|
stbgefltc
|
 |
July 24, 2013, 01:22:55 PM |
|
Multipools, multimining - shooting a bullet in your feet...
Do you guys realize that it's gonna kill coin after coin ? Unless it's the intended goal, of course...
|
|
|
|
barwizi (OP)
Legendary
 Offline
Activity: 882
Merit: 1000
|
|
July 24, 2013, 01:26:46 PM |
|
we'll have to develop a resistance to coinchoose/coinwarz and multipools. Perhaps a readjust that pushes diff up sharply if 5 blocks are solved in under 2 minutes, by a factor of 5? that should instantly kick off the multipools. Then if a block is not solved in 20~30 minutes difficulty reverts to previous level?
|
|
|
|
|
|
stbgefltc
|
|
July 24, 2013, 01:40:06 PM |
|
we'll have to develop a resistance to coinchoose/coinwarz and multipools. Perhaps a readjust that pushes diff up sharply if 5 blocks are solved in under 2 minutes, by a factor of 5? that should instantly kick off the multipools. Then if a block is not solved in 20~30 minutes difficulty reverts to previous level?
Yeah, the core idea in the suggested solution is to give up on fixed retarget intervals. Every block becomes a potential retarget point depending on how fast/slow the last N blocks get found. N needs to be picked so it is high enough to compensate variance, but small enough to react to network spikes. |
|
|
|
|
John122
|
|
July 24, 2013, 09:33:50 PM |
|
Is it stratum only? Because it is a bitch to configure with CudaMiner... Also, I'm really not interested in multimining, so I would rather stay with the old pool.. |
|
|
|
|
feeleep
Legendary
Offline
Activity: 1197
Merit: 1000
|
|
July 27, 2013, 08:58:23 AM |
|
Is it stratum only? Because it is a bitch to configure with CudaMiner... Also, I'm really not interested in multimining, so I would rather stay with the old pool.. I can set up getwork port as well but really want to migrate everyone... |
|
|
|
|
John122
|
|
July 27, 2013, 06:58:56 PM |
|
I can set up getwork port as well but really want to migrate everyone...
Yeah, if you set up getwork I'll migrate. |
|
|
|
|
barwizi (OP)
Legendary
Offline
Activity: 882
Merit: 1000
|
|
July 28, 2013, 03:59:07 AM |
|
we'll have to develop a resistance to coinchoose/coinwarz and multipools. Perhaps a readjust that pushes diff up sharply if 5 blocks are solved in under 2 minutes, by a factor of 5? that should instantly kick off the multipools. Then if a block is not solved in 20~30 minutes difficulty reverts to previous level?
Yeah, the core idea in the suggested solution is to give up on fixed retarget intervals. Every block becomes a potential retarget point depending on how fast/slow the last N blocks get found. N needs to be picked so it is high enough to compensate variance, but small enough to react to network spikes. i think thats great, lets have a super quick diff adjust, quick to rise and quick to fall, it will force multipool and crypto switchers to make a decision. if we make adjustment every 3 blocks with ridiculous rises and immediate falls, it'll rid us of them. |
|
|
|
|
feeleep
Legendary
Offline
Activity: 1197
Merit: 1000
|
|
July 28, 2013, 04:33:47 AM |
|
I can set up getwork port as well but really want to migrate everyone...
Yeah, if you set up getwork I'll migrate. Port created. Here are the details: http://mnrb.coinmine.pl
Getwork: 8197
Stratum: 9197To all users on http://nrb.coinmine.pl - please withdraw your coins and migrate your accounts to above pool. |
|
|
|
oatmo
Member
Offline
Activity: 104
Merit: 10
|
|
July 28, 2013, 06:31:01 AM |
|
Just a note. It looks like the current 4 hour force retarget isn't working. Can we debug why we aren't getting a retarget after 4 hours. Let's get that fixed, and also add the quick retarget up based on 5 or 15 blocks, and target that for 3-4k blocks away.
Or if I am reading the block chain wrong, let me know.. We should have seen a retarget on 28988.
Oatmo
Block Approx. Time Transactions Value Out Difficulty Outstanding Average Age Chain Age % CoinDD
28994 2013-07-28 04:53:19 2 40 3.661 579860 22.8843 51.4496 21.7942%
28993 2013-07-28 04:26:37 2 40 3.661 579840 22.8666 51.431 21.808%
28992 2013-07-28 03:51:19 4 41.0932966 3.661 579820 22.8428 51.4065 21.8262%
28991 2013-07-28 01:09:40 2 40 3.661 579800 22.7314 51.2942 21.9104%
28990 2013-07-28 00:15:53 2 40 3.661 579780 22.6948 51.2569 21.9385%
28989 2013-07-27 23:13:50 2 40 3.661 579760 22.6525 51.2138 21.971%
28988 2013-07-27 22:04:37 3 1194.45011157 3.661 579740 22.6053 51.1657 22.0074%
28987 2013-07-27 19:57:34 2 40 3.661 579720 22.531 51.0775 22.029%
28986 2013-07-27 19:54:54 2 40 3.661 579700 22.5299 51.0757 22.0304%
28985 2013-07-27 18:54:23 2 40 3.661 579680 22.4887 51.0336 22.0625%
28984 2013-07-27 17:56:43 2 40 3.661 579660 22.4494 50.9936 22.0931%
28983 2013-07-27 17:51:17 2 123.00550875 3.661 579640 22.4464 50.9898 22.096%
28982 2013-07-27 17:36:49 4 160.32731993 3.661 579620 22.4378 50.9798 22.1016%
28981 2013-07-27 17:34:05 1 20 3.661 579600 22.4372 50.9779 22.1011%
28980 2013-07-27 17:33:25 1 20 3.661 579580 22.4375 50.9774 22.1015%
|
|
|
|
|
oatmo
Member
Offline
Activity: 104
Merit: 10
|
|
July 28, 2013, 06:41:06 AM |
|
we'll have to develop a resistance to coinchoose/coinwarz and multipools. Perhaps a readjust that pushes diff up sharply if 5 blocks are solved in under 2 minutes, by a factor of 5? that should instantly kick off the multipools. Then if a block is not solved in 20~30 minutes difficulty reverts to previous level?
We have to be careful using single block times to do anything, because this is a random process. I like the 5 blocks in less than 2 minutes causing a retarget. How about the forced retarget is 2 hours, or less than 12 blocks in 1 hour, or alternatively forced at 2 hours, or less than 6 blocks in 30 minutes. Currently we are looking at the last 30 blocks for the retarget, I propose in the 5 blocks in 2 minutes, or <6 blocks in 30 minutes that we use the last 15 blocks to calculate the new difficulty. This will dampen the effect and reduce overshoots. |
|
|
|
|
|
stbgefltc
|
|
July 28, 2013, 11:02:49 AM |
|
Yeah, current implementation is faulty on forced retarget. It will only be triggered if one block takes more than 4 hours to find.
Fix is on its way, haven't had time before.
|
|
|
|
feeleep
Legendary
Offline
Activity: 1197
Merit: 1000
|
|
July 28, 2013, 08:08:21 PM |
|
My both pools lost connections - what is happening? Can anyone provide some addnodes?
|
|
|
|
snaidervp
Sr. Member
Offline
Activity: 333
Merit: 250
"Raven's Cry"
|
|
July 29, 2013, 06:42:52 PM |
|
i cant sync my wallet either  |
yawn
|
|
|
|
John122
|
|
July 29, 2013, 07:04:31 PM |
|
i cant sync my wallet either It's been an hour and my transaction still has 0 confirmation...  |
|
|
|
|
barwizi (OP)
Legendary
Offline
Activity: 882
Merit: 1000
|
|
July 29, 2013, 07:15:01 PM |
|
mine is synced, but i'm guessing this is due to there being very few mining. I'll be looking at the diff code tonight and offer my idea later.
|
|
|
|
|
snaidervp
Sr. Member
Offline
Activity: 333
Merit: 250
"Raven's Cry"
|
|
July 29, 2013, 07:30:59 PM |
|
|
yawn
|
|
|
barwizi (OP)
Legendary
Offline
Activity: 882
Merit: 1000
|
|
July 31, 2013, 04:23:15 PM |
|
Algo update incoming, give me two days. I am thinking that a per block re-target is best instant high and instant low, and in the event of 20min no block solved, it drops to previous diff. it forces you to either stay or leave. The multi mine pools will leave us alone after they notice this.
|
|
|
|
|
|
John122
|
|
July 31, 2013, 06:35:50 PM |
|
Algo update incoming, give me two days. I am thinking that a per block re-target is best instant high and instant low, and in the event of 20min no block solved, it drops to previous diff. it forces you to either stay or leave. The multi mine pools will leave us alone after they notice this.
Hurray!  The multimine pool has kinda been killing the coin recently... |
|
|
|
|
|
jasinlee
|
|
July 31, 2013, 06:56:46 PM |
|
The beauty of which has allowed others to buy up cheap nrb  |
|
|
|
|
John122
|
|
July 31, 2013, 07:12:51 PM |
|
The beauty of which has allowed others to buy up cheap nrb True that!  |
|
|
|
|
|
