kano (OP)
Legendary
Offline
Activity: 4494
Merit: 1805
Linux since 1997 RedHat 4
|
|
November 30, 2015, 11:53:26 PM Last edit: December 01, 2015, 12:13:16 AM by kano |
|
... More to come on that soon as I gather up the data I have and get it over to him for independent analysis.
Make sure you learn about 'confidence level' and apply it to your reply, coz your comment about my pool size and the number of orphans clearly suggests you have no idea about it. Edit: ... As for the unworthy stuff... while you've posted two blocks that didn't make the cut... I'd be more curious as to if you had any blocks that the pool thought wouldn't make the cut that bitcoind (and the bitcoin network) actually accepted. Probably not. Why? Because math is math. Just because bitcoind is in C++ and ckpool is in C and eloipool is in python doesn't change the comparison of two numbers. It's either a block or it isn't. This whole "unworthy" thing is just nonsense. ...
Haven't been a programmer for very long or not much experience with different languages hey?
|
|
|
|
|
|
"There should not be any signed int. If you've found a signed int
somewhere, please tell me (within the next 25 years please) and I'll
change it to unsigned int." -- Satoshi
|
|
|
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction.
|
|
|
|
-ck
Legendary
Offline
Activity: 4102
Merit: 1632
Ruu \o/
|
|
December 01, 2015, 12:42:06 AM |
|
... As for the unworthy stuff... while you've posted two blocks that didn't make the cut... I'd be more curious as to if you had any blocks that the pool thought wouldn't make the cut that bitcoind (and the bitcoin network) actually accepted. Probably not. Why? Because math is math. Just because bitcoind is in C++ and ckpool is in C and eloipool is in python doesn't change the comparison of two numbers. It's either a block or it isn't. This whole "unworthy" thing is just nonsense. ...
This is effectively directed at me so while I was trying not to get dragged into this... We use diff as a double everywhere in ckpool which makes it super fast without nonsense converting to hex and doing comparisons and so on. Doubles are only accurate to integer sizes at 2^53. Given diff is expressed differently (heh) in bitcoind compared to a double, and numbers of 2^53 may one day not be unbelievable in bitcoin, then add to the fact that even the fractional accuracy gets less the closer to 2^53 a double gets, and that while there are standards for how rounding should occur with doubles, but diff isn't expressed directly as a double in the blockchain... testing the extremely rare very close diff share is not going to hurt.
|
Developer/maintainer for cgminer, ckpool/ckproxy, and the -ck kernel 2% Fee Solo mining at solo.ckpool.org -ck
|
|
|
kano (OP)
Legendary
Offline
Activity: 4494
Merit: 1805
Linux since 1997 RedHat 4
|
|
December 01, 2015, 01:15:43 AM |
|
Payout 386010 sent (finally) c8943e4a9a18625e18c333940134efe90481e8d68c86e2f8cc375f88a596c295 and confirmed
|
|
|
|
wizkid057
Legendary
Offline
Activity: 1223
Merit: 1006
|
|
December 01, 2015, 01:32:39 AM |
|
... As for the unworthy stuff... while you've posted two blocks that didn't make the cut... I'd be more curious as to if you had any blocks that the pool thought wouldn't make the cut that bitcoind (and the bitcoin network) actually accepted. Probably not. Why? Because math is math. Just because bitcoind is in C++ and ckpool is in C and eloipool is in python doesn't change the comparison of two numbers. It's either a block or it isn't. This whole "unworthy" thing is just nonsense. ...
This is effectively directed at me so while I was trying not to get dragged into this... We use diff as a double everywhere in ckpool which makes it super fast without nonsense converting to hex and doing comparisons and so on. Doubles are only accurate to integer sizes at 2^53. Given diff is expressed differently (heh) in bitcoind compared to a double, and numbers of 2^53 may one day not be unbelievable in bitcoin, then add to the fact that even the fractional accuracy gets less the closer to 2^53 a double gets, and that while there are standards for how rounding should occur with doubles, but diff isn't expressed directly as a double in the blockchain... testing the extremely rare very close diff share is not going to hurt. You lost me at "we use diff as a double everywhere"...... Jokes aside, you can even submit every share to bitcoind and test that way if you like. Doesn't matter to me. I just don't think skewing stats in the block list based on these non-blocks is reasonable.
|
|
|
|
-ck
Legendary
Offline
Activity: 4102
Merit: 1632
Ruu \o/
|
|
December 01, 2015, 01:34:11 AM |
|
You lost me at "we use diff as a double everywhere"...... Jokes aside, you can even submit every share to bitcoind and test that way if you like. Doesn't matter to me. I just don't think skewing stats in the block list based on these non-blocks is reasonable. Block stats are nothing to do with me. I was just explaining why I submit them to bitcoind.
|
Developer/maintainer for cgminer, ckpool/ckproxy, and the -ck kernel 2% Fee Solo mining at solo.ckpool.org -ck
|
|
|
wizkid057
Legendary
Offline
Activity: 1223
Merit: 1006
|
|
December 01, 2015, 01:38:17 AM |
|
You lost me at "we use diff as a double everywhere"...... Jokes aside, you can even submit every share to bitcoind and test that way if you like. Doesn't matter to me. I just don't think skewing stats in the block list based on these non-blocks is reasonable. Block stats are nothing to do with me. I was just explaining why I submit them to bitcoind. Oh, yeah I already understood that. I personally think the diff as a double thing is a coder speedup more than a performance speedup, but it is what it is and you have to cover edge cases like that for sure. But when bitcoind says "nope!" then it should just be ignored and not thrown in the block list as if it actually meant something. And if it must be for posterity then the next block's stats should at least display correctly. Just weird seeing the stats for real blocks being displayed so poorly due to basically a random share that is worthless.
|
|
|
|
kano (OP)
Legendary
Offline
Activity: 4494
Merit: 1805
Linux since 1997 RedHat 4
|
|
December 01, 2015, 01:42:49 AM Last edit: December 01, 2015, 02:29:28 AM by kano |
|
You lost me at "we use diff as a double everywhere"...... Jokes aside, you can even submit every share to bitcoind and test that way if you like. Doesn't matter to me. I just don't think skewing stats in the block list based on these non-blocks is reasonable. Block stats are nothing to do with me. I was just explaining why I submit them to bitcoind. Oh, yeah I already understood that. I personally think the diff as a double thing is a coder speedup more than a performance speedup, but it is what it is and you have to cover edge cases like that for sure. But when bitcoind says "nope!" then it should just be ignored and not thrown in the block list as if it actually meant something. And if it must be for posterity then the next block's stats should at least display correctly. Just weird seeing the stats for real blocks being displayed so poorly due to basically a random share that is worthless. Well I don't have much money so yeah I'm poor. Edit: oh ok I'll answer properly even though your comment is silly. Yeah on my block page I show all blocks found, even if they are orphans, and also show the 'Unworthy' blocks - yep extra information and nothing hidden ... unlike ... ... If we submit a share to bitcoind it will show there, no matter what, nothing hidden. But then of course if someone is actually trying to look at the pool luck, not at the block by block level where it really doesn't show you anything useful unless you got no idea about statistics but instead at the top there's a great table of block statistics with all sorts of numbers you wont find on the rich Eligius web page Edit2: though that may be coz, if you did put the CDF[Erl] on your web site and showed a year of it, you'd get people screaming "Da Fuk!" why is it so bad?!?
|
|
|
|
wizkid057
Legendary
Offline
Activity: 1223
Merit: 1006
|
|
December 01, 2015, 02:38:12 AM |
|
You lost me at "we use diff as a double everywhere"...... Jokes aside, you can even submit every share to bitcoind and test that way if you like. Doesn't matter to me. I just don't think skewing stats in the block list based on these non-blocks is reasonable. Block stats are nothing to do with me. I was just explaining why I submit them to bitcoind. Oh, yeah I already understood that. I personally think the diff as a double thing is a coder speedup more than a performance speedup, but it is what it is and you have to cover edge cases like that for sure. But when bitcoind says "nope!" then it should just be ignored and not thrown in the block list as if it actually meant something. And if it must be for posterity then the next block's stats should at least display correctly. Just weird seeing the stats for real blocks being displayed so poorly due to basically a random share that is worthless. Well I don't have much money so yeah I'm poor. Edit: oh ok I'll answer properly even though your comment is silly. Yeah on my block page I show all blocks found, even if they are orphans, and also show the 'Unworthy' blocks - yep extra information and nothing hidden ... unlike ... ... If we submit a share to bitcoind it will show there, no matter what, nothing hidden. But then of course if someone is actually trying to look at the pool luck, not at the block by block level where it really doesn't show you anything useful unless you got no idea about statistics but instead at the top there's a great table of block statistics with all sorts of numbers you wont find on the rich Eligius web page Edit2: though that may be coz, if you did put the CDF[Erl] on your web site and showed a year of it, you'd get people screaming "Da Fuk!" why is it so bad?!? So you're saying because the individual block lucks don't show anything it doesn't matter if some are displayed incorrectly......... then why do you display them at all for any blocks if it's so useless? And since you do display them, why do you display some incorrectly and make some blocks look luckier than they actually were? As for Eligius, every block listed is actually a block. Its stats are actually correct for every block on every line ... unlike ... ... lol. While I don't have a table aggregating luck, people interested are welcome to utilize the well defined public API and calculate out whatever they like for themselves... and they do regularly. Upcoming stats 2.0 (planned for Q1) will have a complete overload of information, including interactive luck charts.
|
|
|
|
kano (OP)
Legendary
Offline
Activity: 4494
Merit: 1805
Linux since 1997 RedHat 4
|
|
December 01, 2015, 03:06:37 AM |
|
Hmm why's my pool averaging better block change than your pool, currently for the last 80 blocks ...
|
|
|
|
wizkid057
Legendary
Offline
Activity: 1223
Merit: 1006
|
|
December 01, 2015, 03:07:46 AM |
|
Hmm why's my pool averaging better block change than your pool, currently for the last 80 blocks ...
Love the deflection Edit: Probably should point out that your definition of better is 107ms. And to directly answer your deflection question, the reason is probably some changes I recently made that I'm about to revert that appears to have added a hair of a delay in telling some eloipool instances about the new work.
|
|
|
|
kano (OP)
Legendary
Offline
Activity: 4494
Merit: 1805
Linux since 1997 RedHat 4
|
|
December 01, 2015, 03:55:27 AM |
|
Deflection? I was bored of that last one and thought it was finished anyway. You don't like showing all the failed blocks, I do. You want the failed block information to disappear, I don't. I have a good table at the top that resolves any confusion of what's going on with the pool, you require your users to do that themselves and hope they can all do that, until some future date. -- Yep I'm only ... oh it's got better Me: 1.8732658217225 You: 2.0306842107522 heh now I'm 157ms average faster ... yet I don't send out empty blocks, vs you do every time for this comparison. -- Though of way more interest Payout 386044 sent 0d6ecbb991d08d8a30cee70114fb5714402d9fdc90947730bd602f5045cb96c0 and confirmed
|
|
|
|
wizkid057
Legendary
Offline
Activity: 1223
Merit: 1006
|
|
December 01, 2015, 04:00:46 AM |
|
You don't like showing all the failed blocks, I do. You want the failed block information to disappear, I don't.
I didn't say I want them to disappear, I just said I didn't think they should have an effect on the stats displayed for real blocks.
|
|
|
|
kano (OP)
Legendary
Offline
Activity: 4494
Merit: 1805
Linux since 1997 RedHat 4
|
|
December 01, 2015, 04:14:26 AM |
|
You don't like showing all the failed blocks, I do. You want the failed block information to disappear, I don't.
I didn't say I want them to disappear, I just said I didn't think they should have an effect on the stats displayed for real blocks. Well then that poses the problem that the (Diff) number displayed won't add up if I change it that way. That's the reason I show it that way. If you add up the page, there's no duplicated numbers. The number on the right shows which (rare) ones are the same block to total. Edit: and since your wording is ambiguous, not it doesn't affect the stats at the top. Meh I guess if I get bored and have the time in the dark distant future I'll add a checkbox at the top "Hide the 'bad blocks'" that shows it how wizkids wants it, but with zeros for the 'bad blocks' so the numbers still add up
|
|
|
|
wizkid057
Legendary
Offline
Activity: 1223
Merit: 1006
|
|
December 01, 2015, 04:17:12 AM |
|
You don't like showing all the failed blocks, I do. You want the failed block information to disappear, I don't.
I didn't say I want them to disappear, I just said I didn't think they should have an effect on the stats displayed for real blocks. Well then that poses the problem that the numbers displayed won't add up if I change it that way. That's the reason I show it that way. If you add up the page, there's no duplicated numbers. The number on the right shows which (rare) ones are the same block to total. Sure, that's why you just adjust your asterisk to include, "Don't add this number when adding up the page." Maybe a mouseover tooltip on the values for the real block with the real values? I dunno, just irks me that it's not right for the real block. Do what you want, it's your pool. Just making a suggestion.
|
|
|
|
Rikim4ru
|
|
December 01, 2015, 10:15:42 PM |
|
I think we get it All Champion, please move to your respectful corner! Time is out.
|
|
|
|
sloopy
|
|
December 02, 2015, 05:22:55 AM |
|
Oh no, do not put a hamper on this discussion.
To me there are a few points being made which are being made by both.
Also "to me" there is one extremely important point, which is who mines empty blocks. There is no doubt in any sane persons mind that mining empty blocks is a tremendously negative action towards the entire bitcoin ecosystem which harms every user, miner, and anyone who wants bitcoin to succeed. You may as well yell that you cannot compete on a level playing field and must resort to underhanded tactics, which then must make people with common sense ask what other choices are being made? Mining 0 transaction blocks is a clear statement that you are willing to do whatever it takes to attempt to have an upperhand whether you do or not, whether it is good for all of us or not, or even if it causes longterm harm. It is a direct claim to selfishness over the good of the network and yells loud and clear you do not care about the network as much as you care about yourself.
Picking about because Kano shows us the shares which were as close to blocks as anything but are not even included in the pool's overall solved block calculations which do not even show up on the pool's luck statistics for actual blocks solved and "luck" versus someone pointing every miner on their pool to work on empty blocks as a PRIORITY, well, even if there were other reasons to say good things I haven't found, that single thing should make people pay attention. It has made several as there are many more people talking about it now than were 6 months ago. I wish I could blame it all on Luke JR as I see his influence there and know what kind of morals he has, but I sincerely had hope that others could listen to reason when it is laid out with coherent conversation.
It isn't like arguing over religion which only idiots want to do in this forum, zero transaction blocks is something negatively impacting what all of us do and come to this forum to learn more about everyday. It is a slap in the face of any respectable person in this scene.
If people think KANO.IS will always be at 2, 3,4 or so PH they are crazy. More people are seeing the payouts, the correct stats, the customer service which no other pool can touch, and you will continue to see growth 1 gh at a time. Some people do not get why it takes longer to get the payout here at KANO.IS, but as more people understand and see the fat checks roll in more will get it and stick.
Opinions and talk are one thing. It is the actions which speak the loudest from all pool operators. IF you make more money from doing shady things it is temporary. You will eventually be stabbed in the back by the very people who helped you because all they want is the quick buck and they will jump for the next thing they think is better. That is a fact of life.
|
Transaction fees go to the pools and the pools decide to pay them to the miners. Anything else, including off-chain solutions are stealing and not the way Bitcoin was intended to function. Make the block size set by the pool. Pool = miners and they get the choice.
|
|
|
|
Rikim4ru
|
|
December 02, 2015, 01:26:56 PM |
|
Oh no, do not put a hamper on this discussion.
To me there are a few points being made which are being made by both.
Also "to me" there is one extremely important point, which is who mines empty blocks. There is no doubt in any sane persons mind that mining empty blocks is a tremendously negative action towards the entire bitcoin ecosystem which harms every user, miner, and anyone who wants bitcoin to succeed. You may as well yell that you cannot compete on a level playing field and must resort to underhanded tactics, which then must make people with common sense ask what other choices are being made? Mining 0 transaction blocks is a clear statement that you are willing to do whatever it takes to attempt to have an upperhand whether you do or not, whether it is good for all of us or not, or even if it causes longterm harm. It is a direct claim to selfishness over the good of the network and yells loud and clear you do not care about the network as much as you care about yourself.
Picking about because Kano shows us the shares which were as close to blocks as anything but are not even included in the pool's overall solved block calculations which do not even show up on the pool's luck statistics for actual blocks solved and "luck" versus someone pointing every miner on their pool to work on empty blocks as a PRIORITY, well, even if there were other reasons to say good things I haven't found, that single thing should make people pay attention. It has made several as there are many more people talking about it now than were 6 months ago. I wish I could blame it all on Luke JR as I see his influence there and know what kind of morals he has, but I sincerely had hope that others could listen to reason when it is laid out with coherent conversation.
It isn't like arguing over religion which only idiots want to do in this forum, zero transaction blocks is something negatively impacting what all of us do and come to this forum to learn more about everyday. It is a slap in the face of any respectable person in this scene.
If people think KANO.IS will always be at 2, 3,4 or so PH they are crazy. More people are seeing the payouts, the correct stats, the customer service which no other pool can touch, and you will continue to see growth 1 gh at a time. Some people do not get why it takes longer to get the payout here at KANO.IS, but as more people understand and see the fat checks roll in more will get it and stick.
Opinions and talk are one thing. It is the actions which speak the loudest from all pool operators. IF you make more money from doing shady things it is temporary. You will eventually be stabbed in the back by the very people who helped you because all they want is the quick buck and they will jump for the next thing they think is better. That is a fact of life.
Sloppy, Don't get me wrong. All my hardware is pointed at Kano. Most of the reason why as been covered in this post. Its just that for a second, this thread looked like a Martial Art Forum; Both side have their conviction, and both of them will stick to it and fight for it. It was a polite 'hamper', please take no offence
|
|
|
|
elduderino
Full Member
Offline
Activity: 143
Merit: 100
Using some expensive heaters
|
|
December 02, 2015, 04:14:43 PM |
|
Yeah, baby!
|
Do the things you know you must do first, then worry about the stuff you're not sure about.
|
|
|
Nuttycoins
Newbie
Offline
Activity: 49
Merit: 0
|
|
December 02, 2015, 04:15:40 PM Last edit: December 02, 2015, 05:15:22 PM by Nuttycoins |
|
Not to change the subject... but I guess I am...
Have a question on payouts, that I can not deduct from the FAQ... I realize my small footprint it doesn't make much difference, but I am noticing something about the big boys behavior and trying to understand something completely.
Do the shifts have any bearing on the actual payout, other then where the 5N payout starts and where it ends? In other words, does it matter at all if you would mine hard in one shift and hardly nothing in the other, or you were mining in a short shift vs a long one?
The way I see it now, shifts really don't matter at all, everything is just one big average of the total shifts in the 5N.
Thanks for your time.
|
|
|
|
|