoin.it/wiki/Block_timestamp

Received time BitMinter: 00:59:33
vs.
Slush's Block: 00:59:30

[Lucko]

I'm interested. How do you know it is seconds?

Simple - comapre https://mining.bitcoin.cz/stats/ and http://blockchain.info/ for each of them
Example for round 18050:
 we should have found 236382 if it wasn't invalid at '2013-05-16 00:59:30', but the actual 236382 ( http://blockchain.info/block/000000000000010ff86768774d6c3ea12dbbee106dcfc389e8b7bd3b64e98744 ) was found at '2013-05-16 00:57:38' (timestamp based and received from blockchain at '2013-05-16 00:59:33') from Bitminter and the next one (236383) was also found from Bitminter ... we had no chance to beat that right?

OK but if I click on invalid block I get "Block Not Found" so I can't see timestamp to compare them. I'm interested in time when it was send to network not the time it was fond by the pool because nobody is saying the time is synchronised...

EDIT: just figure it out. It took 2 minutes not 2 seconds(if time is synchronised). This is too long not to detect a new block. For blockchain it doesn't matter but for a pool this it too long... If you have cgminer you can see new block detected msg before you see that info on blockchain... Sometime it is a big difference up to a minute or more.

EDIT2: Look at differences in time for last block.

18078   2013-05-17 21:59:35   2:25:00
Timestamp   2013-05-17 21:59:18

The time received by the network will be the more accurate time, which is only a 2 second difference. The timestamp can be off by up to 2 hours ahead, and the median time of the last 11 blocks behind. Because of this, miners will sometimes fudge the timestamp as somewhat of an extra nonce value.

See: https://en.bitcoin.it/wiki/Block_timestamp

Received time BitMinter: 00:59:33
vs.
Slush's Block: 00:59:30

Isn't receive time diffident for anyone on the network? And if this would be a case Slush should get this one...

[Psyrick]

18078    2013-05-17 21:59:35    2:25:00    20411945    1951    0.00001476    236670    25.11154300
18077    2013-05-17 19:34:35    2:27:43    20693673    2462    0.00268799    236657    25.12165813

Nearly identical round time. Nearly identical shares. 5.49% reward. I don't care how much cheating is being prevented by the score system, you can't convince me this isn't pure bull.
I'm going to BTCGuild.
Later Slush.

[Geez]

18078    2013-05-17 21:59:35    2:25:00    20411945    1951    0.00001476    236670    25.11154300
18077    2013-05-17 19:34:35    2:27:43    20693673    2462    0.00268799    236657    25.12165813

Nearly identical round time. Nearly identical shares. 5.49% reward. I don't care how much cheating is being prevented by the score system, you can't convince me this isn't pure bull.
I'm going to BTCGuild.
Later Slush.

I experienced similar things.

[haveagr8day]

I'm interested. How do you know it is seconds?

Simple - comapre https://mining.bitcoin.cz/stats/ and http://blockchain.info/ for each of them
Example for round 18050:
 we should have found 236382 if it wasn't invalid at '2013-05-16 00:59:30', but the actual 236382 ( http://blockchain.info/block/000000000000010ff86768774d6c3ea12dbbee106dcfc389e8b7bd3b64e98744 ) was found at '2013-05-16 00:57:38' (timestamp based and received from blockchain at '2013-05-16 00:59:33') from Bitminter and the next one (236383) was also found from Bitminter ... we had no chance to beat that right?

OK but if I click on invalid block I get "Block Not Found" so I can't see timestamp to compare them. I'm interested in time when it was send to network not the time it was fond by the pool because nobody is saying the time is synchronised...

EDIT: just figure it out. It took 2 minutes not 2 seconds(if time is synchronised). This is too long not to detect a new block. For blockchain it doesn't matter but for a pool this it too long... If you have cgminer you can see new block detected msg before you see that info on blockchain... Sometime it is a big difference up to a minute or more.

EDIT2: Look at differences in time for last block.

18078   2013-05-17 21:59:35   2:25:00
Timestamp   2013-05-17 21:59:18

The time received by the network will be the more accurate time, which is only a 2 second difference. The timestamp can be off by up to 2 hours ahead, and the median time of the last 11 blocks behind. Because of this, miners will sometimes fudge the timestamp as somewhat of an extra nonce value.

See: https://en.bitcoin.it/wiki/Block_timestamp

Received time BitMinter: 00:59:33
vs.
Slush's Block: 00:59:30

Isn't receive time diffident for anyone on the network? And if this would be a case Slush should get this one...

Since BitMinter doesn't show their generation time to the second, it would be hard to compare. Blockchain.info received theirs 3 seconds after Slush's was generated. But based on our most recent one, 236670, the time for it to propagate from generation on Slush to the time Blockchain.info sees it is about a minute. By that, BitMinter may still have generated it before Slush.

Block 236670 Found at (Slush's Pool): 21:59:35
Block 236670 Received Time (Blockchain.info): 22:00:24

[Lucko]

18078    2013-05-17 21:59:35    2:25:00    20411945    1951    0.00001476    236670    25.11154300
18077    2013-05-17 19:34:35    2:27:43    20693673    2462    0.00268799    236657    25.12165813

Nearly identical round time. Nearly identical shares. 5.49% reward. I don't care how much cheating is being prevented by the score system, you can't convince me this isn't pure bull.
I'm going to BTCGuild.
Later Slush.

It is a 20% difference in shares. You had some problems in 19078. And if they were on the end of the round you are on a loosing side of this. But if there were on the start you would get about the same as for 18077. This is just a way it is with Slush method. Unless you are saying that you were mining all the time and you did monitor it or log it and you didn't see any brakes in mining.

P.S.: You are not big but BTC Guild has 51% problems. Consider any other pool to help them out.